Exploring the thermodynamics and conformational aspects of nicotinic acid binding with bovine serum albumin: a detailed calorimetric, spectroscopic and molecular docking study

Abstract

The present study reports comprehensive energetic and conformational aspects of the binding of an antihyperlipidemic drug, nicotinic acid (NA), with a model transport protein, bovine serum albumin (BSA) by calorimetry, light scattering, spectroscopic (absorption, fluorescence, ¹H-NMR, and circular dichroism) and molecular docking methods. The calorimetric result reveals that NA binds to BSA in a sequential way with a stronger affinity (∼10⁴ M⁻¹) for the first binding site. The study in the presence of various co-solutes (salt, tetrabutylammonium bromide, sucrose, and surfactants) indicates the significant contribution of electrostatic as well as hydrophobic interactions but insignificant contribution of hydrogen bonding to the binding process. In addition, NA was also observed to bind with BSA through π–π interactions as revealed by ¹H-NMR and the molecular docking study. The spectroscopic analysis reveals the formation of a complex via a static quenching mechanism. The presence of two sequential binding events has been successfully explained by calorimetry which has also been supported by the fluorescence study. The changes in the size as well as in the secondary structure of BSA were observed upon binding with NA. The stronger binding of NA at Sudlow site I (subdomain IIA) of BSA has been explored by the molecular docking study in combination with specific site probe experiments. Casting light on such drug–protein interactions helps in better understanding the biomolecular recognition and opens up new approaches in rational drug-design processes.

---

1. Introduction

Understanding and controlling drug–protein binding interactions may help us to modulate the pharmacokinetics and pharmacodynamics of a drug. Once administered, the drugs have the ability to bind with various blood constituents like cells, proteins, etc. The biological activity of a drug is significantly affected by the nature and magnitude of such drug–protein interactions. In the blood plasma, serum albumin is the most abundant protein that can bind, transport, and metabolize various endogenous as well as exogenous substances to their target positions.^1,2 Other physiological functions of serum albumin involve its contribution to osmotic blood pressure and the pH of the blood, in sequestering oxygen free radicals, and in inactivating various toxic metabolites.³ The importance of drug–protein binding study lies not only to determine the biological effect of a drug but it can also provide necessary information related to the therapeutic effect of drugs in pharmacology and pharmacodynamics. The information about the exact location of drug binding site on protein can make us better understand the way of distribution of a drug in the body and its use with other drugs and competitive natural catabolites such as bilirubin.⁴ Among the family of serum albumins, bovine serum albumin (BSA, Fig. 1A (ref. 5)) is a broadly studied model protein not only because of the presence of diverse binding sites and its structural resemblance with human serum albumin⁶ but also because of its medicinal importance, abundance, low cost, easy availability, remarkable ligand binding properties, and wide acceptance in pharmaceutical industries.⁷ BSA, a globular protein having a single polypeptide chain, comprises of 583 amino acid residues along with 17 disulfide bridges and one free thiol group.⁸ The presence of hydrophobic binding pockets on serum albumin increases the apparent solubility of the hydrophobic drugs in the blood plasma and also modulates their in vivo delivery to the cells.⁹ The presence of hydroxyl, carboxyl or other reversible binding sites on the amino acid residues present in the binding pocket of protein enable it to reversibly bind with the drugs through weak interactions such as ionic, van der Waals, hydrogen bonding and hydrophobic interactions.¹⁰ In some cases, the various administered drugs not only bind with serum albumin but also affects the function of protein during blood transportation process by altering its conformation.^11,12 In literature, many studies on the binding interactions among BSA and various ligands have been successfully reported.^13–18

Fig. 1 (A) Structure of BSA showing different subdomains. (B) Structure of nicotinic acid (NA). The aromatic protons are labelled as H₁, H₂, H₃, and H₄ to explain ¹H-NMR titrations.

Vitamin B₃, nicotinic acid (NA), also known as pyridine 3-carboxylic acid (Fig. 1B) belongs to water soluble vitamin B family. It is a precursor of coenzymes NAD (nicotinamide adenine dinucleotide) and NADP (nicotinamide adenine dinucleotide phosphate) in vivo, which play an important role in catabolism and anabolism reactions in the body, respectively. The lack of NA causes pellagra and Alzheimer's disease.¹⁹ It is a broad spectrum antihyperlipidemic drug used for the treatment of hyperlipidemic disorders. At pharmacological doses, NA acts as a potent lipid modifying drug capable of reducing low density lipoprotein cholesterol levels along with triglycerides, apolipoprotein B, and lipoprotein (a) levels but increases the level of high density lipoprotein cholesterol as well as apolipoprotein A-I.^20,21 Hence it is also used to treat cardiovascular diseases. Besides these benefits, the main side effect of NA at high doses is cutaneous flushing caused due to prostaglandin activity along with decreased glucose tolerance and negative gastrointestinal effects.^22,23 Thus the study of such pharmaceutical interactions with serum albumin opens up new ways in the drug design and development process.

To the best of our knowledge, no report on the quantitative and qualitative analysis of the binding of NA with BSA has been found in literature. The present study is focused to characterize the binding interactions, conformational change in BSA and the location of exact binding site for NA on BSA via means of various appropriate techniques like isothermal titration calorimetry (ITC), dynamic light scattering (DLS) in combination with UV-vis, steady-state as well as three dimensional fluorescence, circular dichroism (CD), ¹H-NMR spectroscopic and molecular docking study. This study may provide an accurate and comprehensive valuable information regarding the binding mechanism and the exact position of NA on BSA which is useful as a reference in drug discovery and development process.

2. Experimental

2.1. Materials

Bovine serum albumin (catalog no. B-4287, purity ≥ 98%) and nicotinic acid (purity ≥ 99%) were purchased from Sigma-Aldrich Chemical Company and Sisco Research Laboratories, India, respectively. The drugs cefotaxime sodium (purity ≥ 98%) and ibuprofen (purity ≥ 98%) were used as procured for site marker study from Sigma-Aldrich Chemical Company. Milli-Q water, having specific resistance of 18.2 mega ohm cm, was used to prepare 0.01 M phosphate buffer of pH 7.4. The stock solution of BSA was prepared first by dissolving it in phosphate buffer (0.01 M, pH 7.4) and then overnight dialyzed against proper buffer at 4 °C. The solution of NA was prepared by directly dissolving it in dialysate phosphate buffer (0.01 M, pH 7.4). Systronics μ pH system 362 was used to measure the pH of dialysate. The mass measurements were carried out on Mettler Toledo AB 265-S balance with a resolution of 0.01 mg. Shimadzu-1800 UV-vis spectrophotometer was used to determine the concentration of BSA using molar absorption coefficient value ^1%A_{1 cm} = 6.8 at 280 nm.²⁴ All other reagents used were of analytical grade and used as such without prior treatment but dried in vaccum desiccator before use.

2.2. Methods

2.2.1. Isothermal titration calorimetry (ITC). An isothermal titration calorimeter (iTC₂₀₀, MicroCal, USA) has been used to thermodynamically characterize the binding properties of BSA and NA. The reference cell was loaded with phosphate buffer and the sample cell of capacity 200 μl was filled with 0.061 mM BSA solution or with 0.01 M phosphate buffer. The titrations of NA into BSA were started after equilibrating the calorimeter to obtain baseline stability. The 9.15 mM NA solution was loaded into 40 μl syringe of the calorimeter. In order to ensure rapid and complete mixing of NA–BSA solution, the contents in the sample cell were thoroughly stirred with a stirring speed of 500 rpm. By measuring the enthalpy change at each injection, the raw data, consisting of a series of peaks, were obtained as a plot of heat flow (kJ s⁻¹) against time (min). A plot of observed enthalpy change per mole of injectant (kJ mol⁻¹) against molar ratio was obtained after integration of raw data peaks using Origin 7 software of the instrument. In order to correct the data from dilution effects, the control experiments included the titrations of titrant into buffer, buffer into protein, and buffer into buffer were performed using same concentration of titrant (ligand) and protein, and subtracted from the experimental data.

Keeping in mind the property of serum albumin to provide diverse binding sites to the ligands, different binding site models have been tried to fit the experimental data points. The reduced chi-square value and/or least error associated with the binding parameters were used to establish the best fit. Based upon the above fact, two sequential binding site model was adopted to fit the experimental ITC data.

2.2.2. Molecular docking study. The optimisation of NA was carried out using Gaussian 09. The docking experiments were performed using the GLIDE docking protocol in the Schrodinger 2015. The native structure of BSA (PDB id: 40R0, a familiar co-crystallised with naproxen) obtained from the Protein Data Bank (http://www.rcsb.org/pdb) at a resolution of 2.58 Å was constructed using GLIDE package to perform the protein preparation wizard, followed by binding site prediction using the naproxen binding sites, as per Lamarckian Genetic Algorithm (LGA) docking. The grid size along X, Y, and Z-axes for both sites were set to 60 Å. Supplied coordinates for site I and site II are X = −2.97, Y = 33.44, Z = 102.24 and X = 1047, Y = 21.35, and Z = 123.35, respectively. The obtained docked conformations were visualised using Pymol and Chimera software package. For the further analysis, the conformation having lowest binding free energy was used. The MM-GBSA, QM/MM, and IFD calculations were carried out using binding energy estimation protocol, Qsite, and Glide docking modules.

2.2.3. Dynamic light scattering (DLS) measurements. DLS measurements were carried out at a scattering angle of 173° using Zetasizer NanoZS (Malvern, instruments, U.K) equipped with a He–Ne laser having 4 mW power at 632.8 nm. The temperature of the measurements was maintained at 298.15 K by using built in temperature controller having an accuracy of ±0.1 K. The DLS measurements were performed in disposable polystyrene cuvettes of 1 cm path length. Prior to scanning, the purification process was carried out by centrifugation at 10 000 rpm for 12 min followed by filtration using 0.22 and then by 0.02 μm pore sized syringe filters. In a typical experiment, 2 ml BSA solution was added into the disposable polystyrene cuvette and a total of 6 measurements were performed after each successive addition of suitable aliquots of NA solution. The translational diffusion coefficient (D₀) through the Stoke's–Einstein equation was used to calculate hydrodynamic diameter (d_h) of the particles:

d_h = k_BT/3πηD₀

where k_B is the Boltzmann constant, T is the absolute temperature, and η is the viscosity of the medium.

2.2.4. UV-visible absorption spectroscopy. The absorption spectroscopic measurements were performed with Shimadzu-1800 UV-visible spectrophotometer equipped with quartz cuvettes having 1 cm path length. The concentration of BSA was fixed to 0.01 mM and the titrations were carried out using successive additions of NA solution.

2.2.5. Fluorescence spectroscopy. Hitachi F-4600 spectrophotometer was used to monitor all the steady-state fluorescence measurements at an excitation wavelength of 280 nm. The excitation and emission slit widths of 5 nm and PMT of 400 were used to record the fluorescence emission spectra of BSA. The concentration of BSA was fixed at 0.01 mM and successive titrations of NA solution into BSA were carried out at 298.15 K. The emission spectra were scanned in the excitation wavelength range of 295 to 420 nm. In order to analyze the data, the intensity values were corrected from inner filter effect using the following equation:

I_corr = I_obs × 10^{(A_ex+A_em)/2}

where I_obs and I_corr represent the observed and corrected intensity values, respectively. A_ex and A_em are the absorbance values at the excitation and emission wavelengths, respectively.

The three dimensional fluorescence measurements were carried out under the following conditions: excitation and emission range of 200–600 nm with an increment of 5 nm. All other parameters were kept same as in steady state fluorescence.

2.2.6. Proton nuclear magnetic resonance (¹H-NMR) spectroscopy. ¹H-NMR spectroscopic measurements were recorded on Brüker 500 MHz (AVANCE III HD) spectrometer at 298.15 K. All the solutions for NMR measurements were prepared in phosphate buffer–D₂O ([D₂O] = 10%, v/v) as solvent. The 0.6 ml of the sample was taken in 5 mm tubes and the solvent signal was suppressed before recording NMR spectra. All the chemical shifts were measured on δ scale.

2.2.7. Circular dichroism (CD) spectroscopy. CD measurements of BSA in the absence and presence of NA were carried out on Jasco J-810 spectropolarimeter at 298.15 K. The far UV-CD spectra were obtained over the wavelength range of 190–260 nm with quartz cell having 0.2 cm path length at a scanning speed of 100 nm min⁻¹. An average of three accumulations was taken for each CD spectra under constant nitrogen flush. The spectra were recorded as CD ellipticity in mdeg after baseline subtraction for buffer solution.

3. Results and discussion

3.1. Calorimetric study of the thermodynamics of NA binding

In order to gain deeper understanding regarding the molecular interactions that drive the formation of complex, the prior knowledge about the thermodynamics of the binding is must. Isothermal titration calorimetry (ITC) is an efficient and valuable tool to shed light on the thermodynamics of the binding interactions. The temperature dependence calorimetric experiments were conducted at 288.15, 298.15, 308.15, and 313.15 K for NA binding with BSA. The ITC profile and the corresponding binding isotherm (corrected for all dilution effects) at 298.15 K is presented in Fig. 2A and B, respectively. The integrated heat profile data were fitted using two sequential non-identical binding site model. The values of enthalpy (ΔH), entropy (ΔS), Gibbs free energy (ΔG), and binding constant (K), listed in Table 1, have been obtained after correcting the experimental ITC data from dilution effects. The listed values in Table 1 are the average of two to three independent experiments. The binding of NA with BSA was described by exothermic heat profile with the stronger binding (∼10⁴ M⁻¹) at first site compared to the second site. With rise in temperature from 288.15 to 313.15 K, the decrease in the binding constant (K₁), increase in the exothermicity (ΔH₁), and unfavorable entropy (ΔS₁) have been observed for the stronger binding site. The thermodynamic parameters suggest the larger contribution of enthalpy factor compared to entropy factor in making the binding process to be spontaneous at higher temperatures. The negative value of ΔH₁ indicates the existence of electrostatic interactions among the binding components.

Fig. 2 (A) ITC raw profile, and (B) the corresponding integrated heat profile for the binding of NA with BSA at 298.15 K.

Table 1 Binding constant and thermodynamic parameters for the binding of NA with BSA at various temperatures

T (K)	K1, (M^−1)	K2, (M^−1)	ΔH1, (kJ mol^−1)	ΔH2, (kJ mol^−1)	ΔS1, (J mol^−1 K^−1)	ΔS2, (J mol^−1 K^−1)	ΔG1, (kJ mol^−1)	ΔG2, (kJ mol^−1)
288.15	(2.80 ± 0.08) × 10^4	(310 ± 6.40)	(−20.20 ± 0.22)	(−164.90 ± 2.93)	(15.10 ± 0.53)	(−524 ± 9.99)	(−24.53 ± 0.06)	(−13.74 ± 0.05)
298.15	(1.41 ± 0.02) × 10^4	(234 ± 2.70)	(−29.65 ± 0.20)	(−255.30 ± 2.89)	(−20.00 ± 0.55)	(−810 ± 9.59)	(−23.68 ± 0.04)	(−13.52 ± 0.03)
308.15	(1.18 ± 0.02) × 10^4	(294 ± 4.80)	(−34.57 ± 0.36)	(−274.00 ± 4.05)	(−34.20 ± 1.03)	(−842 ± 13.00)	(−24.02 ± 0.04)	(−14.56 ± 0.04)
313.15	(1.08 ± 0.02) × 10^4	(264 ± 4.70)	(−40.52 ± 0.48)	(−343.90 ± 5.59)	(−52.10 ± 1.38)	(−1050 ± 17.70)	(−24.18 ± 0.05)	(−14.52 ± 0.05)

---

The study of temperature dependence of enthalpy provides information about the change in heat capacity (ΔC_p = (∂ΔH/∂T)_p) of the binding process. The plot of ΔH₁ vs. T (Fig. 3) for the first site gives a straight line with correlation coefficient of 0.9898 and the value of slope yields ΔC_p = (−767.39 ± 77.81) J K⁻¹ mol⁻¹. In aqueous solutions, the change in heat capacity can be linked with the change in the well orderly arranged water molecules present in the hydrophobic binding pocket of protein. The observed negative ΔC_p value for the studied binding system pinpoint the transfer of hydrophobic groups into the interior binding pocket of BSA which leads to the liberation of such orderly arranged water molecules from hydrophobic pocket to the bulk solution. Several other factors like change in the conformation and decrease in the degree of freedom of BSA upon NA binding may also be responsible for negative ΔC_p value.²⁵ The contribution of hydrophobic transfer step (ΔG_hyd) to the free energy of binding may be determined using the following relation:^25,26

ΔG_hyd = (80 ± 10) × ΔC_p

Fig. 3 Plot showing temperature dependence of calorimetric enthalpy for NA–BSA binding system at pH 7.4.

The large negative value of ΔG_hyd = (−61.39 ± 8.57) kJ mol⁻¹ obtained from the above relation suggests the involvement of hydrophobic interactions in the binding process.

Enthalpy–entropy compensation (EEC) phenomenon, associated with reorganization of the solvent molecules, is an important effect that arises due to internal compensation of enthalpy and entropy factors. Several studies on ligand–protein interactions that exhibit such EEC have been reported previously.^27–29 A linear relationship among ΔH₁ and TΔS₁ (Fig. 4) with slope = 0.9827 has been obtained for NA binding with BSA for first binding site. The invariant free energy (ΔG₁) of the system with rise in temperature (Table 1) has been observed due to strong EEC. Such kind of complete EEC is observed in the systems having ΔC_p ≠ 0 and ΔC_p > ΔS₁ and this is true in the present case. In the present system, favorable ΔH₁ values compensate unfavorable ΔS₁ values at higher temperatures. According to the thermodynamic model proposed by Ross et al.,³⁰ the various factors like (1) disruption of hydrogen-bonded water networks from the binding pocket of BSA and/or from around NA, (2) the existence of electrostatic, hydrogen bonding, and van der Waals interactions among the binding components, lead to the evolvement of heat during the binding process. The conformational restrictions (decrease in vibrational and rotational degree of freedom) during the binding process were responsible for the outcome of unfavorable entropy values.³¹

Fig. 4 Plot showing enthalpy–entropy compensation for NA–BSA binding system at pH 7.4.

In order to assay two state binding dictates, van't Hoff enthalpy (Δ_vHH°) values were determined using following equation and compared with calorimetric enthalpy (ΔH₁) values for first binding site.

For binding of NA with BSA, inconsistency among enthalpy values has been observed on comparing Δ_vHH° with ΔH₁. The value of Δ_vHH° at 298.15 K comes out to be −9.81 kJ mol⁻¹ which differs from ΔH₁ value. Such type of inconsistency among enthalpy values has been reported in literature³² and this indicates the absence of two state binding behaviour. The results may point towards the conformational change in protein during binding process which may be interpreted in terms of either with increase in temperature or with NA binding. In order to have a better look on the conformational change in protein due to NA binding, circular dichroism and three dimensional fluorescence spectroscopic experiments have been performed.

3.1.1. Effect of co-solutes on the binding interactions among NA and BSA. The contribution of electrostatic, hydrophobic, and hydrogen-bonding interactions in the binding process were assessed in the presence of various co-solutes.
3.1.1.1. Effect of salt. In order to assess the role of ionic interactions to the binding process, the effect of salt on the bimolecular association is often studied. The strength of ionic interactions weakens through the screening mechanism and hence its contribution to the free energy of binding also decreases only if electrostatic interactions play a meaningful role in the binding process.³³ Therefore, to assay the contribution of electrostatic interactions to the binding process, the binding experiments were conducted in the presence of NaCl at 0.05, 0.2, and 0.33 M concentrations at pH 7.4 and 298.15 K (Fig. S1†). The corresponding binding parameters are summarized in Table 2. The K₁ (high affinity binding constant) value changes from (1.41 ± 0.02) × 10⁴ M⁻¹ in the absence of NaCl to (1.12 ± 0.06) × 10⁴, (6.43 ± 0.03) × 10³, and (4.85 ± 0.29) × 10³ M⁻¹ in the presence of 0.05, 0.2, and 0.33 M NaCl, respectively. The decrease in K₁ value is due to the prior release of counterions followed by the binding of charged ligand.

Table 2 Binding constant and thermodynamic parameters for the binding of NA with BSA in the presence of co-solutes at 298.15 K

Co-solute	Conc. (M)	K1, (M^−1)	K2, (M^−1)	ΔH1, (kJ mol^−1)	ΔH2, (kJ mol^−1)	ΔS1, (J mol^−1 K^−1)	ΔS2, (J mol^−1 K^−1)	ΔG1, (kJ mol^−1)	ΔG2, (kJ mol^−1)
Blank	—	(1.41 ± 0.02) × 10^4	(234 ± 2.70)	(−29.65 ± 0.20)	(−255.30 ± 2.89)	(−20.00 ± 0.55)	(−810 ± 9.59)	(−23.68 ± 0.04)	(−13.52 ± 0.03)
	0.05	(1.12 ± 0.06) × 10^4	(193 ± 9.50)	(−17.13 ± 0.51)	(−225 ± 10.00)	(20.10 ± 1.26)	(−711 ± 33.13)	(−23.11 ± 0.13)	(−13.05 ± 0.12)
NaCl	0.2	(6.43 ± 0.03) × 10^3	(254 ± 9.90)	(−15.27 ± 0.49)	(−179.40 ± 6.27)	(21.70 ± 1.60)	(−556 ± 20.17)	(−21.74 ± 0.02)	(−13.73 ± 0.09)
	0.33	(4.85 ± 0.29) × 10^3	(271 ± 12.00)	(−14.30 ± 0.67)	(−163.50 ± 6.59)	(22.60 ± 1.75)	(−502 ± 21.73)	(−21.04 ± 0.15)	(−13.89 ± 0.11)
TBAB	0.02	(8.32 ± 0.35) × 10^3	(178 ± 6.70)	(−23.00 ± 0.56)	(−230.90 ± 8.71)	(−2.03 ± 1.52)	(−731 ± 28.90)	(−22.37 ± 0.11)	(−12.85 ± 0.09)
	0.05	(6.78 ± 0.25) × 10^3	(252 ± 8.40)	(−20.91 ± 0.49)	(−176.60 ± 5.43)	(3.25 ± 1.34)	(−546 ± 17.93)	(−21.87 ± 0.09)	(−13.71 ± 0.08)
Sucrose	0.5	(1.26 ± 0.02) × 10^4	(347 ± 4.30)	(−20.17 ± 0.18)	(−126.60 ± 1.53)	(10.90 ± 0.47)	(−376 ± 5.03)	(−23.40 ± 0.04)	(−14.49 ± 0.03)
SDS	4.5	(2.39 ± 0.69) × 10^3	(1.49 ± 0.75) × 10^3	(115.70 ± 25.40)	(−136.70 ± 25.50)	(453 ± 87.66)	(−398 ± 80.93)	(−19.28 ± 0.74)	(−18.11 ± 1.39)
	8.6	(582 ± 140)	(4.54 ± 0.87) × 10^3	(219.80 ± 57.40)	(−81.81 ± 12.00)	(790 ± 194.56)	(−204 ± 38.64)	(−15.78 ± 0.62)	(−20.87 ± 0.48)
HTAB	0.34	(9.76 ± 0.34) × 10^3	(351 ± 9.60)	(−17.98 ± 0.36)	(−146.50 ± 3.59)	(16.10 ± 0.92)	(−442 ± 11.81)	(−22.77 ± 0.09)	(−14.53 ± 0.07)
TX-100	0.1	(6.33 ± 0.32) × 10^3	(227 ± 7.90)	(−17.67 ± 0.62)	(−249.30 ± 8.01)	(13.60 ± 1.66)	(−791 ± 26.58)	(−21.69 ± 0.13)	(−13.45 ± 0.09)
	0.31	(7.40 ± 0.22) × 10^3	(213 ± 6.20)	(−24.94 ± 0.46)	(−251.80 ± 6.98)	(−9.53 ± 1.29)	(−800 ± 23.17)	(−22.08 ± 0.08)	(−13.29 ± 0.07)

---

Record et al.³⁴ proposed a linear dependence relationship between the binding constant ‘K₁’ and [Na⁺] ion concentration as:

∂(log K₁)/∂(log[Na⁺]) = −(Zψ)₁

where ‘Z’ is the apparent charge of the bound ligand and ψ is the fraction of sodium counterions bound per BSA molecule. The (Zψ)₁ indicates the number of counterions released upon ligand binding from BSA. The slope of the plot of log K₁ vs. log[Na⁺] gives the value of (Zψ)₁ = 0.43. This value represents the number of sodium counterions released from BSA upon NA binding and suggests the existence of electrostatic interactions in the binding system. The partitioning of polyelectrolyte (ΔG_pe)₁ and non-polyelectrolyte (ΔG_npe)₁ contributions to the observed Gibbs free energy (ΔG₁) (Fig. 5) for the first binding site has been further used to verify the presence of electrostatic interactions in the binding process. The (Zψ)₁ value can be used to determine polyelectrolyte (ΔG_pe)₁ contributions to ΔG₁ by the following relationship:

(ΔG_pe)₁ = (Zψ)₁RT ln[Na⁺]

Fig. 5 Partitioning of ΔG₁ (blue) into (ΔG_pe)₁ (red) and (ΔG_npe)₁ (green) at different concentrations of NaCl at 298.15 K.

This value of (ΔG_pe)₁ was further used to calculate (ΔG_npe)₁ (= ΔG₁ − (ΔG_pe)₁), where ΔG₁ = −RT ln K₁, ‘K₁’ is the binding constant for first site obtained from calorimetry in the presence of NaCl. The (ΔG_npe)₁ is the free energy contribution arising from the non-polyelectrolyte forces (i.e. hydrogen bonding, van der Waals, and hydrophobic interactions) that stabilize the drug–protein complex. The value of (ΔG_pe)₁ corresponds to the polyelectrolyte forces (i.e. ionic interactions) involved in the binding process. For all concentrations of NaCl, the observed large value of (ΔG_npe)₁ as compared to (ΔG_pe)₁ (Table 3) highlights the vital role of non-polyelectrolytic forces in the stabilization of complex. The decrease in the magnitude of (ΔG_pe)₁ with increase in the concentration of NaCl underscores the contribution of polar interactions in the binding process.

Table 3 The values of different free energies (ΔG₁, (ΔG_pe)₁, and (ΔG_npe)₁) and the number of counterions released upon NA binding (Zψ)₁ for NA–BSA binding system at first site in different NaCl concentrations at 298.15 K

NaCl (M)	ΔG1 (kJ mol^−1)	(Zψ)1	(ΔGpe)1 (kJ mol^−1)	(ΔGnpe)1 (kJ mol^−1)
0.05	(−23.11 ± 0.13)	(0.43 ± 0.03)	(−3.19 ± 0.22)	(−19.92 ± 0.22)
0.2	(−21.74 ± 0.02)	(0.43 ± 0.03)	(−0.41 ± 0.03)	(−21.33 ± 0.03)
0.35	(−21.04 ± 0.15)	(0.43 ± 0.03)	(−0.28 ± 0.02)	(−20.76 ± 0.30)

---

3.1.1.2. Effect of tetrabutylammonium bromide (TBAB). An assessment of both hydrophobic as well as electrostatic interactions involved in the binding process is generally analyzed by using TBAB as a co-solute. The four bulky butyl groups present on TBAB along with its ionic nature affects the hydrophobic as well as electrostatic interactions existing among binding components.³⁵ Hence in order to better understand the role of both hydrophobic and electrostatic interactions, the binding of NA with BSA was observed at two TBAB concentrations (0.02 and 0.05 M). The ITC binding profile and the corresponding data at 298.15 K are displayed in Fig. S1 (ESI†) and Table 2, respectively. With increase in the concentration of TBAB from 0.02 to 0.05 M, the value of K₁ reduces from (1.41 ± 0.02) × 10⁴ M⁻¹ (in the absence of TBAB) to (8.32 ± 0.35) × 10³ and (6.78 ± 0.25) × 10³ M⁻¹, respectively. The reduction in K₁ value with increase in the concentration of TBAB may be attributed to the larger interference of TBAB in both hydrophobic as well as electrostatic interactions. The comparison of K₁ value both in the presence of 0.05 M TBAB and NaCl shows the larger decrease in the presence of former as compared to latter. Thus, the above results highlight the contribution of hydrophobic interactions in the binding process.
3.1.1.3. Effect of sucrose. To quantify the contribution of hydrogen bonding to the binding process, sucrose has been used as co-solute.^17,24 Any change in binding parameters can be associated with the alterations in the hydrogen bonding network existing either among the reacting species or among solvent molecules. This is due to the interference of several –OH groups present on sucrose in the binding process. The binding experiment was monitored in the presence of 0.5 M sucrose at pH 7.4 and 298.15 K (Fig. S1†). The value of K₁ = (1.41 ± 0.02) × 10⁴ M⁻¹ in the absence of sucrose changes to (1.26 ± 0.02) × 10⁴ M⁻¹ in the presence of 0.5 M sucrose. Table 2 shows that although K₁ value remains almost same but the binding becomes less exothermic and entropy becomes positive for the first binding site in the presence of sucrose. Such changes in ΔH₁ and ΔS₁ values indicate the reorganization of the solvent structure in such a way that decreases the extent of hydrogen-bonding among the reacting species. The positive increase in ΔS₁ value i.e. the increase in randomness of the system in the presence of sucrose further supports the above fact.
3.1.1.4. Effect of surfactants. As mentioned above, TBAB has been used to explore non-columbic interactions but it also affects the polar interactions present among NA and BSA. Thus in order to further explore the role of non-columbic interactions to the binding process, the binding was studied in the presence of surfactants at pH 7.4 and 298.15 K. The surfactants modify the structure of protein by partial denaturation which decreases/increases the ligand binding sites on protein.³² Fig. S2 (ESI†) and Table 2 shows the ITC raw profile and corresponding ITC data, respectively, for NA binding with BSA in the presence of anionic, cationic, and non-ionic surfactants. The decrease in K₁ and increase in K₂ value has been observed in the presence of anionic surfactant, sodium dodecyl sulphate (SDS), both above and below its CMC (CMC = 8.2 mM).³⁶ The anionic nature of SDS interferes in the electrostatic interactions via blocking the binding sites for NA present on BSA. At the studied concentrations, the ability of SDS to partially denature the protein³⁷ may also be responsible for such change in K (K₁ and K₂) values. The denaturing property of SDS may modify the number and virtue of the binding sites on protein. The observed ΔH₁ value indicates the loss of electrostatic interactions and suggests that the obtained value of K₁ is only due to the contribution from hydrophobic forces present in the binding process.

Cationic surfactant, hexadecyltrimethylammonium bromide (HTAB), leads to the decrease of K₁ value to (9.76 ± 0.34) × 10³ M⁻¹ below its CMC (CMC = 0.91 mM).³⁶ Being positively charged, HTAB does not interfere in the electrostatic interactions but can influence the non-polar interactions existing among NA and BSA. A usual binding profile observed in the presence of HTAB clearly indicates no loss of the virtue of the binding sites on BSA at the studied concentration of HTAB. The interference to the non-polar interactions among binding components may result in the decrease of K₁ value. The presence of triton X-100 (TX-100) both below and above its CMC (CMC = 0.261 mM)³⁸ show marked effect on the binding of NA with BSA. The neutral nature of TX-100 is responsible for its interference in polar as well as non-polar interactions existing among NA and BSA. Below CMC, it decreases the value of K₁ and makes the process enthalpically less favorable and entropically more favorable. The values of ΔH₁ and ΔS₁ indicate the decrease in the extent of electrostatic interactions. But at higher concentration (i.e. above CMC), the slight increase in the value of K₁ has been observed compared to the value obtained at lower concentration (i.e. below CMC). Such increase in K₁ value may be described due to the conformational change of protein in the presence of non-ionic surfactant above its CMC. The binding study in the presence of HTAB and TX-100 clearly points towards the significant contribution of hydrophobic interactions to the binding process.

3.2. Unraveling specific binding sites for NA on BSA

The molecular docking study is a widely accepted technique to understand the preferred binding location and binding interactions of ligand binding with macromolecules. The heart shaped serum albumin consists of three homologous α-helical domains (I, II, III; I (residues 1–195), II (196–383), III (384–585)). Each domain comprehends two subdomains (A and B) that segment common structural motif's.⁵ The hydrophobic cavities in subdomain IIA (Sudlow site I) and IIIA (Sudlow site II), known as warfarin and benzodiazepine binding sites, respectively, are the principal regions of ligand binding with BSA.^39–41 In the present work, the different possible conformations of NA that binds with BSA has been determined using Glide program. The preparation of NA for docking was accomplished using LigPrep module⁴² for the generation of possible states of NA between pH 2–7 at OPLS 2005 force field. The best energy ranked results are tabulated in Table S1† for both extra precision (XP mode) and standard precision (SP mode) docking considering epik state penalties for NA at both sites of BSA (Fig. 6). In addition, molecular docking study shows the weak interactions of NA at subdomain IIB of BSA. The multiple binding sites analysis (sequential docking) was carried out at poses obtained from induced fit docking protocol. Fig. 6 indicates that NA interacts hydrophobically with Tyr149, Leu218, Phe222, Ala290, Ile289, Arg217 and 256, Leu259 and 237, Ile263, Lys221, and His287 present at subdomain IIA (Sudlow site I) of BSA. Moreover, the formation of two hydrogen bonds: one with Ser286 (bond length 2.49 Å) and other with Arg256 (bond length 2.36 Å) has been observed from docking study. The possibility of π–π interaction with the Phe222 and Trp213 was also explored by further optimisation of the minimized protein–ligand complex using the induced fit docking protocol (IFD). The IFD score of NA at site I was 1281.07 and at site 2 was 1279.89. The distance of NA with Phe222 was found out to be 2.91 Å (within the conventional limit of the ArCH/π-interaction⁴³) and with Trp213, it was found out to be 8.1 Å, whereas IFD pose depicted the additional H-bond with side chain of Lys221 at site 1. The further exploration of the interaction between NA and Phe222 was carried out using trimmed protein residues (consideration of the amino acid residue within 10 Å distance) for attaining reasonable computational cost at the higher basis set. The IFD obtained pose and QM/MM optimised pose depicted the possibility of ArCH–π interaction between the Phe222 and NA (Fig. 7). At subdomain IIIA (Sudlow site II), NA interacts hydrophobically with Arg484, Ser488, Leu386, 452, 406 and 429, Phe487, Val432, Phe402, Gly430, Asn390, Thr448. In addition to this, one hydrogen bond formation with Tyr410 (bond length 1.78 Å) and one with Arg410 (bond length 2.10 Å), in subdomain IIIA, has been also observed. The hydrophobic interactions of NA with Ala209, 212 and 349, Val481, Arg347, Lys350, Ser201, Leu480, Phe205, Asn482, Ser479, and two hydrogen bonds with Leu346 and Lys350 at a distance of 2.02 and 1.99 Å were observed in subdomain IIB. The docking and binding energy estimation supported the possibility of π–π interactions at subdomain IIA and IIB with Phe222 and Phe205, respectively. The prime energy predicted for NA at Sudlow site I and II was −25 147.90 and −18 663.00 kJ mol⁻¹, respectively. The binding energy calculations for protein–ligand complex were carried out using MM-GBSA method (generalized born and surface area continuum solvation).⁴⁴ For docked poses A and B (Fig. 7), the value of dG comes out to be −36.63 and −36.92 kJ mol⁻¹ for complex formation at site II and I, respectively. The QM/MM energy for NA–BSA complex was found to be more favorable for Sudlow site I, thus explaining more efficient binding of NA at subdomain IIA. The obtained higher values of entropy might be due to non-consideration of water at the binding sites (Table S1†).

Fig. 6 (A) the most stable docked pose and QM/MM minimised conformation obtained for NA–BSA at Sudlow's site II, and (B) at Sudlow's site I. (C) The optimised structure of protein : ligand complex in 1 : 2 ratio focusing Sudlow's site II, and (D) focusing Sudlow's site I.

Fig. 7 (A) QM/MM optimized structure of IFD resulted best pose at Sudlow site I (subdomain IIA), and (B) corresponding ligand interaction diagram. (C) QM/MM optimised structure of IFD minimized pose at subdomain IIB, and (D) corresponding ligand interaction diagram.

Sequential docking on the most energetically favoured NA–BSA at site I was performed at Sudlow's site II, using the glide docking protocol. The obtained docked pose was further optimised using the QM/MM calculation using B3LYP-6-31g(d) for both molecules of NA and OPLS 2005 for the macromolecule. The results suggested that the binding of first NA molecule at site I is followed by the binding of second NA molecule at second site (Fig. 6) with almost double energy but further increase of NA molecule results in relative decrease in the increase of the optimised energy (Fig. S3 and Table S1†). This implies that NA binds with weaker interactions at subdomain IIB of BSA. The slight inconsistency among calorimetric and docking results may be interpreted in terms of different basis for both techniques.^45,46 In calorimetric experiments (aqueous system), firstly, the protein molecule was free to adopt different conformations as ligand binds and secondly, the solvent (water) molecules may interfere in the binding interactions (hydrogen bonding) among NA and BSA. Whereas, in docking study (static and fixed system), the protein molecule was not allowed to be flexible to adopt different conformations for binding with ligand and also, the role of water molecules at the binding sites of BSA was not considered.^45–47 Thus the overall structural rearrangements in BSA observed as a result of NA binding in aqueous system, plausibly cause the difference in experimental and theoretical results. In addition, some crucial approximations like limited number of ligand positions in the trial and omission of protein dynamics is the basis of docking study.⁴⁵

The location of exact binding site on BSA was further demonstrated by investigating the binding process in the presence of some site probes that are already known to bind at specific site of BSA. In this system, ibuprofen, a non-steroidal anti-inflammatory drug, as site II probe and cefotaxime sodium, an antibacterial drug, as site I probe are used (ESI, Fig. S4†).^1,48 The various thermodynamic parameters obtained from calorimetry for the binding process in the presence of these probes are listed in Table 4. It has been observed that the K₁ value is not affected in the presence of site II probe (ibuprofen) whereas it decreases considerably in the presence of site I probe (cefotaxime sodium). Thus these values of K₁ suggest that NA competes strongly with site I probe for the binding in subdomain IIA (site I) of serum albumin.

Table 4 Binding constant and thermodynamic parameters for the binding of NA with BSA in the presence of site probes at 298.15 K

Site probes	K1, (M^−1)	K2, (M^−1)	ΔH1, (kJ mol^−1)	ΔH2, (kJ mol^−1)	ΔS1, (J mol^−1 K^−1)	ΔS2, (J mol^−1 K^−1)	ΔG1, (kJ mol^−1)	ΔG2, (kJ mol^−1)
Blank	(1.41 ± 0.02) × 10^4	(234 ± 2.70)	(−29.65 ± 0.20)	(−255.30 ± 2.89)	(−20.00 ± 0.55)	(−810 ± 9.59)	(−23.68 ± 0.04)	(−13.52 ± 0.03)
Cefotaxime sodium	(0.88 ± 0.04) × 10^4	(378 ± 16)	(−18.98 ± 0.62)	(−163.70 ± 5.84)	(11.80 ± 1.70)	(−500 ± 19.24)	(−22.51 ± 0.12)	(−14.71 ± 0.11)
Ibuprofen	(1.16 ± 0.04) × 10^4	(328 ± 9.10)	(−24.70 ± 0.47)	(−136.10 ± 3.65)	(−4.98 ± 1.28)	(−408 ± 12.01)	(−23.19 ± 0.09)	(−14.36 ± 0.07)

---

3.3. Alterations in the molecular topology

Insights into the variations in hydrodynamic diameter (d_h) of native BSA upon NA complexation has been obtained by using dynamic light scattering (DLS) measurements. The plot of the hydrodynamic diameter (d_h) of BSA in the absence and presence of NA and the corresponding values are given in Fig. 8 and Table S2,† respectively. The observed d_h value (9.46 ± 0.05) nm of native BSA having lower value of polydispersity (∼19%) corresponds to its monomeric form, in agreement with the reported value.^49,50 The decrease in the d_h value has been observed with increase in NA/BSA molar ratio. This drop in the value of d_h is assigned to the hydrophobic and π–π interactions present between NA and BSA in the interior binding pocket of protein which leads to the decrease in the molecular volume of protein. The possible binding mechanism involves the disruption of the solvent shell around the binding sites of the protein (agrees well with ITC result) which also contributes towards the decrease in the size. Such small decrease in the size of serum albumin due to ligand binding has also been reported in literature.^46,51

Fig. 8 A plot representing the variation in the hydrodynamic diameter (d_h) of BSA as a function of NA/BSA molar ratio.

3.4. Spectroscopic studies

3.4.1. NA induced fluorescence quenching of BSA. As is well known that the various photophysical properties of the fluorophores tryptophan (Trp), tyrosine (Tyr), and phenylalanine (Phe) of protein are highly sensitive to the changes in polarity of the microenvironment. The fluorescence spectroscopy is an apparently susceptible and most commonly used technique to study such changes taking place in their surroundings.⁵² The binding of a ligand either quenches or enhances the intensity of the fluorophores. Various phenomenon's like excited state reactions, molecular rearrangements, energy transfer, ground state complex formation, and collisional quenching occurring during the complex formation may be the basis for quenching of the excited state fluorophores.⁵³ In the present study, BSA was excited at 280 nm to obtain fluorescence emission spectra due to both Trp and Tyr residues. At an excitation wavelength of 280 nm, the energy transfer from the excited state Tyr residues also contributes to the excitation of Trp residues along with the incident radiation.⁵⁴ No emission spectrum for NA was obtained at studied excitation wavelength. Fig. 9 shows the gradual decrease in the fluorescence intensity of BSA upon successive additions of NA. The obtained intensity values were corrected from inner filter effect for further binding analysis. However, during the binding process of NA with BSA, no change in the position of emission maxima (λ_max = 335.8 nm) has been observed, which neglects the possibility of any conformational change in the tertiary structure of protein induced due to NA binding.

Fig. 9 Steady state fluorescence emission spectra of BSA (0.01 mM) at an excitation wavelength of 280 nm with increasing concentrations of NA: (0, 0.001, 0.030, 0.049, 0.069, 0.100, 0.140, 0.220, and 0.338 mM).

In order to assay the type of quenching mechanism i.e. whether static or dynamic, occurring in the complexation process of NA with BSA, the following Stern–Volmer equation was used:⁵⁵

I₀/I = 1 + K_Qτ₀[Q] = 1 + K_SV[Q]

where I₀ and I are the fluorescence intensities of BSA in the absence and presence of NA; [Q] is the quencher concentration (NA in this case); K_SV is the Stern–Volmer quenching constant; K_Q is the bimolecular quenching rate constant; and τ₀ (∼10⁻⁹ s)^17,56 is the average lifetime of BSA in the absence of quencher. A Stern–Volmer plot (Fig. 10A) between I₀/I vs. [NA] was plotted using corrected intensity values. The slope of this plot gives K_SV value = 6.03 × 10⁴ M⁻¹ which was further used to calculate K_Q (= K_SV/τ₀ = 6.03 × 10¹² M⁻¹ s⁻¹). The magnitude of K_Q (∼10¹² M⁻¹ s⁻¹) obtained for the binding of NA with BSA exceeds the limiting diffusion constant of biomolecules (2.0 × 10¹⁰ M⁻¹ s⁻¹).^57,58 This indicates that the static quenching plays a key role in quenching the fluorescence intensity rather than dynamic quenching and clarifies the formation of complex at ground state.

Fig. 10 (A) Stern–Volmer plot of I₀/I vs. concentration of NA ([NA]). (B) Plot of log(I₀ − I/I) vs. log[NA] to calculate the binding constant for NA–BSA system.

3.4.2. Characterisation of binding parameters. The above fluorescence results clearly demonstrate the formation of complex among NA and BSA by static quenching mechanism. For such binding process, the binding constant (K_b) and stoichiometric coefficient (n) were calculated using following equation:⁵⁵

log(I₀ − I)/I = log K_b + n log[Q]

here, the symbols I, I₀, and [Q] are the same as described above. The slope, n = 1.50 and correlation coefficient = 0.9829 was obtained from the plot of log(I₀ − I)/I vs. log[NA] (Fig. 10B). The positive deviation in the value of ‘n’ from unity suggests the binding of more than one molecule of NA with BSA. This result also supports the fitting of ITC data in two sequential binding site model. The value of K_b determined from the intercept (= log K_b) of this plot comes out to be 1.38 × 10⁶ M⁻¹. The order of the overall binding constant K′ (= K₁K₂ = 3.29 × 10⁶ M⁻¹) obtained from ITC agrees well with the order of K_b obtained from fluorescence study.

3.4.3. Absorption spectroscopic studies. To have a look on the structural changes in protein and formation of ligand–protein complex, absorption spectroscopy is an ideal approach to be adopted.⁵⁹ Fig. 11 presents the absorption spectra of BSA with increasing concentrations of NA. An absorption peak at 278 nm reflects the spectral properties of aromatic amino acid residues like Trp, Tyr etc. present in the BSA.⁶⁰ The successive titrations of the solution of NA into BSA were performed by keeping the concentration of BSA as constant (0.01 mM). The binding of NA with BSA resulted in an increase in the absorption (hyperchromic effect) intensity without any shift in the absorption maxima. This indicates the formation of ground state complex among NA and BSA without any change in the microenvironment around aromatic amino acid residues of BSA.

Fig. 11 Electronic absorption spectra (1 → 8) of BSA (0.01 mM) in the absence and presence of increasing concentrations of NA: (0, 0.003, 0.006, 0.009, 0.012, 0.015, 0.018, 0.021 mM). Inset shows the absorption spectra of pure NA.

3.4.4. ¹H-NMR spectroscopy. ¹H-NMR spectroscopy is one of the most recently used techniques to get better perceptive of the molecular interactions existing among proteins and drugs. The changes in the chemical shifts (δ) reflect the unequal contribution of different parts of the drug in stabilizing a complex. In the present work, the effect of the increasing concentrations of NA in BSA has been studied using ¹H-NMR spectroscopy. In order to get better clarity and understanding, the protons of the aromatic ring of NA has been designated as H₁, H₂, H₃, and H₄ as shown in Fig. 1B. The ¹H-NMR spectra and the corresponding δ values of protons (H₁, H₂, H₃, and H₄) are displayed in Fig. S5 (ESI†) and Table 5, respectively. The shift towards the lower field has been observed for all aromatic protons of NA with increase in NA/BSA molar ratio. Such shift in δ values could arise due to magnetic deshielding effect (i.e. the decrease of the electron density around protons). The Δδ value (Table 5) can serve to disclose the different domains around the protons of the bound drug. The results demonstrate the insertion of the lower portion (Fig. 1B) of NA (containing H₁ and H₄ protons) into the hydrophobic cavity (subdomain IIA) of BSA and stabilization of (NA–BSA) complex possibly by π–π stacking interactions. The π–π stacking exists among the lower portion of the electron deficient pyridine ring of NA and aryl residue present in the subdomain IIA of BSA. The stronger binding of NA at subdomain IIA of BSA is already revealed in Section 3.2. Such kind of interactions is known to cause deshielding of aromatic protons in some systems⁶¹ which support our experimental observations. The repulsions between the anionic carboxylate group and π-electron cloud of aryl ring (Phe222, His287) may be responsible for stacking of only the lower portion of the drug instead of upper portion. The stacking is in edge to edge pattern as observed in docking results. A number of biological,^61–63 drug-surfactant,⁶⁴ and supramolecular^65,66 systems showing π–π interactions have been reported in literature.

Table 5 ¹H-NMR chemical shifts (δ) of protons of NA in the free state and after complexation with BSA

Molar ratio (NA/BSA)	δ (ppm)				Δδ (ppm)
	H1	H2	H3	H4	H1	H2	H3	H4
Pure NA	8.538	7.465	8.197	8.866	—	—	—	—
50	8.556	7.529	8.292	8.893	0.018	0.064	0.095	0.027
100	8.625	7.708	8.493	8.946	0.087	0.243	0.296	0.080
150	8.659	7.792	8.588	8.971	0.121	0.327	0.391	0.105

---

The larger downfield shift observed in H₂ and H₃ protons indicate that the upper portion of the drug remain surrounded by polar environment where the carboxylate group binds electrostatically with the oppositely charged groups of BSA. Further, the broadening of the proton signals has been observed in the presence of BSA (Fig. S5†). The merging of the split peaks (due to spin–spin coupling) into one, results in the broadening of the proton signals. Such broadening signifies the decrease in spin–spin relaxation time and increase of correlation time.⁶³ Thus the above results reveal the formation of (NA–BSA) complex via π–π stacking (non-covalent) and electrostatic interactions in addition to the hydrophobic interactions (explored by calorimetry and molecular docking).

3.4.5. Conformational study of BSA: three dimensional fluorescence spectroscopy and circular dichroism. The three dimensional fluorescence spectroscopy is an efficient and powerful technique used to analyze the conformational changes in protein associated due to ligand binding.^67,68 Table 6 represents the characteristic parameters of three dimensional fluorescence spectra of BSA in the absence and presence of NA. Peak 1 and peak 2 are the characteristic emission peaks (Fig. S6†) used to describe the conformational changes in protein. The decrease in the fluorescence intensity has been observed for both peak 1 and peak 2, with a 5 nm red shift in the emission wavelength for peak 2. Here, peak 1 (λ_ex ∼ 280 nm, λ_em ∼ 340 nm) relates to the emission spectral properties of fluorophores (Trp, Tyr) as a result of n → π* transitions in the protein whereas peak 2 (λ_ex ∼ 230 nm, λ_em ∼ 335 nm) exposes the behaviour of the emission spectra of polypeptide backbone structure (i.e. π → π* transitions of C=O of BSA).^69,70 This indicates that the binding of NA induces conformational change in the polypeptide backbone structure of BSA and results in the unfolding of the secondary structure of protein.

Table 6 The three dimensional fluorescence spectral characteristic parameters of BSA and NA–BSA complex in 0.01 M phosphate buffer of pH 7.4 at 298.15 K

Peaks	BSA			NA–BSA
	Peak position (λex/λem) (nm/nm)	Δλ (nm)	Intensity	Peak position (λex/λem) (nm/nm)	Δλ (nm)	Intensity
Peak 1	280/340	60	303.1	280/340	60	296.6
Peak 2	230/335	105	167.9	230/340	110	140.8

---

In order to further elaborate the conformational aspects of protein due to ligand binding, circular dichroism (CD) spectroscopy is a sensitive and efficient technique to be used. The far UV-CD spectra (to monitor the variations in the secondary structure) of BSA in the absence and presence of NA is shown in Fig. 12. Two characteristic negative bands at 209 and 221 nm, attributed to n → π* and π → π* transitions of α-helical structure, have been exhibited by BSA in the far UV-CD region. The decline observed in the negative ellipticity of both bands reflects the decrease in α-helix content in protein upon addition of NA. The α-helix content of BSA in the absence and presence of NA can be evaluated using following equation:⁷¹

α-Helix (%) = {(−MRE₂₀₉ − 4000)/(33000 − 4000)} × 100

where MRE₂₀₉ is the mean residual ellipticity at 209 nm. The MRE value for β-form and random coil is 4000 and for pure α-helix its value is 33 000 at 209 nm. The MRE₂₀₉, used to express variations in the secondary structure of protein can be calculated using following equation:

MRE₂₀₉ = observed CD (mdeg) at 209 nm/(10c_pnl)

where, c_p is the molar concentration of protein (BSA), n is the number of amino acid residues (583 for BSA) and l is the cell path length.

Fig. 12 Far UV-CD spectra of BSA at different NA/BSA molar ratios (0, 4, and 18).

The α-helix content of free BSA was estimated to be 51.37% at 209 nm using above equations. This value decreases to 48.56 and 48.19% on adding NA in the molar ratios of 1 : 4 and 1 : 18, respectively. The slight decrease in the α-helix content manifests the slight change in the secondary structure of protein (i.e. unfolding) due to NA binding. This agrees well with the three dimensional fluorescence spectroscopic results. However, the shape of the spectra does not change suggesting that BSA retains its identity (i.e. remains predominantly α-helix) even after the addition of NA. There is no shift in the peak positions, both at 209 and 221 nm, which indicate the insignificant role of hydrogen bonding in the binding system. This result reinforces the observations obtained from ITC in the presence of sucrose as co-solute. This confirms that the conformational change in BSA induced by NA binding is responsible for the observed inconsistency among Δ_vHH° and ΔH₁ values obtained from ITC.

4. Conclusions

The present work is aimed to analyze the binding interactions among antihyperlipidemic drug, NA and transport protein BSA by various appropriate and efficient techniques. The calorimetric study demonstrates that BSA offers two sequential nature of sites for binding to NA. Among these binding sites, the stronger binding having affinity of the order of 10⁴ has been observed for first site. The significant involvement of electrostatic and hydrophobic interactions to the binding has been successfully interpreted using NaCl, TBAB, sucrose, and surfactants as co-solutes. Site probe and molecular docking studies reveal that subdomain IIA (Sudlow site I) of BSA is the region where the drug binds strongly. The result of UV-vis absorption and fluorescence experiments manifests the formation of ground state NA–BSA complex and quenching of the fluorescence intensity by static quenching mechanism. The order (10⁶) of the overall binding constant obtained from calorimetry is same as obtained from fluorescence measurements. The role of π–π interactions in addition to electrostatic and hydrophobic interactions has been further addressed using molecular docking and ¹H-NMR study. The binding of NA with BSA may proceeds via change in the solvent structure (i.e. hydrogen bonded water network) around the binding sites of BSA as seen by DLS and ITC. The results of conformational study highlight the change in the polypeptide backbone of protein as a result of NA binding. Thus this study provide the energetics of binding, effect on the conformation, and exact location of the binding site on BSA for NA which is useful not only in understanding the effect of the drug on human body but also in designing formulations having better therapeutic effect.

Acknowledgements

Department of Science and Technology, India, for DST-PURSE scheme, and University Grants Commission (UGC), India, for UPE-scheme are thankfully acknowledged. Amandeep Kaur is highly thankful to the UGC, India, for the award of Junior Research Fellowship (JRF). Authors are thankful to Dr Tejwant Singh Kang for help in DLS measurements.

References

1. R. Sharma, S. Choudhary and N. Kishore, Eur. J. Pharm. Sci., 2012, 46, 435–445.
2. Z. Chen, H. Xu, Y. Zhu, J. Liu, K. Wang, P. Wang, S. Shang, X. Yi, Z. Wang, W. Shao and S. Zhang, RSC Adv., 2014, 4, 25410–25419.
3. D. C. Carter and J. X. Ho, Adv. Protein Chem., 1994, 45, 153–203.
4. R. Brodersen, B. Honore, B. Pedersen and I. M. Klotz, Trends Pharmacol. Sci., 1988, 9, 252–257.
5. B. X. Huang, H.-Y. Kim and C. Dass, J. Am. Soc. Mass Spectrom., 2004, 15, 1237–1247.
6. U. Anand and S. Mukherjee, Phys. Chem. Chem. Phys., 2013, 15, 9375–9383.
7. Y. J. Hu, H. L. Yue, X. L. Li, S. S. Zhang, E. Tang and L. P. Zhang, J. Photochem. Photobiol., B, 2012, 112, 16–22.
8. S. Chatterjee and T. K. Mukherjee, Phys. Chem. Chem. Phys., 2014, 16, 8400–8408.
9. S. M. T. Shaikh, J. Seetharamappa, P. B. Kandagal and S. Ashoka, J. Mol. Struct., 2006, 786, 46–52.
10. S. S. Singh and J. Mehta, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2006, 834, 108–116.
11. Y. Yue, J. Liu, J. Fan and X. Yao, J. Pharm. Biomed. Anal., 2011, 56, 336–342.
12. X. Li and S. Wang, New J. Chem., 2015, 39, 386–395.
13. W. Lei, G. Jiang, Q. Zhou, B. Zhang and X. Wang, Phys. Chem. Chem. Phys., 2010, 12, 13255–13260.
14. X. Li, G. Wang, D. Chen and Y. Lu, RSC Adv., 2014, 4, 7301–7312.
15. M. A. Rub, J. M. Khan, A. M. Asiri, R. H. Khan and K. U. Din, J. Lumin., 2014, 155, 39–46.
16. X. Li, G. Wang, D. Chen and Y. Lu, Mol. BioSyst., 2014, 10, 326–337.
17. N. S. Jha and N. Kishore, J. Chem. Thermodyn., 2011, 43, 814–821.
18. U. Anand, L. Kurup and S. Mukherjee, Phys. Chem. Chem. Phys., 2012, 14, 4250–4258.
19. D. Menon, D. Dolimore and K. S. Alexander, Thermochim. Acta, 2002, 392–393, 237–241.
20. L. A. Carlson, J. Intern. Med., 2005, 258, 94–114.
21. K. Cheng, T. J. Wu, K. K. Wu, C. Sturino, K. Metters, K. Gottesdiener, S. D. Wright, Z. Wang, G. O'Neill, E. Lai and M. G. Waters, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 6682–6687.
22. F. Z. A. Zanat, A. M. Qandil and B. M. Tashtoush, Drug Dev. Ind. Pharm., 2011, 37, 1090–1099.
23. S. Tunaru, J. Lattig, J. Kero, G. Krause and S. Offermanns, Mol. Pharmacol., 2005, 68, 1271–1280.
24. S. Choudhary and N. Kishore, Chem. Biol. Drug Des., 2012, 80, 693–705.
25. M. Hossain and G. S. Kumar, Mol. BioSyst., 2009, 5, 1311–1322.
26. A. Basu and G. S. Kumar, J. Hazard. Mater., 2014, 273, 200–206.
27. L. Jen-Jacobson, L. E. Engler and L. A. Jacobson, Structure, 2000, 8, 1015–1023.
28. J. A. Molina-Bolivar, F. Galisteo-Gonzalez, C. C. Ruiz, M. Medina-O'Donnell and A. Parra, J. Mol. Liq., 2015, 208, 304–313.
29. B. Breiten, M. R. Lockett, W. Sherman, S. Fujita, M. Al-Sayah, H. Lange, C. M. Bowers, A. Heroux, G. Krilov and G. M. Whitesides, J. Am. Chem. Soc., 2013, 135, 15579–15584.
30. P. D. Ross and S. Subramanian, Biochemistry, 1981, 20, 3096–3102.
31. F. P. Schwarz, K. D. Puri, R. G. Bhat and A. Surolia, J. Biol. Chem., 1993, 268, 7668–7677.
32. N. Keswani, S. Choudhary and N. Kishore, J. Biochem., 2010, 148, 71–84.
33. T. T. Waldron, G. L. Schrift and K. P. Murphy, J. Mol. Biol., 2005, 346, 895–905.
34. M. T. Record, C. F. Anderson and T. M. Lohman, Q. Rev. Biophys., 1978, 11, 102–178.
35. T. Banerjee and N. Kishore, J. Chem. Thermodyn., 2008, 40, 483–491.
36. K. Takeda, M. Shigeta and K. Aoki, J. Colloid Interface Sci., 1987, 117, 120–126.
37. D. Kelley and D. J. McClements, Food Hydrocolloids, 2003, 17, 73–85.
38. S. K. Singh and N. Kishore, J. Phys. Chem. B, 2006, 110, 9728–9737.
39. D. C. Carter, B. Chang, J. X. Ho, K. Keeling and Z. Krishnasami, Eur. J. Biochem., 1994, 226, 1049–1052.
40. G. Sudlow, D. J. Birkett and D. N. Wade, Mol. Pharmacol., 1975, 11, 824–834.
41. G. Sudlow, D. J. Birkett and D. N. Wade, Mol. Pharmacol., 1976, 12, 1052–1061.
42. Schrödinger Release 2015-4: (a) LigPrep, version 3.6, Schrodinger, LLC, New York, NY, 2015; (b) Prime, version 4.0, Schrödinger, LLC, New York, NY, 2015.
43. P. Chakrabarti and U. Samanta, J. Mol. Biol., 1995, 251, 9–14.
44. S. Genheden and U. Ryde, Expert Opin. Drug Discovery, 2015, 10, 449–461.
45. N. Zaidi, M. R. Ajmal, G. Rabbani, E. Ahmad and R. H. Khan, PLoS One, 2013, 8, e71422.
46. E. Ahmad, G. Rabbani, N. Zaidi, B. Ahmad and R. H. Khan, PLoS One, 2012, 7, e38372.
47. S. Neelam, M. Gokara, B. Sudhamalla, D. G. Amooru and R. Subramanyam, J. Phys. Chem. B, 2010, 114, 3005–3012.
48. Y. J. Hu, Y. Liu and X. H. Xiao, Biomacromolecules, 2009, 10, 517–521.
49. T. Sen, K. K. Halder and A. Patra, J. Phys. Chem. C, 2008, 112, 17945–17951.
50. D. Ray, A. Kundu, A. Pramanik and N. Guchhait, J. Phys. Chem. B, 2015, 119, 2168–2179.
51. J. M. Ruso, P. Taboada, L. M. Varela, D. Attwood and V. Mosquera, Biophys. Chem., 2001, 92, 141–153.
52. C. Jash, P. V. Payghan, N. Ghoshal and G. S. Kumar, J. Phys. Chem. B, 2014, 118, 13077–13091.
53. B. K. Paul and N. Guchhait, Photochem. Photobiol. Sci., 2011, 10, 980–991.
54. J. Steinhardt, J. Krijin and J. G. Leidy, Biochemistry, 1971, 10, 4005–4014.
55. A. Jahanban-Esfahlan, V. Panahi-Azar and S. Sajedi, Biopolymers, 2015, 103, 638–645.
56. E. L. Gelamo and M. Tabak, Spectrochim. Acta, Part A, 2000, 56, 2255–2271.
57. S. K. Jana, S. K. Seth, H. Puschmann, M. Hossain and S. Dalai, RSC Adv., 2014, 4, 57855–57868.
58. G. G. Ariga, P. N. Naik, S. T. Nandibewoor and S. A. Chimatadar, Biopolymers, 2015, 103, 646–657.
59. Z. Chi and R. Liu, Biomacromolecules, 2011, 12, 203–209.
60. X. Zhao, R. Liu, Z. Chi, Y. Teng and P. Qin, J. Phys. Chem. B, 2010, 14, 5625–5631.
61. E. V. Pletneva, A. T. Laederach, D. B. Fulton and N. M. Kostic, J. Am. Chem. Soc., 2001, 123, 6232–6245.
62. G. Wu, X. Jiang, L. Zhou, L. Yang, Y. Wang, G. Xia, Z. Chen and M. Dhan, J. Mol. Struct., 2013, 1045, 47–54.
63. T. Wu, Q. Wu, S. Guan, H. Su and Z. Cai, Biomacromolecules, 2007, 8, 1899–1906.
64. R. K. Mahajan, S. Mahajan, A. Bhadani and S. Singh, Phys. Chem. Chem. Phys., 2012, 14, 887–898.
65. S. Ishihara and S. Takeoka, Tetrahedron Lett., 2006, 47, 181–184.
66. H. J. Schneider, Angew. Chem., Int. Ed., 2009, 48, 3924–3977.
67. Y. Qian, X. Zhou, J. Chen and Y. Zhang, Molecules, 2012, 17, 6821–6831.
68. Q. Wang, Q. Sun, P. Tang, B. Tang, J. He, X. Ma and H. Li, RSC Adv., 2015, 5, 81696–81706.
69. B. K. Paul, D. Ray and N. Guchhait, Phys. Chem. Chem. Phys., 2013, 15, 1275–1287.
70. S. K. Chaturvedi, E. Ahmad, J. M. Khan, P. Alam, M. Ishtikhar and R. H. Khan, Mol. BioSyst., 2015, 11, 307–316.
71. Q. Wang, X. Ma, J. He, Y. Li and H. Li, RSC Adv., 2015, 5, 44696–44704.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra28028a

---