Comparative Study of Toluidine Blue O and Methylene Blue Binding to Lysozyme and Their Inhibitory Effects on Protein Aggregation.

A comparative binding interaction of toluidine blue O (TBO) and methylene blue (MB) with lysozyme was investigated by multifaceted biophysical approaches as well as from the aspects of in silico biophysics. The bindings were static, and it occurred via ground-state complex formation as confirmed from time-resolved fluorescence experiments. From steady-state fluorescence and anisotropy, binding constants were calculated, and it was found that TBO binds more effectively than MB. Synchronous fluorescence spectra revealed that binding of dyes to lysozyme causes polarity changes around the tryptophan (Trp) moiety, most likely at Trp 62 and 63. Calorimetric titration also depicts the higher binding affinity of TBO over MB, and the interactions were exothermic and entropy-driven. In silico studies revealed the potential binding pockets in lysozyme and the participation of residues Trp 62 and 63 in ligand binding. Furthermore, calculations of thermodynamic parameters from the theoretical docking studies were in compliance with experimental observations. Moreover, an inhibitory effect of these dyes to lysozyme fibrillogenesis was examined, and the morphology of the formed fibril was scanned by atomic force microscopy imaging. TBO was observed to exhibit higher potential in inhibiting the fibrillogenesis than MB, and this phenomenon stands out as a promising antiamyloid therapeutic strategy.

Introduction

Binding interaction of various photoactive organic small molecules with proteins has evoked great interest in medicinal chemistry. The nature of protein–ligand binding effects, delivery rate, and therapeutic efficacy are important information for drug-design and development. Detailed biophysical studies on the dye–protein interaction help in understanding the structural features in terms of the bioaffinity and pharmacokinetic behavior of the dyes on the protein domain.[1−3]

Lysozyme (N-acetylmuramide glyconohydrolase, lyz), the protein present in abundance in various protective fluids and lymphatic tissues of most animals, has antimicrobial properties with the potential to destroy the bacterial cell wall. Lyz has wide utility as a food preservative also.[4,5] Anti-inflammatory, antiviral, antiseptic, antihistamine, and antineoplastic properties of the protein are also known. Lyz is a low-molecular (∼14 kDa) globular protein having α-helix, β-sheet turns, and disordered structural elements. Its primary structure has 129 amino-acid residues, including six tryptophan (Trp) and three tyrosine (Tyr) residues and four disulfide bonds. Six Trp residues are located at the substrate-binding sites, among these two are located in the hydrophobic matrix box. Studies have speculated Trp 62 and 108 to be the most supreme fluorophores in lyz.[6−8] Hen egg white lyz shares around 60% homology with its human counterpart, which is included in hereditary nonneuropathic systemic amyloidosis.[9] Lyz has been used as a model protein because of its natural abundance, high stabilization, and compact size for studying the protein-structure principles, function, dynamics, ligand bindings, and aggregation properties.

It is necessary to have correctly folded structure of protein for their biological action to be normal. Most of the diseases, for example, Alzheimer’s, Parkinson’s, Huntington’s, Senile systemic amyloidosis, type II diabetes, and so forth are the root of abnormal function of proteins, which occurs because of the aggregation of proteins.[10,11] In agreement with various studies, lyz aggregates under some destabilizing conditions, such as low pH (pH 2.0) and high temperatures (>50 °C) to form amyloid fibrils.[9−12] Recent reports demonstrated that the synthetic lyz fibrils exhibited nonenzymatic cytotoxicity, including inducing aggregation and hemolysis of human erythrocytes, triggering intermolecular cross-linking of cytoskeletal proteins, and reducing the viability of neuroblastoma cells through apoptotic and necrotic pathways.[13,14] Under in vitro conditions amyloid fibrils are produced by generating harsh conditions such as high temperature, high pressure, acidic or alkaline pH, or mixing of cosolvents, metal ions, lipid assemblies, and surfactants.[15−17] The rupture of fibrillar assemblies and inhibition of amyloid formation are the therapeutic blueprint recommended for the treatment of amyloid-related diseases. Enzymatic inhibitors, antagonists, antibodies, hormones, small peptides, synthetic ligands, and natural polyphenols have been probed to be used as antiamyloidogenic agents. In this context, various polyphenolic compounds reportedly demonstrated the potential to inhibit the amyloid formation.[18,19] Most importantly, lyz protein has effective potential to carry drugs to the target receptors, and the efficacy of such drug targeting depends on their binding capacity. For this strategy, interaction of small organic molecules such as toluidine blue O (TBO) and methylene blue (MB) with lyz which inhibit the fibril formation have been studied and have gained great interest.

Toluidine blue O (2-methyl-3-dimethylamino-7-amino-phenothiazin-5-iumchloride) (Figure A) is a phenothiazinium group planar cationic dye that has anti-tumor activity.[20] TBO has some activity on selective cancer cells and has the potential to be developed as an anticancer drug.[21] TBO is widely known to have antimicrobial properties when used as photosensitizers.[22] TBO is an effective photosensitizer against planktonic bacterial and fungal growth, which has shown to reduce the cell viability of many microorganisms.[23] It has great applications in the field of gynecology during colposcopy and detection of oral premalignant and malignant lesions.[24,25] It is also used as sensors, biosensors, polymerization inhibitors, staining agents, photosensitizers to determine the action of photoactivated microorganisms, and so forth.[26−28]

Figure 1

Chemical structure of (A) toluidine blue O and (B) methylene blue, and structure of (C) lyz.

Methylene blue (3,7-bis(dimethylamino)-phenothiazin-5-ium chloride) (Figure B) is also a phenothiazinium cationic aromatic dye, which has the potential to stain the nucleic acids and has antimalarial activity. Methylene blue is widely used as a photosensitizing agent for photodynamic inactivation of RNA viruses including HIV, hepatitis B virus, and hepatitis C virus in plasma.[29,30] Contemporary studies have suggested that dengue virus can be inactivated by the MB/narrow-bandwidth light system.[31] Also, MB produces singlet oxygen (1O2) from triplet molecular oxygen in the presence of light.[32]

In this work, we have evaluated the binding interaction of TBO and MB to lyz using various spectroscopic and calorimetric techniques and performed molecular docking to predict the binding sites on lyz. Furthermore, a detailed molecular study of the dye-induced fibril inhibition in lyz has been characterized. The amyloid fibril formation from the lyz was investigated by various spectroscopic tools such as thioflavin T (ThT) fluorescence spectroscopy, congo-red (CR) absorption spectroscopy, and far-UV circular dichroism (CD) spectroscopy. Conformational properties of amyloid lyz were analyzed by far UV CD spectroscopy. Furthermore, a structural and morphological view of lyz fibril was examined by the atomic force microscopy (AFM) study.

Thus, understanding the binding characteristics of small molecules to lyz may provide an insight to develop some new small molecular inhibitors in the therapy of amyloid diseases.

Results

Fluorescence Quenching of Lyz Induced by Dye Molecules

Fluorescence studies are used to evaluate the binding interaction of small molecules to proteins[33−36] because of the influence of photophysical sensitization of protein fluorophores to surrounding polarity. Lyz contains six Trp residues at positions 28, 62, 63, 108, 111, and 123 (Figure C).[33] The residues at positions 28, 108, 111, and 123 are located in the α-domain and those at positions 62, 63, and 108 are located in the substrate-binding cleft. Trp-62 and 63 are situated in between the hinge regions of α and β domains. From the literature studies, it has been demonstrated that Trp-62 and 63 are the most exposed residues located at the solvent site and therefore highly susceptible to chemical modifications.[35,37] When lyz is partially unfolded, Trp-62 and 63 become more exposed to light, consequently enhancing the intensity of the protein. Trp 62 and 108 residues are mainly responsible for fluorescence emission of lyz, whereas the residues at positions 28, 63, 108, 111, and 123 are less emissive. The Trp 63 residue resides closely to the active-site and is not buried in the hydrophobic core. However, Trp-62 is fully exposed to the solvent site and Trp-108 is away from the hydrophilic environment.[37] Consequently, more exposed Trp-62 and 63 residues are the main moieties to be affected or quenched by the interaction of the ligand. Hence, spectral variations of intrinsic fluorescence characteristics of lyz with TBO and MB dyes may give the finding of mode of binding interaction. The fluorescence study suggested the information about the surrounding environmental domain of lyz. For this purpose, fluorescence experiments were performed with excitation wavelength at 295 nm, where only Trp residues of lyz are excited, and it provides the emission spectrum maximum around 336 nm. Figure depicts the intrinsic fluorescence of Trp moieties of lyz accompanying quenching with the progressive addition of the dyes, TBO and MB, respectively. In the titration of TBO and MB with lyz, quenching was observed accompanying the formation of a new peak at 348 and 347 nm, respectively.

Figure 2

Steady-state fluorescence spectral changes of lyz (0.85 μM) on addition of 0, 0.17, 1.20, 1.89, 3.26, 6.02, 11.52, 17.02, 22.50, 28.03, 33.54, and 50.05 μM of (A) TBO (curves 1–12) and 0, 0.17, 1.20, 3.95, 6.71, 9.46, 14.96, 20.47, 25.97, 31.48, 42.48, and 53.49 μM of (B) MB (curves 1–12).

Fluorescence quenching occurred mainly by two processes, static quenching by complex formation or dynamic quenching due to the collisional process during photoexcitation.[38] In static quenching, an increase in temperature reduces the stability of the formed complex, consequently lowering the static quenching constant. On the other hand, with the increase of temperature, in the case of dynamic quenching, the number of colliding photoactive molecules is increased as well as larger the diffusion coefficient, resulting in the increase of the quenching constant. Therefore, to reaffirm the quenching mechanism, the fluorescence experiment was done at three variable temperatures, viz at 288.15, 298.15, and 308.15 K. The quenching phenomenon was analyzed by the well-known Stern–Volmer equation.[38]where, Fo and F signifies the fluorescence intensities of lyz (mainly Trp moiety) at wavelength maxima with and without the presence of quencher (dyes), respectively. [Q] represents the quencher concentration, Kq denotes the biomolecular quenching constant, and τo (it is taken to be 108 s–1) is the average lifetime of lyz without the quencher. KSV is the Stern–Volmer quenching constant.

The resulting linear relationship of Fo/F against [Q] primarily shown in Figure suggests that the quenching is either static or dynamic. Furthermore, the values of Kq and KSV (Table ) decreased as the experimental temperature was increased; this result legitimate supported the fact that the quenching is being governed by the static mechanism. Moreover, the magnitudes of the Kq values were greater than 2.0 × 1010 M–1 s–1, and this finding also reinforces the prevalence of the static mechanism.[33−35]

Figure 3

Stern–Volmer plots at three different temperatures for lyz quenching by (A) TBO and (B) MB, respectively, are presented.

Table 1

Binding Data Derived from Spectrofluorimetric Studies at Different Temperaturesa

dye	temperature (K)	KSV (M–1)	Kq (M–1 s–1)	KA (M–1)	N	KLB (M–1)
TBO	288.15	(1.53 ± 0.05) × 105	(1.53 ± 0.05) × 1013	(1.64 ± 0.09) × 105	0.71	(1.52 ± 0.06) × 105
	298.15	(1.01 ± 0.05) × 105	(1.01 ± 0.05) × 1013	(1.13 ± 0.08) × 105	0.81	(1.11 ± 0.06) × 105
	308.15	(7.74 ± 0.04) × 104	(7.74 ± 0.04) × 1012	(8.92 ± 0.08) × 104	1.08	(8.01 ± 0.05) × 104
MB	288.15	(9.93 ± 0.05) × 104	(9.93 ± 0.05) × 1012	(9.70 ± 0.07) × 104	1.09	(9.74 ± 0.05) × 104
	298.15	(7.87 ± 0.04) × 104	(7.87 ± 0.04) × 1012	(8.19 ± 0.06) × 104	0.97	(8.61 ± 0.05) × 104
	308.15	(6.12 ± 0.03) × 104	(6.12 ± 0.03) × 1012	(6.37 ± 0.05) × 104	1.29	(6.48 ± 0.04) × 104

All the data in this table are the average of four determinations.

The phenothiazinium dyes TBO and MB are highly fluorophore compounds. Therefore, the effect of lyz on the fluorescence emission of the dyes was also examined, and the spectra are represented in Figure S1. In both cases, the quenching of emission spectra also suggested the complex formation at the ground state.

Fluorescence Lifetime Study

Fluorescence lifetime measurements were performed to distinguish whether the quenching is static or dynamic.[39] In static quenching, the measured fluorescence lifetime values were invariant with the concentration of a quencher, whereas in dynamic quenching there was a noticeable variation in fluorescence lifetime data.[39] Fluorescence lifetime measurements along with their amplitudes for free and dye-(TBO/MB) bound lyz were deduced from their respective time-resolved fluorescence decay profiles to interpret the dynamics of lyz in the presence of the dyes. Excitation and emission wavelengths were set to be 280 and 337 nm, respectively. The light signal scattered from LUDOX provided the instrumental response function which was utilized for the deconvolution of the fluorescence signals. Decay curves of lyz and its complexes with TBO and MB were fitted with a biexponential function and the goodness of the fits were evaluated from both the χ2 values and the residuals of the function fitted to the data. Fluorescence decay is given by[40]where F(t) = fluorescence intensity at time t and α = preexponential factor with respect to the ith decay time constant, τ. For multiexponential decay, the average lifetime τavg is given bywhere τ = fluorescence lifetime and α = relative amplitude with i ranging between 1 and 2. For free lyz, the fluorescence lifetimes were deduced to be τ1 = 1.07 ns and τ2 = 2.54 ns, whereas the fluorescence lifetime were τ1 = 1.11 ns and τ2 = 2.63 ns in the presence of TBO. In the presence of MB, the fluorescence lifetime values were τ1 = 1.14 ns and τ2 = 2.88 ns. The Trp residues divulge multiexponential decays;[35] therefore we have not assigned independent components but the average fluorescence lifetime values have been reported to obtain a qualitative analysis. Average fluorescence lifetime of lyz was 1.93 ns, whereas its complexes with TBO and MB had average fluorescence lifetime values of 1.94 and 1.95 ns, respectively. Hence, time-dependent fluorescence experiments revealed that the fluorescence lifetime of free and lyz complexes with the dyes were not significantly changed. These studies suggested that the quenching of lyz fluorescence is static in nature and is due to ground state complexation.

Absorbance Titration

Absorbance spectral titration was also performed to support the static quenching mechanism and the absorption changes were recorded in the visible region, that is, in the 450–800 nm wavelength region. The absorption maxima of TBO and MB dyes are 633 and 664 nm, respectively. The interaction of lyz with these dyes is presented in Figure S2. The spectral changes of dye–protein composite systems supported the argument of dye–protein complex formation in the ground state.

Binding Parameter Elucidation

Besides determining the Stern–Volmer quenching constant (KSV), we can also evaluate the apparent binding constant (KA) and the number of binding sites (n) by the given equation as

From the plot of log[(Fo – F)/F] versus log[Q], the apparent binding constants (KA) and number of binding sites (n) were calculated at three variable temperatures and the data are depicted in Table . With the increase of temperature, weakening the apparent binding strength (KA) revealed the static quenching statement as discussed earlier. The calculated number of binding sites “n” obtained from the above equation was nearly about unity, indicating only one type of binding site for the binding interaction, most likely around the Trp residue of the protein.

The quenching data were quantitatively calculated by the double reciprocal plot dictated by the Lineweaver–Burk equation[33]where KLB is the quenching constant calculated from the ratio of the intercept to the slope of the Lineweaver–Burk plot, presenting the efficacy of quenching at the ground state. At three variable temperatures, from the plot of protein–dye interactions, the data were calculated and presented in Table , and it is revealed that TBO has greater affinity toward lyz than MB. Moreover, the variation of KLB with increasing temperature was similar to that of the KSV values. All these results supported the static quenching phenomenon, which is the driving force for the fall in Trp residual emission with the increment of the ligand concentration.

Anisotropy

A steady-state anisotropy result implies the degree of rotational restriction of small molecule binding with macromolecules. Any modulation in the flexibility or tumbling motion of the immediate surroundings of the fluorophore probe can be examined by the anisotropy plot. The anisotropy technique was based on the principle that larger molecules tumble slower than the smaller molecules, resulting in an increase in the anisotropy values that correspond to the increased rotational restriction. From Figure , it is shown that binding of dye molecules to lyz resulted in an increase in the anisotropy value. This indicated that motional restriction is increased because of strong binding. Furthermore, the anisotropy value is increased from 0.109 to 0.118 upon binding of TBO to lyz. However, the change in anisotropy is marginal upon binding of MB to lyz. This observation indicated that rotational rigidity imposed a larger increase of TBO than MB, resulting in the stronger binding interaction of TBO with the protein. Moreover, the binding constant (K) can be measured from the results of anisotropy (r) of dyes with the binding to lyz according to Ingersoll and Strollo’s postulate[34,41]

Figure 4

Anisotropy variations of (A) TBO and (B) MB on treatment with the incremental concentration of lyz. Inset diagram represents corresponding double reciprocal plots of 1/fb vs 1/[L] to calculate the binding constant.

Ingersoll and Strollo describes the following equation aswhereHere, R is the correction fraction which is calculated by the ratio of fb to ff, and fb and ff indicate the fractional fluorescence intensities of the bound and free forms of dyes, respectively. rb and rf are their corresponding anisotropy. [L] represents the lyz protein concentration in mM. Using the above relationship, double reciprocal plots of 1/fb versus 1/[L] are plotted and the corresponding binding constants were calculated from the slopes. The calculated binding constants of TBO and MB to lyz are deduced to be 1.06 × 105 and 6.75 × 104 M–1, respectively; these values are in close agreement with those obtained from spectroscopy measurements.

8-Anilino-1-naphthalenesulfonic acid (ANS) Displacement Study

The ANS displacement study was performed to locate the preferred binding region of dye molecules on lyz protein. ANS was used as a potential probe for getting the information about the hydrophobic binding sites on the protein.[42,43] Studies on binding of dye molecules to lyz were performed in the presence of ANS keeping the lyz to ANS ratio as 1:0, 1:1, and 1:10, and the relative intensity of fluorescence (F/F0) was plotted as a function of concentration of dye molecules. From Figure , it is shown that dye molecules moderately compete with ANS for the hydrophobic binding region of lyz, and it is higher in TBO than in MB. Thus, TBO binds more strongly in the hydrophobic site of lyz by displacing the ANS molecules.

Figure 5

ANS displacement profiles for the fluorescence quenching of lyz by (A) TBO and (B) MB at various lyz/ANS ratios (1:0, 1:1 and 1:10).

Conformational Changes of the Protein from Synchronous Fluorescence Spectroscopy

The ligand binding to lyz causes conformational changes that can be analyzed by synchronous fluorescence.[44] Miller proposed that, if the excitation and emission wavelength difference (Δλ) are set to be at 15 or 60 nm, the synchronous fluorescence spectra of the protein revealed a scenario in the polarity change around the Trp and Tyr amino acid moieties.[45] In the binding of lyz to dyes, there is a systematic decrease in the emission spectra with a large red shift (8 nm for TBO) and a small blue shift (5 nm for MB) of emission maxima when Δλ was set to be 60 nm (Figure ). This is governed by the postulate that there is a change in the Trp moiety toward the hydrophilic environment exposing more to the solvent site in the case of TBO, and the corresponding blue shift signified the reverse phenomenon for Try moieties in MB.[46] On the other hand, there is less change in the emission maxima when Δλ was set to be 15 nm, indicating that there is little transformation occurred in Tyr moieties. This finding reinforced that binding of TBO and MB with lyz causes the change in polarity around the Trp residues and most likely around Trp-62 and Trp-63. This observation suggested, unequivocally, the participation of the Trp moieties in the binding process.

Figure 6

Synchronous fluorescence spectra of lyz (10 μM) on treatment with 0, 2, 6, 12, 20, 28, 36, 44, 52, 68, and 76 μM of (A) TBO (curves 1–10) and (B) MB (curves 1–10) when Δλ = 60 nm, and 0, 6, 12, 24, 40, 52, 68, 84, 108, 124, and 156 μM of (C) TBO (curves 1–11) and (D) MB (curves 1–11) when Δλ = 15 nm.

Circular Dichroism (CD) Spectral Study

CD spectroscopy is extensively employed to scan the structure, conformation, and stability of proteins in solution.[47] Though CD could not represent a specific structure of proteins as CD spectroscopy is considered as a “global reporter”, but it can provide a guesstimate of fraction of the residues in the lyz protein containing α-helix, β-sheet, and random coils. Chicken egg white lyz is an α + β protein with a large α-domain incorporated with four α-helices and a 310-helix and a smaller β-domain involving a triple-stranded antiparallel β-sheet, and an irregular loop containing two disulphide bridges. Therefore, lyz contains mostly of α-helix domains. To examine the α-helical characteristics of lyz, the far UV CD spectrum of native lyz was recorded. Figure A,B shows that lyz contains two minima at 208 and 222 nm, which mainly occurs because of the α-helical domain.[33−35] The 208 nm band occurred because of the π–π* transition of the α-helix and the 222 nm band corresponds to the n−π* transition for both the α-helix and random coil. From the CD spectra, it is evident that with the increase of dye concentration the CD signals decreases systematically, suggesting the alteration in the secondary structure of lyz upon interaction with these dyes. The lowering of the negative ellipticity values suggested the rupture of the α-helix content, indicating unfolding of the peptide strand of native protein on interaction with the dyes. The helical content of the free and bound lyz molecules was measured in terms of mean-residue ellipticity (deg cm[2] dmol–1) as reported earlier.[48,49]where C represents the lyz concentration at molarity, n is the number of amino-acid residues, and l denotes the cuvette path length.

Figure 7

Far-UV spectral changes of lyz (10 μM) with the addition of 0, 5, 10, 15, and 25 μM of (A) TBO (curves 1–5) and 0, 5, 10, 20, and 30 μM of (B) MB (curves 1–5). Near-UV spectral changes of 0, 6, 14, 30, and 64 μM of (C) TBO (curves 1–5) and (D) MB (curves 1–5), respectively.

The α-helical values of free lyz and the corresponding protein bound by dyes were calculated from the relation as

From the above equation, it was calculated that lyz contains 33.48% of the α-helix character, which is in good agreement to literature values.[33−35] The α-helical character on dye binding was reduced and deduced to be 20.25 and 25.36%, respectively, for TBO and MB. Both the dyes induced strong secondary structural changes manifested by the loss of α-helix stability. The binding also caused the unfolding in lyz with the extended polypeptide chains, revealing the hydrophobic cavities with concomitant exposure of the aromatic amino-acid residues.

Near-UV CD spectral (Figure C,D) experiments were conducted to decipher the tertiary structural changes in lyz induced by binding with dyes. In the 250–300 nm region, the CD spectral changes of lyz occurs because of the existence of Trp, Tyr, and Phe residues and disulphide bonds. The CD spectrum exhibits three positive peaks near 283, 289 and 295 nm, which corresponds to the transitions of Trp moieties in lyz.[33−35] There were no drastic changes in this region of CD spectra on interaction of dyes, indicating that dye molecules do not have much effect to induce major tertiary structural changes as no break of disulphide bonds occurred. Though the inference is not conclusive, the CD spectra provide us an overall estimation of structural changes of protein with the interaction of dye molecules. Overall, CD spectroscopy and steady-state fluorescence measurements supported the strong binding interaction of TBO and MB with lyz protein.

Differential scanning calorimetry (DSC) Study

Binding of small molecules to macromolecules generally alter their thermal stability by increasing or decreasing their transition temperature Tm (thermal denaturation temperature). To record the DSC transition temperature (Tm) of lyz protein in the absence and the presence of TBO/MB, DSC was performed. DSC also enables the detection of the energetics associated with the temperature-dependent protein folding–unfolding process.[50,51] The DSC transition profile of lyz showed a single endothermic peak at 339.85 K, and upon complexation with TBO and MB, the transition temperatures of lyz yielded to be, respectively, 338.0 and 339.3 K (Figure S3). Therefore, the DSC result suggested that there is little destabilizing effect of dye molecules when it binds with lyz.

ITC Study

A quantitative estimation of the drug–protein interaction is essential because of its binding propensity, and its effect on its stability will govern the therapeutic efficiency of the drug. The ITC thermograms of lyz–TBO and lyz–MB are presented in Figure . In the top panel, each heat burst spikes of raw ITC profiles signifies single injection of lyz into dye solutions. The bottom panel corresponds to the actual corrected heat of reaction of lyz–dye binding after subtracting the heats of dilution. Here, the points manifested the experimental heats of injection, and the solid lines correspond to the best fitted isotherms.[33−35,52,53] Both the ITC thermograms revealed exothermic reaction. The ITC data of these dye–protein interactions are depicted in Table . The summarized thermodynamics data and their sign provided a way to estimate the binding forces of dye–protein interactions. It is shown from the ITC data that electrostatic interactions are dominant in the lyz–TBO/MB binding as ΔH < 0 and ΔS > 0.[54] In both cases, the interactions are spontaneous in nature as ΔG < 0. However, TBO is more appropriate for ionic interactions as the size and substitutions are comparable with the H-bonding and induce higher binding affinity in the lyz–TBO interactions. Thus, the ITC data corroborated the steady-state fluorescence data. Both the dyes, TBO and MB, bind in a 1:1 ratio as observed from N values (Table ). Both the bindings were driven by positive entropy changes with minimum favorable negative enthalpy changes. It is assumed that large positive TΔS° values are the outcome of perturbation, release of counter ions, and release of solvated water molecules from the lyz–dye binding site.[55] Moreover, the calculated positive ΔS° and negative ΔH° values are accountable for the specific electrostatic interactions of TBO/MB with lyz in aqueous medium.[56]

Figure 8

ITC figures for the titration of lyz–TBO/MB complexation at 298.15 K. Upper panels denote the raw heat for the injection of 1.5 mM lyz into (A) 40 μM of TBO and (B) 50 μM of MB. Control titrations of dilution for the heat of lyz into corresponding buffers are shown at the bottom (not in scale). The bottom panel represents the integrated heat after the subtraction of heat of dilution. The symbols (■) denote the data points, and the solid line represents the fitted curve of the “one set of binding sites” model.

Table 2

Thermodynamics Parameters for the Binding of Dyes with lyz from ITC at Different Temperaturesa

dye	temperature (K)	K (M–1)	N	ΔH° (kcal mol–1)	TΔS° (kcal mol–1)	ΔG° (kcal mol–1)	ΔCp° (kcal mol–1 K–1)
TBO	288.15	(1.78 × 105) ± 0.03	1.65	–0.290 ± 0.04	6.60 ± 0.03	–6.89 ± 0.02
	298.15	(1.20 × 105) ± 0.05	1.28	–0.334 ± 0.03	6.56 ± 0.03	–6.90 ± 0.03	–4.78 ± 0.09
	308.15	(0.82 × 105) ± 0.04	1.57	–0.385 ± 0.03	6.51 ± 0.02	–6.89 ± 0.03
MB	288.15	(1.00 × 105) ± 0.04	1.15	–0.330 ± 0.05	6.24 ± 0.02	–6.57 ± 0.02
	208.15	(0.79 × 105) ± 0.07	1.14	–0.353 ± 0.05	6.31 ± 0.02	–6.66 ± 0.03	–2.90 ± 0.03
	308.15	(0.61 × 105) ± 0.05	1.10	–0.388 ± 0.03	6.33 ± 0.03	–6.72 ± 0.03

All the data in this table are the average of three determinations.

At three different temperatures, ITC experiments were performed to know the nature and vector of forces governing the binding reactions. From the temperature-dependent ITC experiments, the standard molar heat capacity changes (ΔCp°) can be determined using the standard relation as

Binding affinity of dye–protein interactions decreased with the increment of temperatures, suggesting that lyz–TBO/MB complex formation is destabilized at higher temperatures. The variation of ΔH° against T yielded ΔCp° for the interactions, which is depicted in Figure S4. From the slopes of ΔH° versus T, ΔCp° values are calculated as −4.78 and −2.9 cal/mol K for TBO and MB, respectively. Negative ΔCp° values are the characteristic features of many biomolecular interactions.[33−35,52,53,57]

The nonzero negative ΔCp° values imply specific changes involving hydrophobic or polar group hydration on binding of these dyes to lyz. It is also suggested that the shift of the nonpolar surface area also contribute to the observed ΔCp° value.[58,59] A relatively higher value of ΔCp° indicated a much higher extent of hydrophobic desolvation effects and conformational changes in lyz to TBO in conformity with the CD results. The hydrophobic contributions of binding of TBO and MB to lyz were determined from the equation, ΔGhyd° = (80 ± 10) × ΔCp°.[60] The ΔGhyd° values are −334.6 and −203.0 cal/mol K, respectively, for TBO and MB.

Molecular Docking and Dynamic Simulation Study

Our docking experiments were performed with lyz as the candidate protein, and the TBO and MB were considered as candidate ligands. The ambient temperature chosen for the docking studies was 298.15 K. The binding-site dynamics study revealed that there exists three binding pockets in lyz (Figure A). Among these, two pockets (green and fade-pink pocket) is of prime importance in the context of binding to both of the above-mentioned ligands. One of these pockets consist of both the Trp residues 62 and 63 (green pocket). Another amino-acid residue Trp 108 is enclosed within another authoritative pocket (fade-pink pocket). From Figure B,C, it was observed that in the most prominent ligand-bound states, both the ligands have strong interactions with the above-mentioned two pockets.[61] Furthermore, they were found to be well-organized inside the cleft formed in between the two pockets. Trp residue 123 lies within another pocket (light-blue) that barely has significant impact on ligand binding. From Figure B,D, it can also be inferred that TBO has strong interaction with the green pocket zone because almost all the immediately near residues of this candidate ligand belong to that pocket. Similarly, Figure C,E manifests the interaction of MB with the residues in or near to the pink pocket when it is embedded into the cleft. From the protScale output plot, we could infer that in both the cases, the immediate interacting residues are hydrophobic in nature. Thus, it can be inferred that the hydrophobic core is masked when the candidate protein is in interaction with the two ligands.

Figure 9

(A) Probable binding pockets in lyz as predicted by the FT algorithm. Green, blue, and fade pink meshes represent the potential binding sites, which may control the internal dynamics of the protein. Panels B and C represent binding pockets of toluidine blue O (shaded in violet mesh) and methylene blue (shaded in green mesh), respectively, in lyz as obtained as the docking output. Panels D and E depict the nearest binding sites of TBO and MB in lyz, respectively.

In the case of TBO, the residues within 2 Å of the ligand are Trp 108, Trp 62, Trp 63, Val 109, Asn 59, and Asp 48. These are those residues which have an immediate interaction with the ligand when the ligand approaches the cleft of the receptor protein. The protScale output plot shows a hydrophobic propensity from residue 110 to 115 (Figure S5). These are those residues which remain at a distance greater than 2 Å but not more than 5 Å. The residues within 5 Å distance from the ligand are 35, 44, 47, 48, 52, 56, 57, 58, 59, 61, 62, 63, 101, 102, 103, 110, 112, 113, and 114, among which residues 44, 47, 48, and 101 are hydrophilic in nature. This reflects that TBO has a significantly higher propensity to bind to hydrophobic residues as only four out of 19 residues are hydrophilic. On the other hand, when MB binds to lyz, residues Trp 62, Trp 63, Val 109, and Arg 112 happen to be those residues lying with 2 Å distance from the bound ligand. Furthermore, the residues within 5 Å from MB are 34, 35, 44, 45, 47, 48, 50, 52, 56, 57, 75, 108, and 110. Residues 44, 45, 47, 48, and 50 are hydrophilic in nature.

According to our docking experiments, lyz–TBO and lyz–MB associations correspond to the change in free energy −6.93 and −6.58 kcal mol–1, respectively. Using these values, for both the acceptors, we have calculated the binding constants, shown in Table , which almost merge with the experimental values. Here, we used the following equationwhere ΔG is the change in the Gibbs energy, T refers to the temperature in kelvin (K), and KD represents the binding constant of the ligand with the protein.

Table 3

Binding Data from Docking Study

protein	dyes	ΔG° (kcal mol–1)	KD (M–1)
lyz	TBO	–6.93	1.20 × 105
	MB	–6.58	6.69 × 104

Inhibitory Effect on Lyz Fibrillogenesis Monitored by ThT and CR Assay

At low pH and high temperatures, amyloid fibrillogenesis of lyz occurred. The parameters such as temperature, ionic strength, and concentration of lyz effect the lag time of fibril formation.[13,19,62,63] In this article, the inhibition of fibrillogenesis was examined by the presence of TBO and MB with lyz. For this reason, lyz was dissolved (10 mg/mL) in glycine–HCl buffer of pH ∼2.20 and heated at 333.15 K with continuous stirring in the presence and absence of 40 μM TBO and 60 μM MB. To observe the formation and growth of amyloid fibrils, ThT fluorescence and CR absorption spectroscopy experiments were performed.

ThT fluorescence spectroscopy is based on the notable feature of this dye as it forms highly fluorescent complexes with amyloid and amyloid-like fibrils.[64,65] ThT does not interact with globular proteins in a native state. It also does not interact with molten globules and unfolded states or amorphous aggregates of proteins. Therefore, the spectral property of ThT depends on the fibril formation. It is believed that a notable enhancement in the fluorescence quantum yield or fluorescence intensity was observed when the dye is incorporated into the highly ordered linear array of the β-sheet structure of amyloid fibrils. Therefore, ThT shows a sudden spectral enhancement when it binds with fibrils, and this phenomenon can be used to monitor the fibrillogenesis kinetics.[66,67]

The ThT fluorescence intensity at 485 nm for the lyz sample increased sharply by attaining an equilibration plateau (Figure A). To examine the inhibitory effect of TBO/MB on fibrillogenesis, the ThT fluorescence intensities of lyz samples in the presence and absence of the dyes were compared. Figure A represents the sigmoidal curve of the varying ThT fluorescence intensities as a function of time. Each curve represented a lag phase (no significant change in fluorescence intensity) followed by a growth phase (sudden notable increment in fluorescence) and finally, the saturation phase where ThT fluorescence reaches a plateau. From Figure A, it was observed that there is a notable delay in fluorescence increment from 7 to 8 h for TBO and 6 to 7 h for MB, suggesting that both dyes effectively delayed the growth of amyloid fibrils. However, the influence of TBO is more in inhibiting the fibrilogenesis. For lyz the lag phase was observed until 5 h, whereas for lyz samples in the presence of TBO and MB, the lag phase lasted till 7 and 6 h, respectively. This lag phase was accompanied by a growth phase in the range 5–6, 7–8, and 6–7 h for lyz and lyz coincubated with TBO and MB. The growth phase was followed by an equilibration plateau in all the cases. Hence, from the detailed results, it can be concluded that coincubation of lyz with TBO (40 μM) and MB (60 μM) resulted in the reduction of ThT fluorescence, suggesting both the dyes had an inhibitory effect on fibrillogenesis. The percentage of ThT fluorescence reductions at 9 h incubation were calculated using the following equationwhere, Io and I denote the ThT fluorescence of lyz fibrils in the absence and presence of TBO/MB. TBO and MB reduced the ThT fluorescence of mature fibrils by 80 and 76%, respectively, at 9 h incubation.

Figure 10

Variation of (A) fluorescence intensity of ThT at 485 nm and (B) absorbance of CR at respective absorption maxima at different time intervals for lyz samples (■) in the absence and presence of TBO (●) and MB (▲).

The CR binding assay is a complementary path to identify the amyloid fibril formation in lyz protein. Amyloid fibrils showed apple-green birefringence in cross-polarized light when it was stained with CR.[12,68,69] Moreover, a significant reduction in the absorption spectra along with a bathochromic shift at the maximum of CR to ∼540 nm was observed, when lyz fibrils binds with CR. The variations in CR absorption spectra (at respective absorption maxima) for lyz in the absence and presence of TBO and MB at different time intervals were plotted in Figure B. From the plot, it is was noticed that there is a significant increment in the absorption of lyz samples stained with CR along with a concomitant red shift in the absorption maxima. This observation suggested the presence of an ordered fibrillar structure of β strands in the lyz protein. Binding with TBO and MB, there was a decrease in absorption maxima, which revealed the inhibition in the formation of amyloid fibrils associated with the cross-β-pleated sheet. These dyes also delayed the time of fibril formation in lyz (Figure ). Therefore, the ThT fluorescence and CR absorbance technique unequivocally elucidate that both the dyes have a concentration-dependent mitigating effect on the aggregation/fibril formation of lyz.

Conformational Changes in Lyz in the Presence of TBO/MB

To monitor the effect of TBO/MB on the changes of secondary structures associated with the fibril formation in lyz, the far-UV CD spectra were obtained. Initially, the CD spectra of lyz without and with TBO/MB showed minima at 209 and 222 nm, which are the characteristics of the α-helix structure (Figure ). However, upon prolonged heating and continuous stirring, the far-UV CD spectrum of lyz exhibited a noticeable structural transition of α-helix to a highly amyloidogenic β-sheet-rich conformation. The conformational transition was arrested by a characterized minimum at 218 nm. The conformational transition in lyz arose at 6 h (Figure A), but in the presence of TBO and MB, the structural transformations were delayed at around 7.3 h (Figure B) and 6.5 h (Figure C), respectively. From the figures, it can be seen that in the presence of TBO, the α-helix conformation was intact even at 7 h of incubation, but in the presence of MB, the incubation was at 6 h. Therefore, the outcome of the observation revealed that the TBO delayed the α-to-β transition of lyz more potently than MB.

Figure 11

Far-UV CD spectra of (A) lyz at 1 (curve 1), 6 (curve 2), 7 (curve 3), and 8 (curve 4) h of incubation. Panels B and C denote far-UV CD changes of lyz in the presence of 40 μM of TBO and 60 μM of MB, respectively, at 1 (curve 1), 6 (curve 2), 7 (curve 3), and 8 (curve 4) h of incubation.

With the binding of phenothiazinium dyes, the minima of CD spectra varied in the region of 216–220 nm, depending upon the incubation time, nature, and concentration of the dyes used (Figure ).

AFM Imaging of Fibrils

AFM imaging was performed to investigate and gather further insights into the inhibitory effect of the dyes on fibrillogenesis. AFM imaging suggested morphological changes of lyz in the absence and presence of dyes at different incubation time intervals. Figure A shows a bunch of matured and dense fibrils at 9 h of incubation in lyz, but in the presence of TBO (Figure B) and MB (Figure C), the fibrils were found to be less-populated, broken, and shorter in length and height.

Figure 12

AFM images of lyz in the absence (A) and in the presence of TBO (B) and MB (C) after 9 h of incubation.

Thus, AFM imaging unequivocally concluded that TBO and MB inhibits the amyloid fibril formation effectively.

Conclusions

The binding interactions of TBO and MB to the protein lyz were studied by experimental and theoretical biophysical approaches under physiological conditions. Spectroscopic, calorimetric, as well as molecular docking suggested that TBO possesses higher binding affinity toward lyz than MB. Both dyes induced microenvironmental changes of lyz, but the induced alteration was found more in Trp moieties than in Tyr. On the other hand, the intrinsic fluorescence experiment suggested the static quenching in lyz–TBO/MB-binding interactions, and the phenomena were legitimated by the result of lifetime measurements. Both the dyes induced strong secondary structural changes by the loss of α-helix stability. From the ITC experiment, the binding constant was found in the order of 105 M–1 in both the cases, but the higher binding of TBO may be due to the large hydrophobic contributions and H-bonding, which were also observed from the molecular docking study. Moreover, TBO inhibited lyz fibrillogenesis at 40 μM concentration, and for MB, it was 60 μM. Also, TBO delayed the lyz fibril formation in large time-scale than MB. From the study, we can conclude that both dyes strongly bind with lyz and have inhibitory effects on amyloid fibrillogenesis; moreover, TBO showed higher binding affinity as well as greater inhibition on lyz aggregation than MB. Therefore, the current study provides an insight into the antiamyloidogenic role of phenothiazinium dyes and also shed light on the rational design of novel antiamyloidogenic therapeutics.

Materials

MB (>95% pure, Riedel-de Haën, GmBH, Germany) was bought from Sigma-Aldrich GmBH. Hen egg white lyz (purity ≥98%, M = 14.3 kDa) and TBO (CAS no. 92-31-9, purity 80%) was bought from Sigma-Aldrich. The commercial sample of lyz was purified as previously reported.[70] The sample was then desalted on a Sephadex G-50 column, dialyzed in the cold (Spectra/Por MWCO, 3500 membrane), and freeze-dried. Na-phosphate buffer (10 mM Na+, 5 mM Na2HPO4) of pH 7.2 was used to carry out all the studies. Sartorius (GmBH, Germany) PB-11 high-precision pH meter was used to measure the pH of buffer solutions. Buffer solutions were filtered with the help of Millipore filters (Millipore Bangalore) of 0.22 mm. TBO was purified by column chromatography on neutral alumina using ethanol/benzene (7:3 v/v) containing 0.4% glacial acetic acid. The fractions were pooled, concentrated under vacuum, and crystallized. The crystals were dried in a vacuum desiccator at room temperature to give spectrally pure dye.[71] The purity of the MB solution was checked by measuring the ratio of the absorbance at 665 nm to that at 610 nm, which was always greater than 2.1, indicating the absence of any demethylated dye in the sample.[72] The concentration of lyz, TBO, and MB was calculated by measuring the absorbance using a molar absorption coefficient (ε) values of 37 750 M–1 cm–1 (280 nm),[33−35] 30 000 M–1 cm–1 (633 nm),[52] and 76 000 M–1 cm–1 (664 nm),[72] respectively. The dyes obeyed Beers’ law in the concentration range used in the experiment. Stock solutions of the dyes used in this study are freshly prepared and placed in the dark to protect from photochemical changes.

Methods

Steady-State Spectroscopic Study and Lifetime Measurements

A JASCO V660 spectrophotometer (Hachioji, Japan) was used to carry out the absorption study, and the instrument was attached with a thermo electrical controlling cell holder. The experiments were conducted in quartz cuvettes (Hellma, Germany) of 1 cm path length at (293 ± 0.5) K. The titration procedures are described in detail previously.[33−35]

Steady-state fluorescence titrations were conducted either in PTI QM-400 (HORIBA Canada) or using a RF5301-PC spectrofluorimeter (Shimadzu, Kyoto, Japan) in fluorescence-free quartz cuvette (1 cm) as described.[33−35,52] All titrations were executed setting excitation and emission band passes of 5 and 5 nm, respectively. For synchronous fluorescence, the excitation range was 220–340 nm and Δλ was set at 15 and 60 nm, respectively.

Lifetime measurements were performed using PTI QM-400 instrument.

Circular Dichroism Study

The JASCO spectropolarimeter J-815 equipped with a temperature controller was used to perform the CD experiments. The experiment was directed by the thermal programmer (425L/15) of JASCO software at 293.15 ± 0.5 K.[33−35,52,73] The CD spectra were recorded at a scan speed of 50 nm min–1 along with a step size of 0.5 nm. The sensitivity of the CD was 10 millidegree, and the bandwidth was 0.2 nm. To modify the signal to noise ratio, about 15 accumulations were carried out, and the spectra were smoothed within permissible limits.

Moreover, far UV spectra of lyz fibrils in the absence and presence of dyes were also recorded in the 190–250 nm region to monitor the conformational changes. For this purpose, 100 μL of the sample solution was diluted with 200 μL of glycine–HCl buffer, and then the CD spectra were recorded at different time intervals using a 0.1 cm path length cuvette.

Isothermal Titration Calorimetry

Isothermal titration calorimetry measurements were carried out on a VP-ITC MicroCal (MicroCal) instrument at 293.15 K. The revolving syringe was loaded with lyz solutions, and 10 μL of lyz was injected each time into the cell containing dye solutions. The syringe rotation was set to be at the speed of 307 rpm, and the spacing of the injections was 240 s. Control experiments were executed in which lyz solutions were titrated into the buffer solutions under the same conditions. These reference heats of dilution were subtracted from the corresponding dye–lyz complex reactions to obtain the exact heat of the binding processes. The heat of the reaction of dye–lyz binding was evaluated by integrating the area under each heat burst curve.[33−35] The ITC data were analyzed by Microsoft Origin software and fitted with “one site binding model” to get the standard thermodynamics parameters. Binding constant (K), binding stoichiometry (N), standard molar enthalpy change (ΔH°), standard Gibbs energy for the complex formation (ΔG°), and the changes of entropy (TΔS°) were derived from standard relationships, ΔG° = −RT ln K (R = 1.9872 cal/mol K, T = 298.15 K) and ΔG° = ΔH° – TΔS°. Temperature-dependent ITC were conducted by varying the temperature, and it gave the information about the standard molar heat capacity changes (ΔCp°). The calorimeter was calibrated with water–water dilutions as described by the manufacturer so that the mean energy <1.30 μcal for each injection, and the standard deviation was <0.015 μcal.

Differential Scanning Calorimetry

The residual heat capacities of the protein and its dye complexes were estimated from the DSC study. It may also be used to measure thermal melting transition temperatures for the compounds. The experiment was performed on a MicroCal LLC, VPDSC unit, USA (now Malvern Instruments, Malvern, UK). By repeating the buffer scanning, the instrument was settled thermally stable with a scanning rate of 333.15 K per hour and at 25 psi pressure. The baseline was deducted from the complex thermograms. The DSC profiles were scanned by Origin 7.0 software to provide the model independent calorimetric transition enthalpy (ΔHcal).[52,53]

Molecular Docking

With an objective to have a mechanistic insight into the protein—ligand association we resorted to theoretical approaches. We went on to have the docking output by resorting to molecular docking. The co-ordinate information of lyz was retrieved from the Protein Data Bank (PDB) (PDB ID: 1DPX). The energy structure of the candidate ligand molecules (toluidine blue O and methylene blue) was minimized using the Avogadro 1.1.1 molecular editor, and the MOL2 format was used as the ligand input for docking analysis. The predock processing of the lyz co-ordinate file which was involved in the removal of water molecules and all bound ligand molecules was executed using the Dock Prep plugin of UCSF Chimera molecular viewer. Protein–ligand docking was performed with SwissDock, which relies on the back-end software EADock DSS in the CHARMM force field. The lowest binding energy was chosen from among the most possible 10 top order docking conformations. The result was analyzed by DockView plugin of UCSF Chimera.

Binding-Site Dynamics

The virtual study involving the FT algorithm was resorted to predict the probable binding sites of lyz. The study was pertinent as it could help unveil the mechanistic mode by which TBO and MB can dock them over the lyz-binding cleft. The solvent-mapping algorithm in FTSite is actually a direct computational analog of these screening techniques. In this approach, a computational mapping is implemented that actually controls the individual placement of each of the 16 disparate small molecular probes over a dense grid around the protein. It in turn finds the favorable positions employing empirical free energy functions. For each probe type, the individual probes are then clustered and the clusters are also ranked on the average free energy. Then, consensus clusters are recognized as sites in which different probe clusters overlap. Depending upon the total number of nonbonded interactions present between the protein and all probes in the cluster, the consensus clusters are ranked. Here, the consensus cluster in which the highest number of contacts is present is ranked first, also nearby consensus clusters are joined with this cluster. The top ranked predicted ligand binding site comprises those amino acid residues that are in contact with the probes of the newly defined cluster. Clusters with fewer contacts represent lower ranked predictions.

Lyz Fibrils Formation

In glycin–HCl buffer lyz samples (10 mg/1 mL) were dissolved at pH ∼2.20, containing 100 mM Na-chloride and 1.54 mM NaN3.[12] For preparation of amyloid fibrils, 2 mL of lyz stock solution was transferred to a 25 mL glass tube, diluted to 20 mL by glycine–HCl buffer, and then stirred with polytetrafluoroethylene-coated microstirring bars at 220 rpm and at 333.15 K temperature.[12,74,75] The glass tubes are vortexed gently to homogenize the samples at a required time gap, and then 1 mL aliquots of the sample were taken for the following experiments.

ThT Fluorescence Experiments

ThT fluorescence experiments were conducted on a Shimadzu spectrofluorimeter (RF-5301 PC). ThT stock solution was prepared, and its concentration was measured by the procedure reported in the literature.[12,76] For this purpose, 150 μL of lyz aliquots in the absence and presence of TBO/MB were diluted with 760 μL of buffer at desired time intervals. Then, an appropriate amount of ThT was mixed to the solution to make the total ThT concentration to 20 μM. Fluorescence intensities at 485 nm were recorded in slow-speed by exciting the resultant mixtures at 416 nm.

CR Binding Assay

The CR binding experiment was performed to detect the fibril formation in lyz in the absence and presence of dyes. For this purpose, CR was dissolved in dimethyl sulfoxide to prepare 1 mM concentrated solution. 50 μL of lyz solutions were diluted with 1 mL of buffer solutions followed by the addition of CR solution so that the final concentration of CR in the total solution was 10 μM. Then, the solution mixtures were incubated at room temperature for 30 min. The absorption spectra of lyz in the absence and presence of TBO/MB were recorded in the 400–800 nm region. The absorbance value at corresponding absorption maxima of the dyes were recorded and plotted against the corresponding time intervals.

AFM Imaging

For the AFM imaging study, the solutions of lyz fibrils were diluted to 100-fold with water. Then, around 5 μL of this sample solution was adsorbed onto a freshly cleaved muscovite ruby mica sheet (ASTM VI grade Ruby Mica from Micafab, Chennai, India). Thereafter, the mica sheet was dried for 30 min in vacuum in an inert atmosphere. The complexes were incubated for 15 min prior to adsorption onto the mica sheet. AFM was performed in the AAC mode on PicoPlus 5500 ILM AFM (Agilent Technologies, USA), which was attached with a piezo-scanner of a maximum range of 9 μm. Here, microfabricated silicon cantilevers of NANOSENSORS (USA) were used. The resonance frequency of the cantilever oscillation was 146–236 kHz, whereas the force constant was 21–98 N/m. The rate of the scan speed was 0.5 lines/s while taking the images (256 by 256 pixels). All the images were processed by flattening using PicoView software (Agilent Technologies, 1.1 version), whereas their manipulation was conducted by Pico Image Advanced version software.