Spectroscopic exploration of drug–protein interaction: a study highlighting the dependence of the magnetic field effect on inter-radical separation distance formed during photoinduced electron transfer

Abstract

In this article, we study the binding interaction of acridone (AD) with human serum albumin (HSA) using conventional spectroscopic techniques and then decipher the latent occurrence of photoinduced electron transfer (PET) using the laser flash photolysis (LFP) technique in conjunction with a weak magnetic field (MF). The experimental observations are further substantiated by docking results. An absorption study reveals the formation of a ground state complex between AD and HSA while a circular dichroism study implies that AD brings about a substantial change in the secondary and tertiary structures of the protein. A fluorescence study mainly helps in evaluation of the binding and thermodynamic parameters. This is one of the rare reports which utilize time-resolved emission spectra and time-resolved area normalized emission spectra to study drug–protein interactions. The detection of the occurrence of PET would be ignored if only the observations of conventional spectroscopic techniques had been considered. LFP detects the occurrence of PET from the tryptophan residue of HSA to AD and MF effect authenticates the triplet origin of the radical ions involved. Furthermore, the actual reaction pathway is elucidated with the help of the MF effect. Although the MF effect is generally observed in a restricted medium, in the present case the complex structure of the protein offers pseudo-confinement to the radical pairs or radical ion pairs resulting in observation of a substantial MF effect. The significance of maintaining the condition of “pseudo-confinement” or the optimum distance between radical ions originating from the acceptor and donor moieties in the observation of the MF effect has been specifically highlighted in this work by integrating the findings of the present study with our previous reports.

---

1. Introduction

Study of the photochemistry of drugs in the presence of biological targets is required for appropriate evaluation of their phototoxicity and photochemical properties. Usually the intrinsic fluorescence of the aromatic amino acid, tryptophan is utilized to study the binding of therapeutically important molecules with proteins. Other spectroscopic tools like circular dichroism (CD), UV-vis absorption, FTIR, NMR etc. are also utilized to gain information regarding the binding modes of drugs with proteins. However, use of the laser flash photolysis (LFP) technique can be reasonably informative to study the photochemistry of drug molecules in the presence of biological targets as it serves as a potential tool to detect the non-fluorescent intermediates which are formed in the due course of the interaction. Photoinduced electron transfer (PET) reaction can be considered as one of the most celebrated photoinduced phenomena which is ubiquitous in nature. If a system containing a drug and a biological target involves PET reaction then radical pairs (RP) or radical ion pairs (RIP) are formed as intermediates, which can be identified and characterized using LFP technique. Further, a weak magnetic field (MF) may help to delve deeper into the spin dynamics of the RPs and RIPs. The electron spin multiplicity of the excited state is generally preserved during the charge transfer, and the memory of this initial spin multiplicity is retained in the resulting RPs/RIPs until the latter undergoes further reactions. These RPs and RIPs are affected by MF as they contain free electrons. MF effect is essentially an integration of diffusion dynamics, spin dynamics and recombination or free radical formation. Spin multiplicity of RIP can evolve coherently between singlet (S) and triplet (T) states, whose energy differ by 2J, where J is the electron exchange integral that preserves electron indistinguishability. The three T states (T_±, T₀) remain degenerate in absence of an external MF. It is relevant to mention here that J depends on the separation distance between the two RIP centres and falls of exponentially as the components of RIP diffuse apart. The RIPs of the geminate cage diffuse in the medium and attain an optimum distance of separation where J is negligible which leads to ample mixing of S and T states induced by hyperfine interaction (HFI) present as internal MF. Application of an external MF of the order of HFI or higher removes the degeneracy of the triplet state via Zeeman splitting and intersystem crossing (ISC) is reduced as only S → T₀ channel remains operative, eventually leading to an increase in population of initial spin state. In general, geminate RPs/RIPs with singlet correlation lead to the formation of recombination product or exciplex, whereas those with triplet configuration prefer to form escape product (free radicals/radical ions). If the initial spin state of the precursors of PET is singlet then there will be an enhancement in the luminescence of the singlet precursors in presence of the external MF, while if the initial spin state is triplet then there will be increase in absorbance of the radical ions in presence of the external MF. Thus, MF effect serves as a tool to identify the initial spin state of precursors of PET.^1–5 Therefore, LFP in conjunction with an external MF is potent enough to unveil the occurrence of PET in drug–protein interaction and portrays a detailed picture of the spin dynamics involved.

Owing to the planar structure of acridine moiety, its derivative can act as good intercalators and thus, the study of interaction of acridine derivatives with nucleic acids has attracted the attention of researchers since a long time.^6–10 Relatively fewer reports are available in literature which discuss about the interaction of acridine derivatives with proteins compared to DNA and RNA. Acridine drugs are known to be prospective candidates of photosensitizer in photodynamic therapy^11–13 and they also have anti-fungal as well as anti-bacterial properties. Thus, study of interactions of such acridine derivatives with exogenous and endogenous drug-delivery vehicles is of pharmacological importance.

Human serum albumin (HSA), a typical model protein which acts as a potent drug-delivery vehicle,^14,15 is the most abundant carrier protein present in blood. It is composed of 585 amino acid residues and 17 disulfide bridges, and the study of its three dimensional crystalline structure suggests that it has three homologous α-helical domains, I (residues 1–195), II (residues 196–383) and III (residues 384–585), each of which has been further classified into subdomains A and B.¹⁶ Subdomains IIA and IIIA (also known as Sudlow's site I and II respectively) are capable of binding aromatic and heterocyclic ligands. He and Carter reported that the two halves of the model protein form a 10 Å wide and 12 Å deep crevice that house the single tryptophan (Trp) residue at binding site IIA.¹⁷ The presence of this sole tryptophan residue, in HSA makes the study of this protein convenient by use of fluorescence technique.

In the present article, we make an endeavor to investigate the binding interactions between acridone (AD), which is an acridine derivative (as shown in Fig. 1) with HSA using conventional techniques like UV-vis absorption, fluorescence and circular dichroism spectroscopies and substantiate our experimental findings with the results of docking studies. This is one of the rare reports which utilize time-resolved emission spectra (TRES) and time-resolved area normalized emission spectra (TRANES) to study drug–protein interaction. Use of time-resolved anisotropy technique also helps to give a vivid insight into the photo-behavior of the drug in presence of the model protein. Further, this is one of the fewer reports in which an attempt has been made to decipher the possibility of PET in drug–protein interaction using LFP coupled with an external MF, which may otherwise go unnoticed if only conventional spectroscopic findings are considered. The potency of MF effect is not only restricted to detection of occurrence of PET, but also may be used to and assess the proximity of RPs/RIPs during reaction. Previously, this spatial dependence nature has been utilized by our group to obtain appreciable MF effect while studying the interaction of ligands with biomacromolecules like proteins and DNA.^18,19 Actually, a separation distance of about 10–20 Å is required for observing MF effect, which is usually retained by confining the RPs/RIPs within heterogeneous organized assemblies.^20–23 The complex structure of the biomacromolecules provides a restricted environment that maintains the required optimum distance of separation between the RIPs which is necessary to preserve their geminate character as well as their spin correlation, thus exhibiting MF effect. In case of DNA it is found that the partial intercalative mode of binding of the ligand helps to maintain the ideal distance required for observation of MF effect. In case of proteins it is observed that ligands capable of maintaining a distance of 12–13 Å from electron donor amino acid residue of human serum albumin is successful in showing MF effect. Thus, integration of our present results with these previous findings may help to design systems in which proper distance between the RIPs generated through PET can be maintained such that appreciable MF effect is observed in biomacromolecules. Hence, apart from studying the binding interaction of AD with HSA using the usual spectroscopic techniques and detection of PET in drug–protein interaction with the help of LFP coupled with an external MF, this paper highlights the fact that MF effect can be used as a ‘spectroscopic ruler’ to estimate the inter-radical separation distance during PET.

Fig. 1 Chemical structure of AD.

2. Experimental methods

2.1. Materials

AD was obtained from Fluka. Fatty acid- and globulin-free HSA was also procured from Fluka. All the solutions except the stock solution of AD were prepared in a 10 mM phosphate buffer of pH 7.4. Stock solution of AD was prepared in ethanol. UV spectroscopy grade ethanol was purchased from Spectrochem.

2.2. Apparatus

2.2.1. Absorption spectroscopy. UV-vis absorption spectra were recorded in Jasco V-650 absorption spectrophotometer at 298 K using a pair of 1 × 1 cm quartz cuvettes. The concentration of the protein was determined spectrophotometrically using the value of molar extinction coefficient of HSA at 280 nm as 36 600 M⁻¹ cm⁻¹.²⁴

2.2.2. CD spectroscopy. CD measurements were made on a Jasco-720 automatic recording spectrophotometer using a 1 cm path length cuvette. An accumulation of three scans with a scan speed of 20 nm per minute and response time of 2 s was performed to collect data.

2.2.3. Fluorescence spectroscopy. Steady-state fluorescence spectra were recorded in Spex Fluoromax-3 spectrofluorimeter using 1 × 1 cm quartz cuvette at 301, 310, 320 and 328 K with excitation wavelength of 280 nm. The fluorescence of HSA was corrected for inner filter effect owing to absorbance by AD at the excitation and emission wavelengths, using the following equation,¹⁸where, F_corr and F_obs are the corrected and observed fluorescence intensities, respectively, and OD_ex and OD_em are the absorbances at excitation and emission wavelengths, respectively. Fluorescence lifetime in singlet state was measured using a diode laser-based Jobin Yvon Horiba picosecond-resolved time-correlated-single-photon-counting (TCSPC) spectrometer with excitation wavelength at 280 nm using pulsed diode light source Nano LED with pulse duration of 1 ns and repetition rate of 1 MHz. The data were fitted to multiexponential functions after deconvolution of the IRF by an iterative reconvolution technique using IBH DAS 6.2 data analysis software. Analysis of the fluorescence decay data I(t) was done using the following equation:where B_i and τ_i are the pre-exponential factor and the fluorescence lifetime, respectively. The values of reduced χ² and residuals serve as the parameters for goodness of the fit. Using the steady-state and time-resolved fluorescence data the TRES and TRANES were constructed. TRES was constructed by measuring the fluorescence decays across the emission spectrum (290–410 nm) at particular intervals. The fitted fluorescence decays were scaled with the steady-state fluorescence intensities.^25,26 The fractional contribution of each component of the fluorescence spectrum at the wavelength of measurement was calculated according to the following equation:where, I_i(λ) is the fractional contribution and α_i and τ_i are the relative amplitude and lifetime of the ith component, respectively. Now the reconstruction of the time-resolved spectra at different time t was performed using the best fitting parameters as suggested by Maroncelli and Fleming.²⁷ Time resolved anisotropy decay measurements were also carried out in Jobin Yvon Horiba picosecond-resolved TCSPC spectrometer and the sample was excited at 377 nm using pulsed diode light source Nano LED with pulse duration of 100 ps and repetition rate of 1 MHz. Anisotropy r(t) is defined as,²⁴where, I_VV and I_VH are the intensity decays of emission obtained with the excitation polarizer oriented vertically and emission polarizer oriented vertically and horizontally respectively. G is the correction term for the relative throughput of each polarization through the emission optics and is given by:²⁸

The entire data analysis was done using IBH DAS 6.2 data analysis software to construct r(t) and from the fitted curve correlation time τ_r was finally recovered.

2.2.4. Transient absorption measurement. A nanosecond flash photolysis setup (Applied Photophysics) containing a Nd:YAG (Labseries, Model Lab150, Spectra Physics) laser was used for the measurement of transient absorption spectra. The sample was excited at 355 nm (FWHM = 8 ns) laser light. Transients were monitored through absorption of light from a pulsed xenon lamp (150 W). The photomultiplier (R298) output was fed into an Agilent Infiniium oscilloscope (DSO8064A, 600 MHz, 4 Gs/s), and the data were transformed to a computer using the IYONIX software. The software Origin 8.0 was used for curve fitting. The solid curves were obtained by connecting the points using B-spline option. The samples were deaerated by passing pure argon gas for 20 min prior to each experiment. No degradation of the samples was observed during the experiments. The MF (0.08 T) effect on the transient spectra was studied by passing direct current through a pair of electromagnetic coils placed inside the sample chamber.

2.3. Methods

2.3.1. Calculation of mean residue ellipticity (MRE) and α-helical content of proteins by CD spectroscopy. The results of CD were expressed as MRE (in deg cm² dmol⁻¹) according to the following equation:²⁹where, C_p is the molar concentration of the protein, n is the number of amino acid residues, and l is the path length. For HSA, n = 585 (ref. 29) and cuvettes with l = 1 cm was used. The α-helical content of free and bound protein was eventually evaluated from the MRE value at 222 nm using the following formula:²⁹where, MRE₂₂₂ is the observed MRE value at 222 nm.

2.3.2. Förster resonance energy transfer (FRET) calculation. The distance between the fluorophore in the protein and the drug was evaluated using Forster's theory of energy transfer, according to which the efficiency of energy transfer is given by the following equation:²⁸where, r is the distance between the donor and acceptor and R₀ is the critical energy transfer distance, at which 50% of the excitation energy is transferred to the acceptor and is given by the following equation:²⁸

R₀⁶ = 8.79 × 10⁻²⁵κ²η⁻⁴Jϕ (9)

In eqn (9), κ² is the spatial orientation factor describing the relative orientation in space of the transition dipoles of the donor and acceptor, n is the refractive index of the medium, ϕ is the fluorescence quantum yield of the donor in the absence of the acceptor, and J is the overlap integral of the fluorescence emission spectrum of the donor and the absorption spectrum of the acceptor, which is given by,²⁸

where, F(λ) is the fluorescence intensity of the fluorescent donor at wavelength λ, and ε(λ) is the molar absorption coefficient at wavelength λ. Under the experimental conditions, k² = 2/3, n = 1.336,²⁸ and ϕ = 0.118 (ref. 30) for HSA. Values of E and J were obtained experimentally.

2.3.3. Calculation of binding constant (K) and number of binding sites (n). Binding constant (K) and number of binding sites (n) for drug protein interaction can be evaluated using the following equation:³¹where, F₀ and F are the fluorescence intensities in the absence and presence of quencher, and [Q] is the concentration of quencher. log(F₀ – F)/F was plotted against log[Q], and the values of K and n were obtained from the intercept and slope, respectively.

2.3.4. Estimation of thermodynamic parameters. The thermodynamic parameters were determined using van't Hoff and Gibbs–Helmholtz equations as follows:and

ΔG⁰ = ΔH⁰ − TΔS⁰ (13)

ln K was plotted against 1/T and ΔH⁰ as well as ΔS⁰ were obtained from the slope and the intercept of the plot, respectively. The values of ΔG⁰ at different temperatures were determined by making use of eqn (13).

2.4. Molecular docking studies

The crystal structure of HSA (PDB entry 1E78) was downloaded from Protein Data Bank. Acridone structure was obtained from PubChem (http://pubchem.ncbi.nlm.nih.gov/) as 3D sdf text and subsequently it was converted to pdb text with OpenBabel to use as an input to AutoDock. AutoDockVina (http://vina.scripps.edu/) and MGLTools (http://mgltools.scripps.edu/) of The Scripps Research Institute were used to perform the docking calculations.

The ligand molecule was energy minimized and the Gasteiger partial charges were added to the ligand atoms, nonpolar H atoms are merged and rotatable bonds were also defined. Essential H atoms, Kollman united atom type charges and solvation parameters were added with the aid of AutoDock Tools. Docking simulations have been performed using the Lamarckian genetic algorithm^32–34 and the Slis and Wets local search method. Initial position, orientation and torsions of the ligand molecules were set randomly. The output from AutoDock was rendered with PyMOL.³⁵ Here PyMOL was also used to calculate the distances between nearest atoms which interact with each other. The accessible surface areas (ASA) of uncomplexed (only HSA) and complexed (HSA with AD) proteins were calculated from Helix Systems server. The protein–ligand structure corresponding to the minimum score as obtained from the Auto-Dock Vina analysis was chosen.³⁶

3. Results and discussions

3.1. UV-vis absorption spectroscopy

UV-vis absorption study is used to explore the possibility of ground-state interaction between AD and HSA. AD is insoluble in water (or phosphate buffer) and thus, the solution of AD is prepared in ethanol medium. Inset of Fig. 2a depicts the absorption spectrum of AD in ethanol which has considerable absorption in the wavelength region 310–410 nm, whereas the model protein HSA does not absorb light in this region (as shown in Fig. 2a). Differential absorption method is used to investigate whether any ground-state interaction is present between the acridine derivative and the protein. Calculated amount of drug along with buffer solution serves as the reference solution in this case. Absorption spectrum is monitored by adding incremental amount of protein to the same concentration of AD which is used in the reference solution. Such a differential method of absorption study removes the contribution of absorbance of AD in the wavelength region characteristic to the acridine derivative (310–410 nm). Fig. 2b shows that with increase in concentration of HSA, absorbance of peaks at 382 and 402 nm as well as a shoulder at 365 nm steadily increases. Emergence of these two new peaks accompanied by a shoulder in the region where HSA has no absorption and contribution of AD has already been eliminated implies that a ground state complex is formed between HSA and AD and the complex so formed has characteristic peaks at 382 and 402 nm accompanied by a shoulder at 365 nm. It is pertinent to mention here that earlier reports suggest that hydrogen bond formation is associated with red shift of absorption spectrum.^37–39 In the present case, the peaks corresponding to the ground state complex (382 and 402 nm) is red shifted compared to that of free AD (379 and 397 nm) implying the possible involvement of hydrogen bonding in drug–protein interaction.

Fig. 2 (a) Absorption spectrum of 1.25 × 10⁻⁵ (M) HSA in phosphate buffer solution showing maximum absorption at 280 nm and no absorbance in the wavelength region 310–410 nm. Inset depicts the absorption spectrum of 4.5 × 10⁻⁵ (M) AD in ethanol medium showing considerable absorption in the wavelength region 310–410 nm and having absorption maxima at 379 and 397 nm. (b) Absorption spectra showing ground-state complex formation between AD and HSA in phosphate buffer solution; [AD] = 7.5 × 10⁻⁵ (M) and [HSA] are: (a) 3.70 × 10⁻⁵ (M) and (b) 5.55 × 10⁻⁵ (M).

3.2. Circular dichroism spectroscopy

Perturbation in the secondary and tertiary structures of a protein can be detected using CD spectroscopy. Two negative bands in the far-UV region (190–250 nm) at 208 and 222 nm are characteristic to native HSA, typifying the α-helical structure of proteins. Fig. 3a shows the change in CD spectrum of HSA in the far-UV region, which accounts for the alteration of secondary structure of the protein on binding with AD. It is customary to express helical content of a protein in terms of mean residue ellipticity at 222 nm, i.e. MRE₂₂₂ as discussed in Section 2.3.1. The α-helix content of the model protein in presence and absence of AD is calculated using eqn (6) and (7) and is summarized in Table 1. It is observed that α-helix content of HSA increases from ∼64% to 87% in presence of AD (as shown in Table 1) implying the stabilization of protein on binding with the acridine derivative³¹ and it is accompanied by formation of an isodichroic point at 206 nm (as shown in Fig. 3a).⁴⁰ Alteration in the tertiary structure of HSA is evident from the substantial change in near-UV CD spectrum of the protein in presence of the acridine derivative as shown in Fig. 3b. Increase in ellipticity in the wavelength range of 250–270 nm is observed with gradual addition of AD to the solution of serum albumin. Near-UV CD is characteristic to the aromatic chromophores as well as the disulphide bonds present in the protein and reflects its tertiary structural organization. The three aromatic amino acid residues in a protein have characteristic wavelength profiles. Tryptophan has a peak close to 290 nm, tyrosine has a peak between 275 and 282 nm and phenylalanine has sharp fine structures between 255 and 270 nm.⁴¹ Furthermore, near UV absorption of disulphide bonds occurs near 260 nm.⁴¹ Thus, the results of near-UV CD (as depicted in Fig. 3b) imply that addition of AD to HSA results in conformational changes around phenylalanine residues and/or disulphide bridges. It is pertinent to mention here that HSA has 17 disulphide bridges⁴² and thirty three phenylalanine⁴³ residues.

Fig. 3 (a) Far-UV CD spectrum showing the variation of ellipticity at 208 and 222 nm on addition of AD to HSA in phosphate buffer solution of pH 7.4; [HSA] = 1 × 10⁻⁶ (M) and [AD] varies from 0 to 7.5 × 10⁻⁶ (M). (b) Near-UV CD spectrum of HSA in phosphate buffer solution of pH 7.4 in presence of AD. [HSA] = 7 × 10⁻⁶ (M) and [AD] varies from 0 to 1 × 10⁻⁵ (M).

Table 1 Variation of α-helix% of HSA in phosphate buffer with increase in concentration of AD

[HSA] (M)	[AD] (M)	MRE222 nm	α-helix%
1 × 10^−6	0	−17079.5	64.09
1 × 10^−6	5.0 × 10^−6	−20947.4	76.85
1 × 10^−6	7.5 × 10^−6	−24279.2	87.85

---

3.3. Fluorescence spectroscopy

AD has fluorescence maxima at about 420 and 440 nm whereas the fluorescence spectrum of HSA displays a maximum at 348 nm. The intrinsic fluorescence of the protein in quenched on addition of AD accompanied by enhancement of the fluorescence of the drug and an isoemissive point is obtained at 396 nm. Fig. 4 shows the fluorescence quenching of HSA in phosphate buffer with increase in concentration of AD at 301 K with an inset depicting the emergence of the isoemissive point. It is evident from Fig. 4 that λ_max of fluorescence spectrum of HSA undergoes gradual blue shift with increase in concentration of AD implying that the tryptophan residue (Trp 214) in the protein experiences a hydrophobic environment in presence of the acridine derivative. Steady-state fluorescence data is analyzed with the help of Stern–Volmer (SV) equation.²⁸ A non-linear SV plot for steady-state fluorescence is obtained as depicted in Fig. 5, which possibly indicates the occurrence of combined static and dynamic quenching.²⁸ The existence of static quenching mechanism is evident from absorption spectroscopic study. To verify the presence of dynamic quenching, time resolved fluorescence study is carried out and it is found that the fluorescence lifetime of HSA at 340 nm is reduced in the presence of AD. The corresponding SV plot (as depicted in the inset of Fig. 5) is found to be linear as it accounts exclusively for dynamic quenching. HSA exhibits bi-exponential decay and its average lifetime (as shown in Table 2) has been used to obtain a qualitative picture.

Fig. 4 Fluorescence spectra showing quenching of intrinsic fluorescence of HSA with increase in concentration of AD; λ_ex = 280 nm at 301 K. [HSA] = 4.0 × 10⁻⁶ (M) and [AD] = (a) 0, (b) 1.5 × 10⁻⁵ (M), (c) 3.0 × 10⁻⁵ (M), (d) 4.5 × 10⁻⁵ (M), (e) 7.5 × 10⁻⁵ (M), (f) 1.2 × 10⁻⁴ (M) and (g) 1.7 × 10⁻⁴ (M). Inset depicts the same set of spectra in wavelength range of 290 to 550 nm showing the isoemissive point; λ_ex = 280 nm.

Fig. 5 SV plot for quenching of steady-state fluorescence of HSA by AD. [HSA] = 4.0 × 10⁻⁶ (M) and [AD] ranges from 0 to 1.7 × 10⁻⁴ (M). Inset shows time-resolved SV plot for quenching of fluorescence of HSA by AD; [HSA] = 5 × 10⁻⁶ (M) and [AD] ranges from 0 to 2.8 × 10⁻⁵ (M).

Table 2 Variation of fluorescence lifetime of HSA (〈τ〉) with increase in concentration of AD. [HSA] = 5 × 10⁻⁶ (M). λ_ex = 280 nm and λ_em = 340 nm

[AD] (M)	A1	τ1 (ns)	A2	τ2 (ns)	〈τ〉^a (ns)	χ^2
a .
0	23.76	2.99	76.24	7.27	6.78	1.14
6.9 × 10^−6	37.9	3.07	62.1	7.06	6.22	1.11
1.38 × 10^−5	45.26	3.13	54.74	6.96	5.92	1.19
2.07 × 10^−5	42.15	3.20	57.85	6.45	5.58	0.98
2.76 × 10^−5	55.67	3.22	44.33	6.45	5.20	0.98

---

The values of Stern–Volmer constant (K_sv) and quenching rate constant (k_q) obtained from steady-state fluorescence data by considering the low concentration range of AD where the SV plot is linear⁴⁴ are 1.98 × 10⁵ M⁻¹ and 2.92 × 10¹² M⁻¹ s⁻¹ respectively. The values of K_sv and k_q obtained from time-resolved fluorescence data are 1.05 × 10⁵ M⁻¹ and 1.56 × 10¹² M⁻¹ s⁻¹ respectively. Such high values of k_q indicates that energy transfer may probably take place from HSA to AD as the upper limit of the value of k_q for a diffusion controlled process is of the order of 10¹⁰.⁴⁵ Another indication of energy transfer is the quenching of fluorescence of HSA with simultaneous increase in emission of AD accompanied by formation of an isoemissive point as shown in the inset of Fig. 4. However, the most convincing proof of the occurrence of energy transfer in the singlet in state is the decrease in fluorescence lifetime of HSA in presence of AD as depicted in Table 2. Previously, Barik et al.⁴⁶ and Patel et al.⁴⁴ have also confirmed the occurrence of FRET by observing reduction in lifetime of the donor on addition of the acceptor moiety. The required condition for FRET, i.e., the overlap of the absorption spectrum of the acceptor and the fluorescence spectrum of the donor, is satisfied for AD–HSA system as evident from Fig. 6. Using eqn (8) and (9) the Förster distance and the mean distance between Trp-214 and AD have been found to be 21.9 Å and 19.6 Å respectively. Similar distances have been reported earlier for drugs that bind to Sudlow's Site I as chlorine derivatives⁴⁴ or ochratoxin.⁴⁷

Fig. 6 Spectral overlap between (a) emission spectrum of the donor (HSA) and (b) absorption spectrum of the acceptor (AD); [AD] = [HSA] = 5 × 10⁻⁵ (M).

Increase in fluorescence anisotropy of a fluorophore is a result of increase in rigidity of its neighbouring environment.²⁸ In time-resolved anisotropy measurement, increase in rotational correlation time of a fluorophore reflects the rigidity of its neighbouring environment. In order to find out the effect of confinement of AD in the protein environment, the time-resolved anisotropy of AD is monitored in absence and presence of HSA. According to Stokes–Einstein relationship:⁴⁸

where,

Here, τ_r is the rotational correlation time, V is the hydrodynamic molecular volume, T is the absolute temperature, D_r is the rotational diffusion co-efficient and η is the viscosity of the medium. If AD in the excited state forms a complex with HSA then the hydrodynamic molecular volume is expected to increase, which in turn may increase its rotational correlation time. Considering this fact, time-resolved anisotropy measurements are carried out with free AD solution and a solution containing AD as well as HSA. For a solution containing only AD, the anisotropy decay fits to a single exponential function and the value of τ_r is evaluated to be 0.235 ns. However, for a solution containing a mixture of AD and HSA, the anisotropy appears to be bi-exponential with two correlation times – a shorter component of 0.227 ns (34.02%) and a longer component of 8.48 ns (65.98%). The existence of two components indicates the presence of two species, free AD and AD bound to HSA. The shorter component of 0.227 ns represents free AD in the system as its value is almost same as that of τ_r of a solution of AD (0.235 ns), whereas the motional restriction imposed on AD in the excited state upon complex formation with HSA is responsible for the emergence of the longer component of the rotational correlation time. High value of correlation time which has been attributed to the complex formed between AD and HSA imply buried nature of the binding site of the ligand in the protein.⁴⁹ Time-resolved anisotropy decay curves for AD are depicted in Fig. 7.

Fig. 7 Time-resolved anisotropy decay curves for (a) a solution of 7 × 10⁻⁶ (M) AD and (b) a solution of 7 × 10⁻⁶ (M) AD + 2 × 10⁻⁵ (M) HSA; λ_ex = 377 nm and λ_em = 420 nm.

TRES and TRANES are constructed from steady-state and time-resolved fluorescence spectra. TRANES is a more convenient method of analysis of wavelength dependent fluorescence decay compared to TRES as the former is a model free method as suggested by Koti and Periasamy.⁵⁰ TRANES analysis helps to determine the number of emissive species present in the system. Fig. 8a and b show the TRES and TRANES profiles respectively of a solution containing HSA and AD using λ_ex = 280 nm. As evident from Fig. 8b, the TRANES profile contains only one isoemissive point at 330 nm implying that arises from the presence of two species, viz., free HSA and HSA bound to AD.

Fig. 8 (a) TRES of AD (7 × 10⁻⁶ M) + HSA (2.50 × 10⁻⁵ M) in phosphate buffer solution between time 1 and 16 ns. (b) TRANES of AD (7 × 10⁻⁶ M) + HSA (2.50 × 10⁻⁵ M) in phosphate buffer solution between time 1 and 16 ns.

The values of binding constant and binding site number can be assessed from fluorescence data by making use of eqn (11) as shown in Table 3. Fluorescence study at various temperatures helps in estimating the thermodynamic parameters such as ΔG⁰, ΔH⁰ and ΔS⁰ associated with drug–protein interaction and according to Ross and Subhramanian⁵¹ the signs of the thermodynamic parameters help in predicting the nature of forces involved in drug–protein interaction. Eqn (12) and (13) are used to evaluate the thermodynamic parameters and Table 3 summarizes the values thus obtained. The value of K decreases with rise in temperature indicating the formation of a complex between HSA and AD which loses its stability at higher temperature. It is observed that ΔG⁰ is negative at all temperatures implying that the reaction is spontaneous in nature. Table 3 shows that AD–HSA system involves negative values of ΔH⁰ and ΔS⁰ signifying that hydrogen bonding and van der Waals interactions are possibly involved in the binding interaction.⁵¹ The probable involvement of hydrogen bonding has already been predicted by UV-vis absorption study as discussed earlier. It is to be noted that the magnitude of ΔH⁰ outweighs that of ΔS⁰ which means that the interactions are enthalpy driven and not entropy driven.

Table 3 Determination of binding constant K, number of binding site n and thermodynamic parameters for AD–HSA system at different temperatures

Temperature (K)	n	K (M^−1)	ΔG^0 (kJ mol^−1)	ΔH^0 (kJ mol^−1)	ΔS^0 (J mol^−1 K^−1)
301	1.32	3.98 × 10^5	−30.62	−87.98	−190.56
310	1.03	2.70 × 10^4	−28.91
320	1.01	2.51 × 10^4	−27.00
328	0.73	1.65 × 10^4	−25.48

---

3.4. LFP in conjunction with MF

LFP is used to study the interaction of AD with HSA in the excited state. The effect of addition of HSA in phosphate buffer to a solution of AD is depicted in Fig. 9A. The transient spectra of 2 × 10⁻⁴ (M) AD and 2 × 10⁻⁴ (M) AD + 9 × 10⁻⁵ (M) HSA in phosphate buffer of pH 7.4 at 0.6 μs after the laser flash at 355 nm is shown in Fig. 9A. Study of literature suggests that transient absorption spectral peaks of TrpH˙⁺ are formed at 350 nm (ref. 52) as well as 560 nm (ref. 53) and that of Trp˙ is at 520 nm.⁵⁴ It has already been reported by our group that triplet–triplet transient absorption spectrum of AD has a broad peak in the wavelength region spanning from 400 to about 500 nm with a maximum at 430 nm accompanied by a small peak at 350 nm and the characteristic peak of AD˙⁻ appears around 510–520 nm.⁵⁵ Decrease in absorbance around 420–430 nm on addition of HSA in Fig. 9A signifies the quenching of ³AD* in the presence of the protein, whereas emergence of a new peak at 550 nm on addition of the model protein indicates the formation TrpH˙⁺. However, formation of TrpH˙⁺ should have been reflected as increase in absorbance at 350 nm, which is not detected probably because of substantial contribution of ³AD* at the same wavelength that leads to quenching of the small peak at 350 nm in the presence of HSA, a behavior similar to that at 420–430 nm. The most significant change in the spectrum of ³AD* on addition of HSA is the enormous increase in absorbance in the wavelength region 450–520 nm. While studying the interaction of thioxanthone with tryptophan, a similar observation of formation of a broad peak in same region has been reported by Das and Nath⁵⁶ which they assigned to Trp˙ although the characteristic peak of the amino acid radical is reported to appear at 510 nm. Moreover, presence of AD˙⁻ is also responsible for enhancement of absorbance at 510 nm.⁵⁵ Thus, in the present case cumulative contribution from Trp˙ and AD˙⁻ is responsible for the huge increase in absorbance around 450–520 nm as depicted in Fig. 9A. The decay profiles of 2 × 10⁻⁴ (M) AD in presence and absence of HSA at 510 nm are portrayed in Fig. 9B, which shows an increase in yield of Trp˙ and AD˙⁻ on addition of the serum albumin to a solution of AD. Fig. 9C and D depict the transient absorption spectra of 2 × 10⁻⁴ (M) AD at 0.2, 0.6, 1.0 and 2.0 μs time-delay after the triggering of the laser pulse in absence and presence of HSA respectively. Fig. 9C shows that the characteristic peaks of ³AD* at 350 nm and 430 are stable at longer time-delay. The time-resolved spectra in Fig. 9D show a persistent peak in the wavelength region of 450–520 nm which reflects the simultaneous presence of Trp˙ and AD˙⁻, which are undisturbed even at longer time delay. Substantial quenching of peak of ³AD* at 350 nm and appearance of signature peak of TrpH˙⁺ at 550 nm on addition of HSA to a solution of AD can be vividly observed on comparing Fig. 9C and D.

Fig. 9 (A) Transient spectra of (a) 2 × 10⁻⁴ (M) AD (■) and (b) 2 × 10⁻⁴ (M) AD + 9 × 10⁻⁵ (M) HSA (•) in phosphate buffer of pH 7.4 at 0.6 μs after the laser flash at 355 nm. (B) Decay profiles of (a) 2 × 10⁻⁴ (M) AD and (b) 2 × 10⁻⁴ (M) AD + 9 × 10⁻⁵ (M) HSA in phosphate buffer of pH 7.4 at 510 nm after laser flash at 355 nm. (C) Time-resolved transient absorption spectra of 2 × 10⁻⁴ (M) AD at 0.2 μs (■), 0.6 μs (•), 1.0 μs (▲) and 2.0 μs (▼) after the laser flash at 355 nm. (D) Time-resolved transient absorption spectra of 2 × 10⁻⁴ (M) AD in presence of 9 × 10⁻⁵ (M) HSA at 0.2 μs (■), 0.6 μs (•), 1.0 μs (▲) and 2.0 μs (▼) after the laser flash at 355 nm.

Emergence of AD˙⁻ and TrpH˙⁺ can be rationalized in the light of PET taking place from the tryptophan residue of the protein to ³AD* leading to the formation of these two radical ions accompanied by quenching at 420–430 nm characteristic to ³AD*. Now the pertinent question is what leads to the formation of Trp˙? Deprotonation of radical cation of tryptophan (TrpH˙⁺) to form the corresponding radical of the amino acid (Trp˙) is a well-known reaction pathway followed by TrpH˙⁺.^52,57,58

TrpH˙⁺ → Trp˙ + H⁺

To make a detailed study on the spin dynamics of PET, the effect of application of a weak MF on the reactions is studied. Fig. 10A shows the consequence of application of a MF of 0.08 T on the transient absorption spectrum of 2 × 10⁻⁴ (M) AD + 3 × 10⁻⁵ (M) HSA in phosphate buffer of pH 7.4 at 0.1 μs after the laser flash at 355 nm. It is observed that absorbance around 510 and 560 nm are enhanced in the presence of MF which probably indicates the increased yield of AD˙⁻ and TrpH˙⁺ on application of MF. Thus, the MF effect is exhibited by the spin correlated geminate pair (AD˙⁻ TrpH˙⁺), which arises out of coupling of its spin and diffusion dynamics indicating that precursors of PET are triplet in origin. It is noteworthy that the presence of TrpH˙⁺ is thus authenticated by MF effect at 560 nm, which was previously undecided as its signature peak at 350 nm is quenched due to the dominant effect of ³AD* at the same wavelength, as discussed earlier. Appreciable MF effect is generally observed in confined media^20–23 which provides the optimum distance of separation of the radical ions of about 10–20 Å such that their spin-correlation is conserved. In the present case pseudo-confinement of the radical ions are provided by the complex structure of the protein and from fluorescence studies (19.6 Å) as well as docking (13.5 Å) the distance between Trp and AD is found to be apt for observation for MF effect. Table 4 shows that the lifetime of AD˙⁻ and TrpH˙⁺ is fairly enhanced in the presence of the external MF indicating that they are triplet-born. Fig. 10B shows the time-resolved transient absorption spectra of 2 × 10⁻⁴ (M) AD + 3 × 10⁻⁵ (M) HSA in presence of MF at 0.2, 0.6, 1.0 and 2.0 μs time-delay after the triggering of the laser pulse. However, it is pertinent to mention here that in the present study MF effect is observed to be small. Mohtat et al. have also observed small MF effect while studying the interaction of triplet benzophenone with bovine and human serum albumins and suggested that such an observation is a result of poor confinement of the radicals by the protein environment compared to micelles where the confinement is efficient and often large MF effect is observed.⁵⁹ The proposed reaction mechanism, thus unveiled by the use of LFP with an associated MF, is represented in Scheme 1.

Fig. 10 (A) Transient absorption spectra of 2 × 10⁻⁴ (M) AD + 3 × 10⁻⁵ (M) HSA in phosphate buffer of pH 7.4 at 0.1 μs after the laser flash at 355 nm in (a) absence (■) and (b) presence (•) of MF. (B) Time-resolved transient absorption spectra of 2 × 10⁻⁴ (M) AD + 3 × 10⁻⁵ (M) HSA in presence of MF at 0.2 μs (■), 0.6 μs (•), 1.0 μs (▲) and 2.0 μs (▼) time-delay after the laser flash at 355 nm.

Table 4 Variation of lifetime of radical ions with MF for a solution of 2 × 10⁻⁴ (M) AD + 9 × 10⁻⁵ (M) HSA in phosphate buffer of pH 7.4

Wavelength	MF (T)	Lifetime (s)
520 nm (AD˙^−)	0	2.02 × 10^−6
	0.08	2.21 × 10^−6
560 nm (TrpH˙^+)	0	1.67 × 10^−6
	0.08	1.82 × 10^−6

---

Scheme 1 Proposed reaction pathways.

3.5. Docking results

AD is docked to HSA to corroborate the spectroscopic observations. As discussed in the introduction, the model protein contains 17 disulfide bridges and a series of nine loops, assembled in three domains (I, II, III), each containing two subdomains, A and B and the sole Trp-214 is located in binding site IIA. The docked conformation of HSA with AD is shown in Fig. 11 indicating that AD binds to domain IIA (Sudlow's site I) of the serum albumin.

Fig. 11 Docked conformation of HSA with AD.

Accessible surface area which is abbreviated as ASA, is an index of forming contacts between the atoms on the surface of a protein and the solvent (water) molecules. The detailed description of interacting residues of HSA is given in Fig. 12 which shows that maximum ASA change occurs with Arg 209 residue of the protein. The accessible surface area of HSA changes from 26 999.16 Ų to 26 825.37 Ų in presence of AD. Residues Arg 209, Lys 212, Ala 213, Val 216, Asp 324, Leu 327, Gly 328, Ala350, Glu354 of HSA are probably involved in the binding process since each of them undergoes a change of more than 10 Ų in their ASA after binding with AD. Binding interaction of the sequestered drug with these residues of the protein impart motional restriction to AD which is reflected in the findings of time-resolved anisotropy studies as discussed earlier. The free energy change of binding as obtained from docking study is −32.34 kJ mol⁻¹ which is comparable with the experimental value at 301 K.

Fig. 12 Plot of change in ASA of HSA against amino acid residues.

The distances of various atoms of Trp residue from different atoms of AD as obtained from docking studies are provided in Table 5. It is observed that there is a discrepancy between the drug-receptor distance obtained from fluorescence studies (19.6 Å) and that from docking studies (13.5 Å, considering the distance from imino-N or indole N of Trp). This may be attributed to the fact that since the direction of transition dipole moment of AD is not known it is difficult to predict the actual value of κ² in eqn (9) correctly. Owing to this limitation in the calculation of FRET, a difference is obtained between the experimental and docked protein-receptor distances. Besides, docking studies involve the attainment of optimized structure as well as orientations of the drug and the protein molecules so that the total free energy of the system is minimized. Thus, docking offers more significance to minimum energy conformation, although the involvement of other conformations is also feasible during the experiments. This may also lead to difference in the results of docking studies when compared to the experimental findings. Another issue which is pertinent in this context is that in the present system of AD and HSA, energy transfer is not the sole cause of fluorescence quenching of the protein. The phenomenon of PET, which has been confirmed by LFP technique, is also responsible for the quenching of intrinsic fluorescence of HSA in presence of AD. This may also result in slight disagreement of the experimentally and theoretically obtained values of protein–ligand distance.

Table 5 Distances of various atoms of Trp residue from different atoms of AD as obtained from docking studies

Drug	Trp 214 (IIB)	Distances (Å)
O [9-O]	N-amino	7.9
O [9-O]	N-imino	13.5
N [10-N]	O	9
C [1-C]	C	10.7

---

4. Conclusion

The present study primarily focuses on the photochemical interaction of AD with HSA. Differential absorption study indicates that AD forms a ground state complex with HSA, while CD spectroscopic study implies that AD brings about substantial alteration in the secondary and tertiary structures of the carrier protein. Fluorescence spectroscopy is utilized to determine the binding constant, binding site number and the thermodynamic parameters associated with the interaction. The values of binding constant (3.98 × 10⁵ M⁻¹ at 301 K) and number of binding site (1.32 at 301 K) are estimated from the quenching of tryptophan emission of HSA. The negative values of ΔH⁰ (−87.98 kJ mol⁻¹) and ΔS⁰ (−190.56 J mol⁻¹ K⁻¹) obtained from fluorescence study are indicative of involvement of hydrogen bonding and van der Waals interaction. The phenomenon of energy transfer in the singlet state is found to occur in AD–HSA system, which has been exploited to evaluate the mean distance between the tryptophan residue of the protein and the ligand moiety. Motional restriction imposed upon the acridine derivative on binding with the protein is clearly demonstrated by the time-resolved anisotropy study. TRES and TRANES analyses, which are quite rare in reports of drug–protein interaction, are carried out to take a deeper insight into the binding mechanism. However, the actual mode of interaction is unfolded by the use of LFP coupled with a weak MF. The occurrence of PET in AD–HSA system is confirmed by detecting the transients formed in the due course of reaction and their original spin state is authenticated by the use of MF. Although MF effect is usually obtained in micellar medium owing to proper confinement of the RIPs, nevertheless protein environment in the present system is also successful in providing pseudo-confinement to the RIPs. Previously, our group has reported that although PET takes place from hen egg white lysozyme to menadione on photoexcitation, but no MF effect is observed probably because of close vicinity of the acceptor and donor moieties (2.79 Å), where exchange interaction is not negligible, thus preventing spin conversion.⁵² A similar case of absence of MF effect is also obtained by our group while studying the interaction of 4-nitroquinoline-1-oxide and hen egg white lysozyme as the donor-acceptor separation distance is about 3.1 Å, which inhibits spin conversion.⁶⁰ On the contrary, in the present case of AD–HSA system the separation distance (13.5 Å) between the donor and the acceptor moieties of PET is fitting for appreciable MF effect. Our previous report shows that during interaction with HSA, two other acridine derivatives, viz. proflavine and acridine yellow maintain a separation distance of 12.93 Å and 13.34 Å from the Trp-214 residue, the electron donor moiety of the carrier protein, which results in observation of prominent MF effect.¹⁸ Thus, it may possibly be inferred that for acridone, acridine yellow and proflavine the bulky structure of acridine along with the functional groups do not allow these compounds to penetrate deep into the crevice at site IIA which house Trp-214, but strategically maintains the optimum distance from the tryptophan residue required for observation of MF effect. The functional groups attached to the acridine moiety for all the three derivatives of acridine prefer to participate in hydrogen bonding and it is also observed that contribution of hydrophobic interaction during binding to protein is not very significant in these cases. Therefore, these compounds do not enter deep into the hydrophobic crevice and cannot come in close vicinity with the buried tryptophan residue which results in retention of ideal distance required for conservation of spin correlation between RPs/RIPs. The significance of usage of a MF in a protein environment can be understood from the observation of Scaiano et al. that application of an external MF can at as a co-carcinogen in biological systems as MF-enhanced free radical attack may weaken the biological response to genotoxic stress and consequently promote increased mutagenesis and cancer.⁶¹ Hence, the knowledge obtained by integration of the present results with the previous findings may help to strategically choose and synthesize molecules which can act as ligands and exhibit prominent MF effect while interacting with biomacromolecules. This may provide sufficient guidelines to stimulate predictive insight into biological reactions and throw some light in the unexplored avenues of photoinduced phenomena involving drug–protein interaction.

Acknowledgements

Financial support from the Chemical and Biophysical Approaches for Understanding Natural Processes (CBAUNP) and Biomolecular Assembly, Recognition and Dynamics (BARD) projects, SINP of Department of Atomic Energy (DAE), Government of India is greatly acknowledged. P. M. acknowledges the Senior Research Fellowship from Council of Scientific and Industrial Research (CSIR), India. We would like to thank Dr Mousumi Banerjee, SINP for her suggestions. We would like to acknowledge Mr Ajay Das and Mrs Chitra Raha, SINP for their technical support.

References

1. U. E. Steiner and T. Ulrich, Chem. Rev., 1989, 89, 51.
2. K. Bhattacharya and M. Chowdhury, Chem. Rev., 1993, 93, 507.
3. B. Chakraborty and S. Basu, Chem. Phys. Lett., 2009, 477, 382.
4. B. Chakraborty and S. Basu, Chem. Phys. Lett., 2010, 487, 51.
5. B. Chakraborty and S. Basu, Chem. Phys. Lett., 2010, 493, 76.
6. R. W. Armstrong, T. Kurucsev and U. P. Strauss, J. Am. Chem. Soc., 1970, 92, 3174.
7. N. Higashi, M. Takahashi and M. Niwa, Langmuir, 1999, 15, 111.
8. M. Read, R. J. Harrison, B. Romagnoli, F. A. Tanious, S. H. Gowan, A. P. Reszka, W. D. Wilson, L. R. Kelland and S. Neidle, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 4844.
9. H. Baruah, C. S. Day, M. W. Wright and U. Bierbach, J. Am. Chem. Soc., 2004, 126, 4492.
10. N. H. Campbell, G. N. Parkinson, A. P. Reszka and S. Neidle, J. Am. Chem. Soc., 2008, 130, 6722.
11. C. Fritsch, G. Goerz and T. Ruzicka, Arch. Dermatol., 1998, 134, 207.
12. E. D. Varnell and H. E. Kaufman, Infect. Immun., 1973, 7, 518.
13. H. Satonaka, K. Kusuzaki, T. Matsubara, K. Shintani, T. Wakabayashi, T. Nakamura, A. Matsumine and A. Uchida, Anticancer Res., 2007, 27, 3339.
14. V. M. Rosenoer, M. Oratz and M. A. Rothschild, Albumin Structure, Function and Uses, Pergamon Press, Oxford, U.K., 1977.
15. C. Bertucci and E. Domernici, Curr. Med. Chem., 2002, 9, 1463.
16. J. Li, C. Ren, Y. Zhang, X. Liu, X. Yao and Z. Hu, J. Mol. Struct., 2008, 885, 64.
17. X. M. He and D. C. Carter, Nature, 1992, 358, 209.
18. B. Chakraborty, A. Singha Roy, S. Dasgupta and S. Basu, J. Phys. Chem. A, 2010, 114, 13313.
19. D. Dey, N. R. Pramanik and S. Basu, J. Phys. Chem. B, 2009, 113, 8689.
20. N. J. Turro and G. C. Weed, J. Am. Chem. Soc., 1983, 105, 1861.
21. Y. Tanimoto, S. Takase, C. Jinda, M. Kyotani and M. Itoh, Chem. Phys., 1992, 162, 7.
22. Y. Akimoto, Y. Fujiwara and Y. Tanimoto, Chem. Phys. Lett., 2000, 326, 383.
23. T. Miura, K. Maeda and T. Arai, J. Phys. Chem. A, 2006, 110, 4151.
24. F. Moreno, M. Cortijo and J. G. Jimenez, Photochem. Photobiol., 1999, 70, 695.
25. A. S. R. Koti, M. M. G. Krishna and N. Periasamy, J. Phys. Chem. A, 2001, 105, 1767.
26. A. S. R. Koti and N. Periasamy, J. Chem. Phys., 2001, 115, 7094.
27. M. Maroncelli and G. R. Fleming, J. Chem. Phys., 1987, 86, 6221.
28. J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Plenum Press, New York, 2006.
29. D. Sarkar, A. Mahata, P. Das, A. Girigoswami, D. Ghosh and N. Chattopadhyay, J. Photochem. Photobiol., B, 2009, 96, 136.
30. T. K. Maiti, K. S. Ghosh, A. Samanta and S. Dasgupta, J. Photochem. Photobiol., A, 2008, 194, 297.
31. B. Ahmad, S. Parveen and R. H. Khan, Biomacromolecules, 2006, 7, 1350.
32. G. M. Morris, D. S. Goodsell, R. S. Halliday, R. Huey, W. E. Hart, R. K. Belew and A. J. Olson, J. Comput. Chem., 1998, 19, 1639.
33. G. M. Morris, D. S. Goodsell, R. Huey and A. J. Olson, J. Comput.-Aided Mol. Des., 1996, 10, 293.
34. D. S. Goodsell and A. J. Olson, Proteins: Struct., Funct., Genet., 1990, 8, 195.
35. W. L. DeLano, The PyMOL, Delano Scientific, San Carlos, CA, USA, 2004, http://www.pymol.org.
36. O. Trott and A. J. Olson, J. Comput. Chem., 2009, 31, 455.
37. G. C. Pimentel, J. Am. Chem. Soc., 1957, 79, 3323.
38. F. D. Mamedov and M. A. Salimov, J. Appl. Spectrosc., 1971, 14, 333.
39. H. E. Strickland, M. Wilchek, J. Horwitz and C. Billups, J. Biol. Chem., 1972, 247, 572.
40. M. Hossain, A. Y. Khan and G. Suresh Kumar, PLoS One, 2011, 6, e1833.
41. S. M. Kelly and N. C. Price, Curr. Protein Pept. Sci., 2000, 1, 349.
42. S. Sugio, A. Kashima, S. Mochizuki, M. Noda and K. Kobayashi, Protein Eng., 1999, 12, 439.
43. G. M. Barenboim, A. N. Domanskii and K. K. Turoverov, Luminescence of Biopolymers and Cells, Plenum Press, New York, 1969.
44. S. Patel and A. Datta, J. Phys. Chem. B, 2007, 111, 10557.
45. N. Zhou, Y. Z. Liang and P. Wang, J. Photochem. Photobiol., A, 2007, 185, 271.
46. A. Barik, B. Mishra, A. Kunwar and K. I. Priyadarsini, Chem. Phys. Lett., 2007, 436, 239.
47. Y. V. Il'ichev, J. L. Perry and J. D. Simon, J. Phys. Chem. B, 2002, 106, 460.
48. B. Valuer, Molecular Fluorescence Principles and Applications, VCH, Weinheim, Germany, 2002.
49. M. Mukherjee, P. Saha Sardar, S. K. Ghorai, S. K. Samanta, A. Singha Roy, S. Dasgupta and S. Ghosh, PLoS One, 2013, 8, e60940.
50. A. S. R. Koti and N. Periasamy, J. Chem. Phys., 2001, 115, 7094.
51. P. D. Ross and S. Subhramanian, Biochemistry, 1981, 20, 3096.
52. S. Banerjee, S. Dutta Choudhury, S. Dasgupta and S. Basu, J. Lumin., 2008, 128, 437.
53. D. V. Bent and E. Hayon, J. Am. Chem. Soc., 1975, 97, 2612.
54. D. Creed, Photochem. Photobiol., 1984, 39, 537.
55. B. Chakraborty and S. Basu, Appl. Magn. Reson., 2012, 42, 5.
56. D. Das and D. N. Nath, J. Phys. Chem. A, 2008, 112, 11619.
57. S. Dutta Choudhury and S. Basu, J. Phys. Chem. B, 2006, 110, 8850.
58. S. Sheth, A. Baron, C. Herrero, B. Vauzeilles, A. Aukauloo and W. Leibl, Photochem. Photobiol. Sci., 2013, 12, 1074.
59. N. Mohtat, F. L. Cozens, T. Hancock-Chen, J. C. Scaiano, J. McLean and J. Kim, Photochem. Photobiol., 1998, 67, 111.
60. M. Banerjee, S. Maiti, I. Kundu, A. Chakrabarti and S. Basu, Photochem. Photobiol., 2010, 86, 1237.
61. J. C. Scaiano, F. L. Cozens and J. McLean, Photochem. Photobiol., 1994, 59, 585.

---