Fluorescence Resonance Energy Transfer, Small-Angle Neutron Scattering, and Dynamic Light Scattering Study on Interactions of Gemini Surfactants Having Different Spacer Groups with Protein at Various Regions of Binding Isotherms

Abstract

The binding interactions of three gemini surfactants having different spacer groups (12-4-12, 12-8-12, and 12-4(OH)-12) with a high concentration (150 μM) of bovine serum albumin (BSA) at various regions of binding isotherms have been studied by means of steady-state fluorescence and fluorescence anisotropy, time-correlated single-photon counting fluorescence of trans-2-[4-(dimethylamino)styryl]benzothiazole, small-angle neutron scattering (SANS), and dynamic light scattering (DLS) measurements. The fluorescence resonance energy transfer phenomenon between the twisted intramolecular charge transfer fluorescent molecule, trans-2-[4-(dimethylamino)styryl]benzothiazole as an acceptor, and tryptophan 213 (Trp-213) of BSA as a donor has been successfully used to probe the binding interactions of gemini surfactants with protein at all regions of binding isotherms. The increasing order of energy transfer efficiency at a higher concentration range of surfactants is 12-8-12 > 12-4-12 > 12-4(OH)-12. Stronger binding of micelles of gemini surfactant molecules having a comparatively more hydrophobic spacer group with the hydrophobic segments of the protein results in closer approach of trans-2-[4-(dimethylamino)styryl]benzothiazole molecules solubilized in micelles to Trp-213. The average excited-state lifetimes become shorter with a trend of increase in contribution from the fast component and decrease in contribution from the slow component to the decay with increasing concentration of a surfactant. The nonradiative rate constant of trans-2-[4-(dimethylamino)styryl]benzothiazole increases with increasing concentration of a surfactant because the average microenvironment around it in protein–surfactant aggregates is more polar as compared to that in native protein. SANS and DLS measurements were carried out for the study of the structural deformations in the protein, on enhancement of the concentration of the gemini surfactants. The necklace and bead model has been used for the analysis of SANS data for the protein–surfactant complexes. At a higher concentration range, 12-8-12 and 12-4-12 have a slightly smaller fractal dimension and a larger correlation length as compared to 12-4(OH)-12. DLS data show that the increasing order of hydrodynamic diameter for the complexes of protein with three gemini surfactants in their high concentration range is 12-4(OH)-12 < 12-4-12 < 12-8-12.

1. Introduction

Protein is one of the main building blocks of life. Serum albumins belong to a multigene family of proteins.¹ They play major roles in various physiological functions as they can be found in the human circulatory system. They even contribute to the osmotic blood pressure and also act as a plasma carrier.² They also take part in transportation of a variety of molecules such as some anesthetic drugs, amino acids, fatty acids, steroids, metabolites, and so on.³ Bovine serum albumin (BSA) is one such pivotal protein commonly used for research purposes because of its stability, water solubility, and flexible binding property. Basically, BSA consists of 583 amino acid residues in a single polypeptide chain (molecular weight = 66.5 kDa). It comprises nine loops which are bound together by 17 disulfide bonds, enclosed in three domains (I, II, and III), each consisting of two subdomains, A and B. BSA has two tryptophan (Trp) residues at positions 134 and 213 of the amino acid sequence. Tryptophan 134 (Trp-134) is found in the hydrophilic subdomain, whereas tryptophan 213 (Trp-213) is present in the hydrophobic core of BSA.⁴

Protein can interact with different types of ligands such as fatty acids, drugs, metal ions, surfactants, and so on.⁵⁻⁸ The study of protein–surfactant interactions has been an attractive area of research to the scientific community over the past few decades. This research area covers a variety of fields starting from the estimation of protein molecular weights to the preparation of effective washing powder enzymes and other such products required for personal hygiene. A number of sophisticated instrumental tools, e.g., circular dichroism,⁹ nuclear magnetic resonance (NMR),¹⁰ microcalorimetry,¹¹ light scattering,¹² small-angle neutron scattering (SANS),¹³ steady-state and time-dependent fluorescence,^14,15 electron spin resonance,¹⁶ were witnessed over a short span of time for more accurate and sophisticated analysis of protein–surfactant complexes. Apart from its use as a model system to understand the mechanism of interactions between protein and ligands in biological systems, the knowledge of behavior of protein–surfactant interactions is also useful in several applications such as biotechnological processes, cosmetic systems, detergents, drug delivery, biosciences,¹⁷⁻¹⁹ food, biochemistry, and the pharmaceutical industry.²⁰ Surfactants are known to uncoil the original protein structure. Proteins comprise both hydrophobic and hydrophilic amino acids and because of this, amino acids allow the surfactant molecules to bind with the proteins.²¹ In many cases, the chemical nature of surfactants governs the interactions occurring between protein and surfactant molecules. There are several reports²²⁻²⁵ where intrinsic fluorescence of BSA, that is, Trp fluorescence, has been employed for the analysis of the structure and dynamics of BSA and also for the examination of different binding aspects between protein and surfactant.²⁶

Several reports detailing the interaction of diverse single-chain surfactants with protein are available. In recent times, a special category of surfactants called gemini surfactants has garnered tremendous significance as compared to conventional monomeric surfactants.²⁷ A gemini surfactant molecule possesses at least two ionic or polar headgroups and two hydrophobic chains connected by a spacer group at their heads.²⁸ A spacer group can be hydrophilic or hydrophobic, flexible or rigid.²⁸ The nature of spacer groups controls the aggregation behavior and thermodynamics of micellization of gemini surfactants.^29,30 An ionic headgroup may have positive (ammonium) or negative (phosphate, sulfate, carboxylate) charge, whereas a nonionic headgroup can be polyether or sugar. A vast majority of gemini surfactants have symmetrical structures with two identical chains and two identical headgroups. A gemini with two C_m (C_m is the number of alkyl carbon atoms) tails and an s (s is the number of alkyl carbon atoms) spacer separating the quaternary nitrogen atoms may be represented as C_m-s-C_m. The present study deals with the binding interactions of gemini surfactants of varying spacer groups with protein, BSA.

Recently, charge transfer probe molecules have been widely explored for their high responsiveness in order to interpret the binding interactions between protein and probe and also to probe the uncoiling of protein by various denaturing agents.^31,32 De et al.³³ investigated the interactions between BSA and three different monomeric surfactants, namely, sodium dodecyl sulfate (SDS), cetyl trimethylammonium bromide, and TX-100, using an external fluorescent probe, 1-anilino-8-naphthalene sulfonate, by utilizing the twisted intramolecular charge transfer (TICT) fluorescence properties. Mukherjee and his group used a probe, ethyl p-(dimethylamino)cinnamate, showing charge transfer fluorescence to study protein–SDS interactions.³⁴ Ghosh et al.³⁵ explored trans-3-(4-monomethylaminophenyl)-acrylonitrile as a charge transfer fluorescent probe for the study of binding of the probe to BSA. They located the exact binding site of the probe using silver nanoparticles with the support of molecular docking. In the present study, the protein–gemini surfactant interactions in different zones of binding isotherms have been probed exploring the fluorescence resonance energy transfer (FRET) between a TICT fluorescent probe, trans-2-[4-(dimethylamino)styryl]benzothiazole (DMASBT) (Scheme 1), and BSA. The study reveals that Trp-213, located in the hydrophobic domain of BSA, acts as a donor and DMASBT acts as an acceptor. Gemini surfactants used in this study are 12-4-12, 12-4(OH)-12, and 12-8-12 (Scheme 1). The effect of the chemical nature of the spacer group on interactions between gemini surfactant and BSA has been studied. Earlier, we had studied BSA–SDS interactions probed with FRET between DMASBT and BSA. In the present study, we have shown how FRET data corroborate well with the other fluorescence data such as fluorescence intensity and anisotropy. FRET data also have been used to support the effect of the spacer group on binding interactions. In addition to using the FRET method, we have also explored SANS and dynamic light scattering (DLS) measurements to highlight the microstructures of protein–surfactant complexes.^36,37 The neutron-scattering data on interactions between lithium dodecyl sulfate (LDS) and BSA were analyzed based on the “necklace and bead” model of protein–surfactant aggregates in which LDS micelles were distributed along the polypeptide chain.³⁶ A recent study of SANS analysis^36,37 on BSA and OVA (ovalbumin) complexes with SDS has inferred that the necklace and bead structure of protein surfactant aggregates is the reason behind the scattering behavior of the system. To have proper scattering signals in SANS measurements, we had to perform experiments with high concentration of BSA (1 wt %, i.e., 150 μM). When doing this study with a high concentration of BSA, we found some differences as compared to our previous study with a low concentration of BSA.³⁸ With a high concentration of BSA, we have found the expulsion of some percentage of DMASBT molecules from beads, that is, micelles with the addition of a high concentration of gemini surfactants.

Scheme 1. Molecular Structures of DMASBT and Gemini Surfactant Molecules.

2. Results and Discussion

2.1. Binding of DMASBT with BSA

2.1.1. Absorption and Steady-State Fluorescence Emission of DMASBT in BSA

DMASBT shows a broad and structureless absorption band in (4-(2-hydroxyethyl))-1-piperazineethanesulfonic acid (HEPES) buffer maintained at pH 7.4, thereby displaying a peak maximum at 402 nm. A wide, unstructured fluorescence band is observed having an emission maximum at 505 nm in aqueous HEPES buffer solution (pH = 7.4) upon excitation at 370 nm. It is to be noted that at this excitation wavelength, DMASBT molecules only selectively get excited without exciting BSA. Hardly any change in absorption spectra is noted in the presence of BSA. However, fluorescence spectra of DMASBT are immensely affected with variation in the concentration of BSA. Figure 1 demonstrates the changes in the fluorescence intensity of DMASBT as the concentration of BSA is increased. In our previous study, we recorded similar spectra up to 125 μM concentration of BSA.³⁹ However, in the present case, spectra have been taken up to 200 μM BSA because denaturation of BSA by the gemini surfactant has been studied for 150 μM BSA. The alterations in the fluorescence peak maxima of DMASBT with enhancement in BSA concentration are represented by Figure 1 (inset). It is noteworthy that the hydrodynamic radius of native BSA at 150 μM (discussed in Section 2.6) is found to be ∼4.1 nm. It is in good agreement with the reported value of 4.0 nm.⁴⁰ The hydrodynamic radius of native BSA at 5 μM is measured as ∼4.1 nm. The reported hydrodynamic radius of native BSA at 15 μM is 3.8 nm.⁴¹ Thus, the hydrodynamic radius of native BSA at a low concentration is very close to that at 150 μM within the limit of experimental error. Therefore, we can assume that there is as such no aggregation of protein molecules at the high concentration chosen in the present study.

Figure 1.

Fluorescence spectra of DMASBT (5 μM) as a function of concentration of BSA. Inset: Variation in fluorescence peak maxima of DMASBT as a function of concentration of BSA (λ_ex = 370 nm).

It has been reported in the literature that the intensity of fluorescence of DMASBT that occurs from the TICT state in a polar environment is low because of the high rates of nonradiative processes.⁴² As the protein concentration is gradually increased, a greater number of DMASBT molecules reside in a region of lower polarity, that is, the hydrophobic region. As a result, fluorescence intensity is increased with a blue shift in fluorescence peak maximum. Fluorescence band maximum of DMASBT gets blue-shifted by 30 nm at 110 μM BSA as compared to the buffer solution. The stoichiometry of the complex between BSA and DMASBT and binding constant, K, determined from the Benesi–Hildebrand plot^43,44 (see Note 1 and Figure S1 in the Supporting Information for details) are found to be 1:1 and 15561 M^–1, respectively, at 298.15 K. The standard Gibbs-free energy change (ΔG°) calculated using the K value was −24 kJ mol^–1 at 298.15 K. The spontaneity of binding is represented by the negative value of ΔG°.⁴⁵

2.1.2. Steady-State Fluorescence Anisotropy of DMASBT in BSA

Steady-state fluorescence anisotropy experiments were performed in order to inspect the motional restriction imparted by the protein on the DMASBT molecule. The changes in fluorescence anisotropy (r) with increasing concentration of BSA is represented by Figure S2. The fluorescence anisotropy rises with the gradual increase in the concentration of BSA, thereby denoting that the flexibility of the environment around DMASBT decreases as it binds with BSA. The effective binding of DMASBT with the protein is confirmed by a significantly large fluorescence anisotropy value of 0.37 at 50 μM BSA.

2.2. Unfolding of BSA by Gemini Surfactants Probed by DMASBT

The fluorescence spectra of DMASBT in aqueous HEPES buffer solution of 150 μM BSA with increasing concentration of each of the three gemini surfactants have been recorded. The fluorescence spectra for 12-8-12 are represented by Figure 2. Similar fluorescence spectra are observed with the other two gemini surfactants, 12-4-12 and 12-4(OH)-12, as well. The concentration range chosen for each surfactant has started from a concentration much below the critical micelle concentration (cmc) and goes up to a concentration much above the cmc, that is, 0–20 mM. The cmc values of 12-4-12, 12-4(OH)-12, and 12-8-12 reported earlier are 1.17, 0.96, and 0.72 mM, respectively.^46,47Figure 3a demonstrates the changes in TICT fluorescence intensity ratios (F/F_o) of DMASBT at 475 nm as the concentration of each of the three gemini surfactants varies. With increase in the concentration of a surfactant, fluorescence intensity initially increases and then decreases. After reaching a minimum, again it starts increasing and attains a maximum. Thereafter, once again it starts decreasing. The concomitant changes in fluorescence peak maximum are represented by Figure 3b.

Figure 2.

Changes in fluorescence spectra of DMASBT (5 μM) with variation of concentration of 12-8-12 (0–50 mM) in 150 μM BSA.

Figure 3.

(a) F/F_o, (b) fluorescence peak maximum (λ_max^fl), and (c) steady-state fluorescence anisotropy (r) of DMASBT with varying concentrations of gemini surfactants (-■-■- 12-4-12, (red)-□-□- 12-8-12, (green)-□-□- 12-4(OH)-12). [DMASBT] = 5 μM, [BSA] = 150 μM, λ_exc = 370 nm, λ_em = 480 nm. The insets in (a,b) show changes more clearly at a low concentration range.

Initially, at a very low concentration, a few surfactant molecules bind to the high-energy binding sites of BSA. It has been reported in the literature that this specific binding makes the protein structure more compact.³⁴ As a result, DMASBT molecules experience a comparatively less polar environment, leading to increase in TICT fluorescence intensity (Figure 3a). DLS measurements were performed for BSA–gemini complexes with 0.1 mM of each of the three surfactants. The hydrodynamic diameters of these complexes are found to be 8.1 and 7.9 nm for 12-4-12 and 12-4(OH)-12, respectively, as compared to 8.3 nm for the native protein. The data are indicating the compactness of the protein structure with a low concentration of surfactant, expecting that the changes in sizes should be low. However, we did not notice any significant change for 12-8-12, which could be because a surfactant with a long spacer chain interacts less effectively at a low concentration range with the protein.³⁸ Therefore, the effect on fluorescence intensity because of a slightly higher hydrophobicity imparted to the native BSA as a result of binding of a few surfactant molecules cannot be ruled out. On further increase in the concentration of surfactant, the protein molecules get unfolded. The probe molecules, DMASBT, get exposed to the polar environment because of noncooperative binding of the surfactant. Because of this, there is a gradual reduction in the fluorescence intensity with a notable red shift in the fluorescence band (Figure 3b). Having attained a minimum intensity at ∼2.5 mM gemini surfactant, there is a massive cooperative binding of the gemini surfactant molecules with BSA in the form of micelles giving necklace and bead kind of structures.^48,49 DMASBT molecules solubilized in these micelles experience a comparatively less polar environment, resulting in enhancement of fluorescence intensity (Figure 3a) with a slight hypsochromic shift (Figure 3b). In contrast to the observation with a low concentration of BSA,³⁸ in the present case, with a high concentration of BSA, the decrease in fluorescence intensity noted at a high concentration of a surfactant is because a few DMASBT molecules get expelled into the bulk aqueous phase.

To further demonstrate the binding interactions of gemini surfactants with BSA protein, the steady-state fluorescence anisotropy measurements have been accomplished.⁵⁰Figure 3c represents the variation in steady-state fluorescence anisotropy (r) of DMASBT at a fixed concentration of BSA (150 μM) with an increasing concentration of each of the three gemini surfactants. With denaturation of protein, the anisotropy is decreased as the probe molecules are exposed to the comparatively less rigid environment. The tumbling motions of DMASBT molecules become feasible. After reaching a minimum, fluorescence anisotropy starts increasing in the cooperative massive binding region of surfactant molecules as the DMASBT molecules start getting solubilized in a rigid environment in micelles formed along the protein chain. Beyond this region, fluorescence anisotropy is decreased as DMASBT gets expelled.

To check to what extent the change in fluorescence intensity is correlated with the change in fluorescence anisotropy, we have overlaid these plots in Figure 4 for 12-8-12 as a representative one. The same is done for the other two surfactants as well. The change in fluorescence anisotropy corroborates quite well with the change in fluorescence intensity except at a certain concentration range of surfactant. Figure 4 depicts some random changes in fluorescence anisotropy in a noncooperative interaction zone. Such an alteration in anisotropy would reflect that the random change in conformation of protein is taking place when the surfactant molecules noncooperatively bind with the protein. The flexibility of the binding environment surrounding DMASBT is expected to be altered with a conformational change of protein. The two other gemini surfactants, 12-4-12 and 12-4(OH)-12, show similar changes as well. Protein conformational changes upon binding with various molecules have been reported in the literature.⁵¹ The results show that fluorescence anisotropy of DMASBT is quite sensitive toward the alterations in the microenvironment, thereby allowing it to probe this conformational change of protein molecules. The commonly used probe, pyrene, lacks this ability.⁵² It is to be noted that the fluorescence anisotropy value of DMASBT in BSA–12-8-12 aggregates is comparatively higher (0.27) than that in the system containing pure micelles of 12-8-12 (0.23) at its 20 mM concentration. Also, the emission peak maximums of DMASBT in BSA–12-8-12 aggregates and in protein-free micelles of 12-8-12 are 473 and 518 nm, respectively, at 20 mM 12-8-12. Thus, DMASBT in a protein–surfactant system encounters a comparatively hydrophobic and rigid environment, as 12-8-12 micelles remain swathed by the protein chain, leading to the formation of necklace and bead kind of structures. Further, the fluorescence anisotropy of DMASBT in BSA-12-8-12 aggregates with a low concentration of 12-8-12 is comparatively greater than that in BSA-12-4-12 and BSA-12-4(OH)-12 systems. This is because the micelles of 12-8-12 surfactant molecules residing along the protein chain bestow a relatively more rigid environment to the fluorophore because of the longer spacer group as compared to 12-4-12 and 12-4(OH)-12.⁴⁴ Thus, the BSA–gemini surfactant complex, characterized by TICT fluorescence properties of DMASBT, accords quite well with the necklace and bead model.⁵³

Figure 4.

Changes in TICT fluorescence intensities and fluorescence anisotropies of DMASBT as a function of the concentration of 12-8-12; [BSA] = 150 μM, λ_ex = 370 nm.

2.3. Fluorescence Lifetimes of DMASBT in BSA–Gemini Surfactant Aggregate Systems

Fluorescence lifetimes of DMASBT have been recorded in BSA–gemini surfactant systems at their various regions of binding isotherms. The fluorescence lifetime of DMASBT in pure water is very short, which could not be measured as it is beyond our instrument’s detection limit.⁵⁴ DMASBT possesses a short lifetime in aqueous medium because there is a high rate of nonradiative processes occurring between the emitting TICT state and the triplet state as well as the ground state.⁵⁵ However, we could measure the lifetime in a proteinous environment as on an average the microenvironment around DMASBT is nonpolar as compared to bulk water. As a result, the emitting state gets destabilized and the rates of nonradiative processes diminished. To have decays selectively from DMASBT in native BSA and also in BSA–surfactant systems, it was excited at 375 nm and decays were recorded at 510 nm. BSA neither has any absorbance at 375 nm nor has any emission at 510 nm. Decays of DMASBT are mostly biexponential in native BSA as well as in BSA–surfactant systems. Many other groups have reported a multiexponential decay of different types of probes in a proteinous environment.⁵⁶ The lifetimes of decay components along with their weightage are given in Tables 1–3 for DMASBT (10 μM) in BSA (150 μM)–12-4-12, BSA (150 μM)–12-4(OH)-12, and BSA (150 μM)–12-8-12 systems, respectively. The average excited-state lifetime for a biexponential decay has been calculated using eq 1

1

where τ₁ and τ₂ are the lifetimes of the two components, and a₁ and a₂ are the corresponding pre-exponential factors, respectively.⁵²

Table 1. Excited Singlet-State Lifetimes (τ), Fluorescence Quantum Yields (Φ_f), and Radiative (k_r) and Nonradiative (k_nr) Rate Constants of DMASBT in BSA (150 μM)-12-4-12 Systems^a.

[12-4-12] (mM)	τ1 (ns)(a1)	τ2 (ns) (a2)	χ2	⟨τ⟩ (ns)	Φf	kr (s–1) × 10–6	knr (s–1) × 10–9
0.0	0.29 (0.54)	1.09 (0.46)	1.05	0.66	0.0013	1.98	1.52
0.5	0.37 (0.59)	1.24 (0.41)	1.12	0.73	0.0010	1.38	1.37
2.5	0.35 (0.71)	1.14 (0.29)	1.04	0.58	0.0021	3.45	1.72
3.5	0.33 (0.75)	1.09 (0.25)	1.19	0.52	0.0007	1.35	1.92
5.0	0.21 (0.75)	0.75 (0.25)	1.05	0.35	0.0010	2.90	2.90
10.0	0.18 (0.77)	0.63 (0.23)	1.07	0.28	0.0009	3.18	3.52
20.0	0.16 (0.78)	0.53 (0.22)	1.00	0.24	0.0008	3.31	4.14

^aλ_ex = 375 nm, λ_em = 510 nm, [DMASBT] = 10 μM.

Table 3. Excited Singlet-State Lifetimes (τ), Fluorescence Quantum Yields (Φ_f), and Radiative (k_r) and Nonradiative (k_nr) Rate Constants of DMASBT in BSA (150 μM)–12-8-12 Systems^a.

[12-8-12] (mM)	τ1 (ns) (a1)	τ2 (ns) (a2)	χ2	⟨τ⟩ (ns)	Φf	kr (s–1) × 10–6	knr (ns–1) × 10–9
0	0.29 (0.54)	1.09 (0.46)	1.05	0.66	0.0013	1.98	1.52
0.5	0.47 (0.58)	1.38 (0.42)	1.08	0.85	0.0017	2.00	1.17
2.5	0.32 (0.71)	1.02 (0.29)	1.17	0.52	0.0011	2.10	1.91
3.5	0.35 (0.77)	0.98 (0.23)	1.16	0.50	0.0016	3.23	2.02
5	0.22 (0.71)	0.67 (0.29)	1.05	0.35	0.0011	3.13	2.85
10	0.18 (0.75)	0.52 (0.25)	1.08	0.27	0.0009	3.39	3.77
20	0.18 (0.76)	0.49 (0.24)	1.05	0.25	0.0006	2.35	3.93

^aλ_ex = 375 nm, λ_em = 510 nm, [DMASBT] = 10 μM.

Table 2. Excited Singlet-State Lifetimes (τ), Fluorescence Quantum Yields (Φ_f), and Radiative (k_r) and Nonradiative (k_nr) Rate Constants of DMASBT in BSA (150 μM)–12-4(OH)-12 Systems^a.

[12-4(OH)-12] (mM)	τ1 (ns) (a1)	τ2 (ns) (a2)	χ2	⟨<τ⟩ (ns)	Φf	kr (s–1) × 10–6	knr (s–1) × 10–9
0	0.29 (0.54)	1.09 (0.46)	1.05	0.66	0.0013	1.98	1.52
0.5	0.36 (0.51)	1.24 (0.49)	1.14	0.79	0.0011	1.47	1.26
0.7	0.32 (0.55)	1.21 (0.45)	1.12	0.72	0.0018	2.54	1.39
4.5	0.32 (0.76)	0.97 (0.24)	1.07	0.48	0.0007	1.47	2.10
5	0.23 (0.73)	0.80 (0.27)	1.04	0.38	0.0010	2.60	2.60
10	0.18 (0.75)	0.62 (0.25)	1.01	0.29	0.0008	2.76	3.45
20	0.14 (0.71)	0.46 (0.29)	1.07	0.23	0.0010	4.30	4.29

^aλ_ex = 375 nm, λ_em = 510 nm, [DMASBT] = 10 μM.

The radiative and nonradiative rate constants were calculated using eqs 2 and 3, respectively

2

3

where τ, Φ_f, and k_r and k_nr denote singlet state lifetime, fluorescence quantum yield, and radiative and nonradiative rate constants, respectively. k_r and k_nr data obtained are also given in Tables 1–3. The data show that in native BSA as well as in BSA–surfactant systems, the fast component has a major contribution than slow components toward the decay. As compared to native BSA, the lifetimes of both fast and slow components in BSA–surfactant aggregates with 0.5 mM of each of the three surfactants increase. This could be because of specific binding of surfactants making the environment of DMASBT more nonpolar. Beyond this concentration, the lifetimes of both the components become shorter. There is a trend of increase in contribution from the fast component and decrease in contribution from the slow component with increasing concentration of a surfactant. As a result, the average lifetimes become shorter with increasing concentration of a surfactant. The data also show that the nonradiative rate constant increases with increasing concentration of a surfactant. All these results are in support of the fact that the average microenvironment around DMASBT in BSA–surfactant aggregates is more polar as compared to that of native BSA. The expulsion of DMASBT molecules from the hydrophobic environment of protein supports our abovementioned observation of decrease in fluorescence intensity with increasing concentration of surfactant at their higher concentration range.

2.4. Fluorescence (Förster’s) Resonance Energy Transfer

In the current work, we have demonstrated the binding interactions of all three gemini surfactants with BSA by carrying out a FRET study. FRET is an electrodynamic phenomenon involving radiationless energy transfer taking place between two dye molecules. It is applicable in different innovative fields of research.⁵⁷⁻⁵⁹ Energy is transferred from a donor to an acceptor through dipole–dipole interaction without the emission of a photon. As a result, the donor molecule’s fluorescence gets quenched, and the acceptor molecule becomes excited with increase in fluorescence intensity.⁶⁰ As per Förster’s theory, the main factors that control the FRET efficiency are as follows: (i) the magnitude of overlap between the donor fluorescence and the acceptor absorption, (ii) the distance between the donor and the acceptor (acceptable range is 0.002–0.01 μm),⁶¹ and (iii) the correct manner in which the transition dipoles of the donor and the acceptor molecules remain oriented.⁵⁰ Trp-134 and Trp-213 in BSA may behave as donors in energy transfer to DMASBT (acceptor). The BSA was excited at 280 nm. According to literature reports,^62,63 at this excitation, the fluorescence often arises mostly from the Trp moieties. The spectral overlap occurring between the emission spectrum of the donor (BSA) and the absorption spectrum of the acceptor (DMASBT) is displayed in Figure 5. This is the main basis for the resonance energy transfer.

Figure 5.

Overlap of emission spectrum of BSA with absorption spectrum of DMASBT in HEPES buffer; [BSA] = 150 μM and [DMASBT] = 10 μM.

In order to demonstrate the FRET in various systems, many experiments have been conducted by enhancing the DMASBT concentration at a fixed concentration of BSA and each of three gemini surfactants. These experiments have also been done with various different concentrations of gemini surfactants in the entire range of the binding isotherm. Figure 6 represents the gradual changes in the fluorescence spectra of the acceptor (DMASBT, peak maximum at ∼472 nm) and the donor (Trp in BSA) through the isoemissive point at ∼428 nm with the variation of concentration of DMASBT. The concentration of each of the three gemini surfactants present in solution was 5 mM (see Figure S3 in the Supporting Information for spectra in the presence of 10 mM surfactants). Figure 7 represents the changes in the fluorescence intensities of the donor and the acceptor for a noncooperative binding zone of surfactants. This behavior is a manifestation of the FRET between BSA and DMASBT.^64,65 It is pertinent to note from Figure 6 that the fluorescence band corresponding to Trp residues of BSA gets blue-shifted as the concentration of DMASBT is increased. It should be noted that whereas the fluorescence peak maximum of Trp residues in free BSA is found at 352 nm, the same in the presence of 10 μM of DMASBT is located at 348 nm. It depicts that DMASBT molecules provide a hydrophobic microenvironment around the Trp residues of BSA.⁶⁶ It also indicates that the DMASBT molecule is present closer to the Trp-213 residue in comparison to the Trp-134, conceptualizing that the environment of Trp-134 is quite hydrophilic.⁶⁷ These results also indicate that the probable donor–acceptor couple is Trp-213–DMASBT.

Figure 6.

Fluorescence emission spectra of BSA (150 μM) in the presence of 5 mM (a) 12-4-12, (b) 12-4(OH)-12, and (c) 12-8-12 as a function of DMASBT concentration; [DMASBT] = 0, 2.5, 5, 7, 10, 15, 20, and 25 μM, λ_ex = 280 nm.

Figure 7.

Decrement in fluorescence intensity of donor (Trp in BSA) with concomitant enhancement in fluorescence intensity of acceptor (DMASBT) with increasing concentration of DMASBT in the FRET process in the BSA–gemini system with 5 mM gemini surfactant and 150 μM BSA.

For the quantitative analysis of the FRET phenomenon, the Fürster’s distance (R_o), the actual distance (r) between DMASBT and Trp residues of BSA, and the efficiency of energy transfer (E_T) for native BSA, noncooperative, cooperative, and saturated binding regions of BSA–gemini surfactant (all three) systems have been calculated.^56,68 The details of calculation of all these parameters are given as Note 2 in the Supporting Information and the values are tabulated in Table 4. The R_o value is found out to be 2.45 nm for native BSA. This calculated R_o value complies with the literature report.⁵⁶ The E_T value for native BSA is calculated to be 0.28. The distance r is 2.87 nm for native BSA. The data clearly indicate that DMASBT resides in close proximity to the Trp moiety. The R_o values calculated for different concentrations of gemini surfactants are in the range of 2.35–2.54 nm, which is in good agreement with the range (2.0–9.0 nm) for the occurrence of FRET.⁵⁰Figure 8a,b shows the changes in E_T and r, respectively, with the variation in the concentration of gemini surfactants in the entire range of the binding isotherm. The trends are well corroborated with the changes in fluorescence intensity and anisotropy as discussed above. One can see from these figures that while there is a specific binding, the E_T increases with a sharp decrease in r. We know that the compactness of BSA increases in this region. Therefore, the distance between the donor and the acceptor decreases, causing an increase in energy transfer efficiency. In the noncooperative binding region, the energy transfer efficiency decreases with a change in the rate of decrease in r. In case of 12-4-12, in fact there is an increase in the value of r. In the cooperative binding region, the energy transfer efficiency again increases with a concomitant decrease in donor–acceptor distance. Strikingly, FRET takes place between BSA and DMASBT even in the uncoiled state of BSA. It symbolizes that DMASBT is acting as a surface probe.⁴⁹ It resides at the surface of micelles and is situated at a proper distance from Trp moiety for resonance energy transfer to take place with the wrapping up of the protein chains around the micelles. It can be seen from Figure 8a that the energy transfer efficiency is minimum in the absence of any gemini surfactant. In this case, Trp moiety is located quite far away from the DMASBT molecule.

Table 4. FRET Parameters Calculated for the BSA (150 μM)–DMASBT (10 μM) System with Varying Concentrations of Gemini Surfactants (12-4-12, 12-8-12, and 12-4(OH)-12).

surfactant	concentration (mM)	J (λ) × 10–16	Ro (nm)	r (nm)	ET
no surfactant	0	1.05	2.45	2.87	0.28
12-4-12	0.5	1.22	2.41	2.15	0.66
	2.5		2.72	2.28	0.74
	3.5		2.29	1.98	0.71
	5.0		2.26	1.96	0.70
	10.0		2.25	1.87	0.75
	20.0		2.25	1.83	0.77
	30.0		2.23	1.82	0.78
12-8-12	0.5	1.23	2.62	2.25	0.71
	2.5		2.44	2.01	0.76
	3.5		2.32	1.93	0.74
	5.0		2.30	1.92	0.75
	10.0		2.31	1.86	0.78
	20.0		2.33	1.81	0.81
	30.0		2.35	1.82	0.82
12-4(OH)-12	0.5	1.21	2.44	2.22	0.64
	0.7		2.63	2.21	0.74
	4.5		2.28	2.01	0.68
	5.0		2.27	2.017	0.67
	10.0		2.26	1.93	0.72
	20.0		2.27	1.88	0.76
	30.0		2.31	1.856	0.79

Figure 8.

(a) Plot of energy transfer efficiency (E_T) vs gemini surfactant concentrations. (b) Plot of actual distance (r) between Trp residue of BSA and DMASBT vs gemini surfactant concentrations; [BSA] = 150 μM, [DMASBT] = 10 μM.

The effect of the alkyl spacer chain length of gemini surfactants on the binding interactions with BSA can be explained based on energy transfer data. More or less, the increasing order of energy transfer efficiency is 12-8-12 > 12-4-12 > 12-4(OH)-12. It depicts that the energy transfer efficiency increases with increasing hydrophobicity of the spacer group. Micelles formed by gemini surfactant molecules with a comparatively more hydrophobic spacer group can bind with the protein effectively. DMASBT molecules solubilized in these micelles can approach Trp-213 more closely. One can also see from Figure 8a that the energy transfer efficiency in the cooperative binding region increases sharply up to ∼10 mM concentration of each of the three gemini surfactants. However, the rate of increase in energy transfer efficiency decreases beyond this concentration range. This might indicate that up to 10 mM concentration of surfactant, micelles are formed along the protein chain; beyond that, the aggregation number of micelles only increases. Thus, the binding interactions of gemini surfactants with BSA have been successfully probed by the FRET study.

2.5. SANS Study

The formation of structures of BSA–surfactant complexes when these two components interact with each other has been studied using SANS technique. In SANS, one usually measures the differential scattering cross section as a function of wave vector transfer [Q = (4π sin θ)/λ], which can be given as below (eq 4)⁶⁹

4

where φ is the volume fraction of the scatterer, V is the volume of the individual scatterer, and ρ_p and ρ_s are scattering length densities of particles and solvent, respectively. P(Q) is the intraparticle structure factor which decides the geometry of the particles⁶⁹ and S(Q) is the interparticle structure factor which originates from the interparticle interactions⁷⁰ (see Note 3 in the Supporting Information for details).

The SANS data of the pure BSA solution have been analyzed using an oblate ellipsoidal model, and BSA molecules are found to have semiminor and semimajor axes of 13.0 and 42.0 Å, respectively, consistent with the values reported by others.⁷⁰ On the other hand, data of pure micelles are fitted by P(Q) of prolate ellipsoidal, whereas S(Q) is calculated by utilizing the Hayter Penfold method for charged spheres interacting via screened Coulomb interactions.⁷¹ As representative values, the semiminor and semimajor axes of micelles of 20 mM 12-4-12 are found to be 12.4 and 28.5 Å, respectively. The SANS data corresponding to the interaction of the two components are shown in Figure 9. These measurements were performed at a fixed concentration of BSA (150 μM) and varying concentrations of gemini surfactants. The data are collected only for a higher concentration range of surfactants as the scattering signals are very low at a lower concentration range. It can be observed from Figure 9 that the data for surfactant concentrations up to 20 mM show a linearity in the intermediate Q range, which is characteristic of the formation of the necklace and bead structures. These necklace and bead complexes are usually characterized by fractal structures where the fractal dimensions represent the degree of the formation of the complexes.⁷⁰ The cutoffs of the linear range of the data at low and high Q values are related to the formation of the extent of the complex and the size of the individual micelles in the complex, respectively. The slope of the scattering data evaluated on the log–log scale is useful for the determination of the value of the fractal dimension, D (Note 3 in the Supporting Information), of the complex. This feature of the linearity in the scattering profile can be seen for all the three gemini surfactants and the slope of the linearity changes with surfactant concentration, suggesting the changes in the fractal dimension with concentration, in all three cases. The fitted parameters are listed in Tables 5 and 6. It can be seen from the tables that correlation length increases whereas the fractal dimension decreases with increasing surfactant concentration, indicating the formation of a larger and more open structure. At a higher concentration (50 mM) of surfactant, the linearity in the data disappears and one can observe the emergence of a correlation peak, which is a signature of interacting micelles suggesting the existence of excessive individual micelles. Therefore, these data have been analyzed by combining the scattering contributions of the fractal complexes of protein and surfactant and free individual micelles.

Figure 9.

SANS plot for systems containing BSA (150 μM), DMASBT (10 μM), plus varying concentrations of gemini surfactants, (a) 12-4-12, (b) 12-4(OH)-12, and (c) 12-8-12 (5, 10, 20, 50 mM).

Table 5. Fitted Parameters of SANS Analysis of the Protein–Surfactant Complex for 150 μM BSA with Gemini Surfactant of Various Concentrations (5–20 mM)^a.

150 μM BSA (1 wt %) + x (mM) 12-m-12	building block (Å) radius	correlation length (Å)	fractal dimension	model
[12-4-12](mM)
5	14.4	18.5	2.8	necklace and bead model
10	14.4	21.8	2.1
20	15.0	27.3	1.2
[12-4(OH)-12](mM)
5	14.0	21.8	2.8
10	14.0	22.7	2.2
20	14.7	24.0	1.7
[12-8-12](mM)
5	14.0	19.1	2.8
10	14.1	23.8	2.1
20	14.4	25.0	1.5

^aData are fitted for prolate ellipsoidal shape of the protein macromolecule.

Table 6. Fitted Parameters of SANS Analysis (Semimajor, Semiminor Axes, Charge, and Model) of the Protein–Surfactant Complex for 150 μM BSA with 50 mM Gemini Surfactants.

150 μM BSA + 50 mM surfactant	semimajor axis (Å)	semiminor axis (Å)	charge	model
12-4-12	35.7	16.5	7.4	necklace and bead model + ellipsoidal micelles
12-4(OH)-12	50.2	16.5	6.5
12-8-12	27.9	15.0	7.3

2.6. Dynamic Light Scattering

The details about the dimensional changes of BSA protein upon addition of gemini surfactant in aqueous buffer solution have been obtained using the DLS technique. In DLS, the fluctuations in the intensity of scattered light are measured as a function of time.⁷² These fluctuations happened because of the Brownian motion of the particles randomly colliding with each other. Small particles undergo swift diffusion leading to fast fluctuations, whereas large particles and aggregates yield comparatively slow fluctuations. The rate of the fluctuations may be calculated using the method of autocorrelation analysis. The analysis of intensity fluctuations helps in the determination of the diffusion constant (D), which can then be transformed to an effective hydrodynamic size using the Stoke–Einstein relation (eq 5)

5

where D is the translational diffusion coefficient, η is the viscosity coefficient, k_B is Boltzmann’s constant, and R_h is the hydrodynamic radius.

The sizes of particles such as BSA–surfactant aggregates at a higher concentration range of all three surfactants have been measured. DLS plots for systems with 150 μM BSA (1 wt % BSA) plus various concentrations of surfactants are shown in Figure 10. One can observe the systematic broadening of the intensity autocorrelation function with increasing surfactant concentration in Figure 10. The analysis clearly indicates that all the data, fitted to the single diffusion coefficient for the protein–surfactant complex, are independent of the surfactant concentration. The effective hydrodynamic diameters obtained from the data are tabulated in Table 7. There is an increase in size with an increase in surfactant concentration, which is due to progressive uncoiling of the protein chain. The diffusion coefficient (being inversely proportional to the hydrodynamic radius) diminishes with the increase of surfactant in the complex. This might happen either when the surfactant molecules bind as individuals or when the micellelike clusters are formed with the protein. There remains a significant difference between intensity autocorrelation function for pure BSA and the BSA–gemini complexes. It is noteworthy that by and large the increasing order of hydrodynamic size obtained for the three gemini surfactants in the concentration range 5–20 mM is as follows: 12-4(OH)-12 < 12-4-12 < 12-8-12, which indicates the formation of larger aggregates of protein with the gemini surfactant having a spacer group of greater hydrophobicity.

Figure 10.

DLS plot for 150 μM BSA (1 wt % BSA) + 10 μM DMASBT with different concentrations of gemini surfactants; (a) 12-4-12, (b) 12-4(OH)-12, and (c) 12-8-12 at their various concentrations; [12-m-12] = 0, 5, 10, 20, 50 mM.

Table 7. DLS Data for 150 μM BSA (1 wt % BSA) + 10 μM DMASBT with Varying Gemini Surfactant Concentrations.

	effective hydrodynamic diameter (nm)
150 μM BSA (1 wt %) + x (mM) 12-m-12	12-4-12	12-4(OH)-12	12-8-12
0	8.3	8.3	8.3
5	15.8	12.2
10	22.0	16.0	20.0
20	24.1		29.0
50	28.1	40.1	33.1

3. Conclusions

The binding interactions of a TICT fluorescent probe, DMASBT with the native protein, BSA and different BSA–gemini surfactant aggregates have been studied. The changes in fluorescence properties of DMASBT with changing the microenvironment have been explored to study the binding interactions of gemini surfactants with varying spacer groups at different binding isotherms. The fluorescence lifetimes and radiative and nonradiative rate constants of DMASBT have been measured at different regions of binding isotherms. The lifetimes of both the decay components in a biexponential decay become shorter with a trend of increase in contribution from the fast component and decrease in contribution from the slow component with increasing concentration of a surfactant. Thus, the average lifetimes become shorter with increasing concentration of a surfactant. The nonradiative rate constant increases with increasing concentration of a surfactant. The average microenvironment around DMASBT in BSA–surfactant aggregates is more polar as compared to that in native BSA. BSA–gemini surfactants’ binding interactions at various binding isotherms have been probed by the FRET phenomenon between Trp-213 of BSA as a donor and DMASBT as a potential acceptor. The energy transfer efficiency is minimum in native protein. It increases in the presence of a low concentration of gemini surfactant as the donor and acceptor moieties come closer to each other because of specific binding. At a higher concentration of surfactant, energy transfer efficiency decreases as a result of unfolding of protein. The increasing order of energy transfer efficiency at a higher concentration range of surfactants is 12-8-12 > 12-4-12 > 12-4(OH)-12. Micelles of gemini surfactant molecules with a comparatively more hydrophobic spacer group can bind with the hydrophobic segments of the protein more strongly. DMASBT molecules solubilized in these micelles can approach Trp-213 more closely. SANS and DLS measurements have been carried out to study the structural changes in the protein, BSA, on increasing concentration of the gemini surfactants. SANS data well assist the necklace and bead model for the formation of micelles along the protein chain. The fractal dimension reduces, whereas the overall size of the protein–surfactant complexes enhances with increasing concentration of surfactant. At a higher concentration range, 12-8-12 and 12-4-12 have a slightly smaller fractal dimension and larger correlation length as compared to 12-4(OH)-12, indicating formation of better micellar aggregates along the unfolded protein chain in the former two cases. It could be due to the greater hydrophobicity of the spacer groups of the former two surfactants as compared to the latter one. DLS data show that by and large the increasing order of hydrodynamic diameter of the BSA–surfactant aggregates for the three gemini surfactants in the concentration range 5–20 mM is as follows: 12-4(OH)-12 < 12-4-12 < 12-8-12. It indicates the formation of larger aggregates of protein with the gemini surfactant having a spacer group of greater hydrophobicity.

4. Experimental Section

4.1. Materials

Gemini surfactants, 12-4-12, 12-8-12 and 12-4(OH)-12, were synthesized following a reported method.^73,74 The synthesized compounds were then recrystallized quite a number of times using a mixture of methanol and ethyl acetate, and then their structures were confirmed by FT-IR and ¹H NMR data.⁴⁶ BSA was purchased from Sigma Chemical Company and used as received. The concentration of the protein, BSA, for all measurements was kept fixed at 1 wt % (150 μM). DMASBT was obtained from Aldrich Chemical Company (WI, USA). Its recrystallization process and purity inspection have been reported earlier.⁷⁵ HEPES was obtained from SRL, India. Ludox (aqueous solution) was procured from Aldrich Chemical Co. In time-correlated single-photon counting (TCSPC) measurements, the Ludox solution was used for recording the lamp profile.

4.2. Methods

Freshly prepared buffer solution was used for sample preparation. NaOH and H₂SO₄, used for pH adjustments of the buffer solution, were purchased from Merck, India. HEPES buffer solution (0.05 M) was prepared using Milli-Q water with an adjustment of pH 7.4. HEPES buffer was used to prepare BSA solutions. A 0.5 mM DMASBT solution was prepared as stock in pure methanol in order to record UV–vis absorption and fluorescence spectra. In order to prepare solutions of varying concentrations of BSA with a particular concentration of DMASBT, 0.02 mL of a methanol solution of DMASBT was mixed with the required amount of an aqueous solution of BSA. The final volume of this solution was adjusted to 2 mL. For the FRET study, 120 mM of solution of each of the three gemini surfactants was prepared as stock in HEPES buffer. Then, the concentration of DMASBT was consequently increased by adding required volumes from the stock solution to the buffer solution of BSA and surfactant, and the final volume was adjusted to 2 mL. Solutions were prepared for different concentrations of surfactants. The fluorescence quantum yields were calculated, with respect to that of quinine sulfate in 0.1 N H₂SO₄, as 0.55. A JASCO (model V-650) UV–vis spectrophotometer was used for absorption measurements and FluoroMax-4 (Horiba Jobin Yvon) spectrofluorimeter was used for steady-state emission measurements. Every fluorescence spectrum was corrected for instrument sensitivity. The slit width for both the slits was set at 3 nm for fluorescence measurements. The same spectrofluorimeter with a polarizer attachment was used for fluorescence anisotropy measurements. Its details can be found elsewhere.^52,58 Fluorescence lifetimes were measured from time-resolved intensity decay following the method of TCSPC using a FluoroCube-01-NL (Horiba Jobin Yvon) picosecond instrument. The light source was a picosecond diode laser (NanoLED 375L, IBH, U.K.) of wavelength 375 nm. Fluorescence decays were recorded at a given wavelength at a magic angle of 54.7° with a vertical excitation beam using a TBX photon detection module (TBX-07C). The typical instrument response function after deconvolution using a liquid scatterer was ∼165 ps for the 375 nm laser. The fluorescence decays were analyzed by IBH DAS6 software. In order to measure the lifetime of DMASBT in protein and protein–gemini surfactant solutions, the analysis of emission decay curves was performed with a biexponential fitting using an iterative program supported by IBH. The quality of fitting was confirmed by χ² value as well as an examination of the fitted function residuals to the data. All measurements were carried out at room temperature (298.15 5 ± 1 K).

Samples for SANS experiments were prepared by dissolving a known amount of BSA and gemini surfactants in a buffer solution which is prepared in D₂O. Use of D₂O as solvent instead of H₂O provides better contrast and low incoherent background in cases when we are dealing with hydrogenous components in neutron scattering experiments. The experiments were performed using the SANS diffractometer at the Dhruva Reactor, Bhabha Atomic Research Centre, Trombay, Mumbai, India.⁷⁶ The mean wavelength of the incident neutron beam was 5.2 Å, with a wavelength resolution of approximately 10%. The scattered neutrons were detected using a one-dimensional ³He detector. The samples were kept in a quartz cell of thickness 2 mm and the temperature of the experiments was kept fixed at 303.15 ± 1 K throughout the measurements. The measured SANS data were corrected and then normalized to a cross-sectional unit using standard procedures.

DLS is a suitable technique complementing SANS, which gives necessary size description through the measurement of the diffusion coefficient of the complex and the Brownian motion in the system. DLS measurements were carried out using an SZ-100 particle size analyzer (Horiba, Japan) having a 10 mW-diode-pumped solid-state laser having a wavelength of 532 nm. The temperature throughout all the measurements was uniformly kept at 303.15 ± 1 K.

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsomega.8b01471.

• Benesi–Hildebrand plot of 1/(F – F₀) versus 1/[BSA] for BSA–DMASBT binding (for [BSA] = 10, 20, 30, 40, 50, 60, 70, 120, 150, 200 μM); plot representing steady-state fluorescence anisotropy of DMASBT with increasing concentration of BSA (λ_ex = 370 nm, λ_em = 475 nm); and fluorescence emission spectra of BSA (150 μM) in the presence of 10 mM (a) 12-4-12, (b) 12-4(OH)-12, and (c) 12-8-12 with varying concentrations of DMASBT; [DMASBT] = 0, 2.5, 5, 7, 10, 15, 20, and 25 μM, λ_ex = 280 nm (PDF)

Author Present Address

^§ Department of Materials and Environmental Chemistry, Arrhenius Laboratory, Stockholm University, SE-106 9, Stockholm, Sweden, SE-106 91.

The authors declare no competing financial interest.