Interaction of Bovine Serum Albumin (BSA) with ... - ACS Publications

Jul 7, 2014 - Honorata M. Ropiak , Peter Lachmann , Aina Ramsay , Rebecca J. Green , Irene Mueller-Harvey , Jamshidkhan Chamani. PLOS ONE 2017 12 ...
0 downloads 0 Views 591KB Size
Article pubs.acs.org/JPCB

Interaction of Bovine Serum Albumin (BSA) with Novel Gemini Surfactants Studied by Synchrotron Radiation Scattering (SR-SAXS), Circular Dichroism (CD), and Nuclear Magnetic Resonance (NMR) W. Gospodarczyk, K. Szutkowski,* and M. Kozak* Department of Macromolecular Physics, Faculty of Physics, Adam Mickiewicz University, ul. Umultowska 85, PL61614 Poznań, Poland ABSTRACT: The interaction of three dicationic (gemini) surfactants3,3′-[1,6-(2,5-dioxahexane)]bis(1-dodecylimidazolium) chloride (oxyC2), 3,3′-[1,16-(2,15-dioxahexadecane)]bis(1-dodecylimidazolium) chloride (oxyC12), and 1,4-bis(butane)imidazole-1-yl-3-dodecylimidazolium chloride (C4)with bovine serum albumin (BSA) has been studied by the use of small-angle X-ray scattering (SAXS), circular dichroism (CD), and 1H nuclear magnetic resonance diffusometry. The results of CD studies show that the conformation of BSA was changed dramatically in the presence of all studied surfactants. The greater decrease (from 56 to 24%) in the α-helical structure of BSA was observed for oxyC2 surfactant. The radii of gyration estimated from SAXS data varied between 3 and 26 nm for the BSA/oxyC2 and BSA/oxyC12 systems. The hydrodynamic radius of the BSA/surfactant system estimated from NMR diffusometry varies between 5 and 11 nm for BSA/ oxyC2 and 5 and 8 nm for BSA/oxyC12.



INTRODUCTION Understanding and controlling the interactions between proteins and surfactants may enable new drug delivery systems for improved medical treatments.1−3 In addition, consumer applications related to the cosmetics and food preservation industries would benefit from new protein−surfactant systems.4−6 Surfactants influence protein aggregation, so a detailed experimental understanding of the character and type of interactions between proteins and surfactants as a function of concentration, pH, temperature, and chemical structures is required in order to advance these fields.7−10 For example, amyloidogenic and related proteins are known to aggregate, and it is believed that these aggregated “plaques” play a key role in diseases like Alzheimer’s disease.11,12 Gemini surfactants, which are dicationic or dimeric surfactants, exhibit unique physicochemical properties relative to their single alkyl chain counterparts, and their higher surface activity and concomitant lower critical micelle concentration (CMC) make them attractive for development.13,14 Rational selection of the headgroup, spacer group, and hydrophobic chain length should enable selectivity with respect to physicochemical, toxicological, and morphological properties.14−18 It is believed that the lyotropic structures of gemini surfactants are structurally biocompatible with cellular membranes.5 This current contribution focuses on the influence of gemini surfactants with varying alkyl spacer groups on bovine serum albumin (BSA) aggregation.2 The surfactant−protein interactions and structural changes were studied with the use of synchrotron radiation small angle X-ray scattering (SR-SAXS), circular dichroism (CD), and 1H nuclear magnetic resonance diffusometry. © XXXX American Chemical Society

Bovine serum albumin is a transport protein with a mass of around 60 kDa. A single BSA molecule has a globular shape with a corresponding Stokes radius of around 3.5 nm. In general, serum albumins carry steroids and fatty acids and they are known to bind and interact with many types of physiological metabolites and various organic/inorganic ligands; therefore, BSA is often used as a model protein molecule.19 Dimeric surfactants were already proposed as substances preventing aggregation of amyloidogenic proteins. Cao et al. have observed that dicationic surfactant micelles of hexamethylene-1,6-bis(dodecyl dimethylammonium bromide) (C12C6C12Br2) effectively delay the formation and effectively disassemble Aβ(1−40) fibrils in vitro.11,12 Furthermore, the influence of another surfactant on BSA conformation, namely, hexamethylene-1,3-bis(tetradecyldimethylammonium bromide), was characterized by synchronous fluorescence spectroscopy and 3D fluorescence spectral techniques, where the conformation of BSA was altered dramatically by binding surfactant molecules to tryptophan and tyrosine residues.5 Also, monomeric cationic surfactants were studied where the influence (induction of protein denaturation) of cethyltrimethylammonium bromide (CTAB)a surfactant with antimicrobial propertieson BSA structure was studied.20 Received: May 14, 2014 Revised: July 3, 2014

A

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B



Article

EXPERIMENTAL SECTION Materials. The following gemini surfactants were used in the study: 3,3′-[1,6-(2,5-dioxahexane)]bis(1-dodecylimidazolium) chloride (oxyC2), 3,3′-[1,16-(2,15-dioxahexadecane)]bis(1-dodecylimidazolium) chloride (oxyC12), and 1,4-bis(butane)imidazole-1-yl-3-dodecylimidazolium chloride (C4). The surfactants were synthesized according to the earlier described procedure.21 Their chemical structures are shown in Figure 1. Bovine serum albumin (BSA), HEPES, and sodium dihydrogen phosphate dihydrate were purchased from Aldrich.

Table 1. Structure Parameters for BSA Obtained from CD Spectraa surfactant

csurf (mM)

csurf:cBSA

α-helix (%)

β-sheet (%)

turns (%)

others (%)

oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC12 oxyC12 oxyC12 oxyC12 oxyC12 C4 C4 C4 C4 C4 C4 C4 C4

0 0.001 0.006 0.013 0.063 0.125 0.250 0.525 1.050 0 0.001 0.013 0.043 0.063 0 0.001 0.013 0.048 0.095 0.190 0.380 0.760

0:1 0.5:1 2.8:1 6.1:1 29:1 58:1 117:1 245:1 491:1 0:1 0.5:1 6.4:1 21:1 31:1 0:1 0.5:1 6.2:1 23:1 45:1 91:1 182:1 364:1

54 54 54 50 50 24 31 42 40 55 56 57 51 35 56 57 54 50 42 37 41 41

9 9 9 12 10 27 20 12 14 7 8 7 9 21 7 7 9 11 15 18 13 12

11 11 11 12 11 13 12 14 13 11 11 11 12 12 11 11 11 11 12 13 13 14

26 26 26 27 28 37 36 31 33 26 26 25 28 32 26 25 27 28 31 32 33 32

a

Concentrations of surfactant and BSA are denoted as csurf and cBSA, respectively.

Table 2. Gyration Radii Obtained from SR-SAXS for 4 mg/ mL BSA for Different Concentrations of oxyC2, oxyC12, and C4 Surfactantsa

Figure 1. Chemical formulas of gemini surfactants in the study: (a) 3,3′-[1,6-(2,5-dioxahexane)]bis(1-dodecylimidazolium) chloride (oxyC2); (b) 3,3′-[1,16-(2,15-dioxahexadecane)]bis(1-dodecylimidazolium) chloride (oxyC12); (c) 1,4-bis(butane)imidazole-1-yl-3dodecylimidazolium chloride (C4).

As different concentrations of the protein were needed for different methods, the molar ratio of protein to surfactant (see Tables 1 and 2) was first and foremost retained. The BSA concentration for the CD experiments was kept constant at 0.15 mg/mL, and a 25 mM solution of sodium phosphate (pH 7.7) was used as a buffer. The surfactant concentration was varied from 1 μM to 1.05 μM according to the data shown in Table 1. For SR-SAXS experiments, the BSA concentration was higher and kept at 4 mg/mL and the pH was 7.7. The concentration of gemini surfactants was also higher (from 0.05 to 20 mM) for SR-SAXS experiments in order to maintain the BSA/gemini ratio. For NMRD measurements, the BSA concentration was set at 8 mg/mL and 50 mM HEPES in D2O (pH 7.5) was used as a buffer. Circular Dichroism (CD). Circular dichroism measurements were carried out using a Jasco J-815 CD spectrometer. The CD spectra were collected at room temperature over the range 182−260 nm using a quartz cuvette with an optical path of 1 mm. The scanning rate was 50 nm/min, and the spectral width was 0.5 nm. All measurements were performed in a nitrogen atmosphere at a nitrogen flow rate of 20 mL/min. For each sample, five spectra were accumulated and averaged to get the final CD spectrum. The CD data were deconvoluted on the Dichroweb server.22 The CDSSTR program module23 was

surfactant

csurf (mM)

csurf:cBSA

Rg (nm)

oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC2 oxyC12 oxyC12 oxyC12 oxyC12 oxyC12 oxyC12 oxyC12 oxyC12 oxyC12 oxyC12 C4 C4 C4 C4 C4 C4

0 0.05 0.10 0.20 0.50 1.00 2.00 5.00 10.00 20.00 0 0.05 0.10 0.20 0.50 1.00 2.00 5.00 10.00 20.00 0 0.5 1.0 1.9 3.8 7.6

0:1 0.8:1 1.7:1 3.3:1 8.3:1 17:1 33:1 83:1 167:1 333:1 0:1 0.8:1 1.7:1 3.3:1 8.3:1 17:1 33:1 83:1 167:1 333:1 0:1 9:1 18:1 35:1 69:1 138:1

3.2 3.2 3.3 3.6 14.9 17.0 ----7.8 7.3 3.2 3.2 3.3 8.6 14.2 25.7 --------3.0 ----4.8 4.8 5.3

The symbol “---” indicates that Rg could not be determined (SAXS polydispersity region).

a

B

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

low as 8 mg/mL. The concentration of ionic HEPES concentration was set to 50 mM for all the samples, and the sample concentrations were kept within an accuracy of 2%. The pulse sequence for NMR diffusometry is based on the spin−echo experiment with pulses of a magnetic field gradient which tags positions of nuclear spins during time called the diffusion time Δ. The Stejskal−Tanner equation describes the diffusion decay of the spin−echo M(g) as follows:

chosen for computational analysis and the SP175 database as reference.24 Synchrotron Radiation Small Angle X-ray Scattering (SR-SAXS). Synchrotron radiation SAXS measurements were performed on the BM29 Beamline of ESRF (Grenoble)25 using synchrotron radiation (λ = 0.0995 nm). SAXS patterns were collected using a photon counting Pilatus 1 M pixel detector (253 × 288 mm2) at a sample to detector distance of 2867 mm within the scattering vector range 0.08 nm−1 > s > 3.8 nm−1 (where s = 4π sin θ/λ and 2θ is the scattering angle). The scattering vector range was calibrated using the diffraction patterns of silver behenate.26 Solutions of a known concentration (∼3 mg/mL) of xylose/glucose isomerase from Streptomyces rubiginosus were used as additional references for the molecular weight calibration.27 The SR-SAXS data were collected for a series of surfactant/BSA solutions in 50 mM sodium phosphate (pH 7.7) containing the surfactant at increasing concentrations (0−20 mM). All measurements were performed using a capillary cell (sample volume 20 μL) and automated filling at 15 °C. The data was collected in 10 successive 1 s frames. Only data frames without radiation damage were further processed (integrated and averaged). In addition, the integrated SAXS data was corrected for detector response and normalized to the incident beam intensity. Also, the scattering signal of a buffer was subtracted using the software package PRIMUS.28 Nuclear Magnetic Resonance Diffusometry (NMRD). Translational self-diffusion coefficients D for BSA/gemini surfactant samples were obtained by 1H nuclear magnetic resonance diffusometry. The diffusion data was collected using a Bruker Avance DMX 9.4 T spectrometer using a Diff25 10 mm diffusion probe capable of generating a maximum pulsed magnetic field gradient of 10 T/m. The standard pulsed field gradient stimulated echo (PGSTE) is shown in Figure 2. The

A (g ) =

M (g ) = A exp( −Dγ 2δ 2g 2(Δ − δ /3) M0

(1)

where D is the diffusion coefficient, γ is the magnetogyric ratio for the protons, δ is the duration of the diffusion gradient, g is the gradient amplitude, M0 is the amplitude for g = 0, and Δ is the diffusion time during which the motion of spins is tagged. For non-heterogeneous systems, a simple model involving distribution of diffusion coefficients is introduced A(g ) = AF exp( −Dγ 2δ 2g 2(Δ − δ /3)) + (1 − AF)

∫0



G(Ddist ) exp( −Ddist γ 2δ 2g 2(Δ − δ /3)) dD (2)

where AF is the population of species with a single diffusion coefficient and (1 − AF) is the population of species characterized by a Gaussian distribution of diffusion coefficients: G(Ddist ) =

⎛ 1 ⎡ D − D ⎤2 ⎞ 1 dist exp⎜ − ⎢ ⎥⎦ ⎟ σ 2π σ ⎝ 2⎣ ⎠

(3)

The diffusion coefficient Ddist is directly related to the mean hydrodynamic radius of diffusing species, and accordingly, the particle size distribution is characterized by σ.



RESULTS AND DISCUSSION Three complementary methods were selected to expand the information regarding the interactions between gemini surfactans and BSA: circular dichroism (CD), synchrotron radiation small-angle X-ray scattering (SR-SAXS), and nuclear magnetic resonance diffusometry (NMRD). While CD analysis provides detailed information on the secondary structure of BSA, SR-SAXS gives an average dimension of BSA molecules upon interaction with surfactants. All in all, NMR diffusometry provides important information on guest−host interactions, solvation dynamics, and overall diffusivity of BSA−surfactant complexes in the solution. Although our methods are complementary, different BSA concentrations in the sample were required due to experimental limitations, albeit molar ratios were kept so that comparison was possible. The BSA concentrations were 0.15 mg/mL for the CD sample, 4 mg/mL for SR-SAXS, and 8 mg/mL for NMRD accordingly. Analysis of BSA Secondary Structure. The BSA molecule can be described by using a three-domain structure.31 At pH 7, two domains are negatively chargeddomain I (charge −10) and domain II (charge −8)and one is neutraldomain III.32 Furthermore, the effect of pH on BSA structure is well described by changes in isomeric forms for which BSA undergoes reversible conformational transitions between the following forms: extended form (E-form), fast form (F-form), normal form (N-form), basic form (B-form), and aged form (A-form). Each form is specifically characterized

Figure 2. PGSTE pulse sequence. The following parameters were used during the experiments: diffusion time Δ = 50 ms, gradient duration δ = 2 ms, gradient amplitude g was varied between 0 and 4.5 T/m in 32 steps.

spin echo signal was Fourier transformed to establish the diffusivities of selected compounds.29,30 The following spectral ranges were selected for analysis of the gemini surfactants and BSA diffusion: (1) 0−2 ppm and (2) 7−8 ppm. The initial PGSE data analysis was made with Bruker TopSpin 2.1. The data was Fourier transformed with the exponential line broadening factor 30 Hz, and signal decays for 0−2 and 7−8 ppm regions were exported for further analysis. Subsequently, the analysis of signal decays was carried out with OriginLab Origin 8.5. Initially, a set of calibration experiments was made for samples with and without BSA for oxyC2, oxyC12, and C4 gemini surfactants in 50 mM HEPES/D2O. Accordingly, the minimum detectable concentration of BSA was identified as C

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

suggests that BSA is first transformed to the F-form and then for higher concentrations to the completely unfolded E-form, characterized by very low α-helix content.33,36 The decrease in the α-helix content takes place for concentrations higher than 0.063 mM (molar ratios 30−60:1) for oxyC2, 0.043 mM (molar ratios 20−30:1) for oxyC12, and 0.095 mM (molar ratios 45−91:1) for C4 gemini surfactant accordingly. The observed loss in the α-helix content was accompanied by an increase in the content of β-sheet structures and others. It can be noticed that the drop in α-helix content versus surfactant concentration is more pronounced for both oxyC2 and oxyC12 surfactants. Otherwise, the C4 gemini surfactant requires higher concentrations in order to induce the N−E transition (Table 1). The reason for that are two ether functional groups in the oxyC2 and oxyC12 spacer groups and the dipole moment which contribute to dipole−dipole and dipole−electric interactions with both water and charged BSA molecules, respectively. The dipole moment can be estimated to fall between dipolar moments of 1.14 D (diethyl ether) and 1.3 D (dimethyl ether). One has to note oxyC12 is more hydrophobic than oxyC2 due to the longer and more flexible spacer, which is probably reflected in the extent of structural changes due to the more favorable affinity of the oxyC12 surfactant for the BSA molecule. Furthermore, the mechanism of α-helices unfolding may be related to gradual swelling of the BSA molecule with surfactant molecules, followed by the interactions of hydrophobic sites of the protein with hydrophobic alkyl chains of the surfactant, resulting in subsequent destabilization of the hydrogen bonds within α-helices.34 At very low surfactant concentrations in a solution, the content of α-helices in BSA secondary structure is preserved, albeit at this stage of interactions the tertiary structure of the protein might be disrupted.37 The Radius of Gyration Obtained from SR-SAXS. Experimentally obtained SR-SAXS curves are presented in Figure 4. The Guinier plots log I(s2) from which the radii of gyration were determined are presented in Figure 5. Radii of gyration Rg obtained by SR-SAXS are summarized in Table 2. Upon interactions of oxyC2 with BSA, the protein radius of gyration is slowly increasing for surfactant concentration up to 0.2 mM. For higher concentrations ranging from 0.5 to 1 mM, the BSA radius of gyration is increasing from around 3 nm up to 17 nm. At higher concentrations of oxyC2 of 10 and 20 mM, Rg is decreasing up to 7 nm. According to CD data, the overall content of α-helix is decreased by 14% at 1 mM oxyC2 in the solution. Such a high value of Rg obtained by SR-SAXS confirms that the BSA molecule is swollen with surfactant molecules and reaching a maximum value at concentrations between 1 and 10 mM for oxyC2. The maximum value of the radii of gyration for BSA was observed for oxyC12 surfactant, also at a concentration of 1 mM (Table 2). Although we were unable to obtain CD data at this particular concentration, the decrease in α-helix content of BSA for oxyC12 was even more pronounced for much smaller concentrations of 0.06 mM (almost 20%, see Table 1). For C4 concentrations higher than 1.9 mM, the BSA radius of gyration varied from 4.8 to 5.3 nm, which confirms weak interactions of C4 with BSA at high concentrations, although CD data shows that for smaller concentrations the content of α-helix is also decreasing by approximately 15% at around 0.8 mM. For most samples with gemini surfactant concentrations ranging from 1 to 20 mM, the loss of solution clarity or the

by α-helix content at specific pH. For example, the B−N transition takes place for pH values below 8 and at the same time is characterized by the highest α-helix content (from 48 to 55%). In turn, the F−N transition is characterized by significantly lower solubility and low α-helix content (down to 35%).33 The experimental CD spectra (expressed as the ellipticity θ versus the incident light wavelength) for BSA/oxyC2, BSA/ oxyC12, and BSA/C4 systems are presented in Figure 3. The detailed secondary structure information derived from these curves is shown in Table 1. The results of CD measurements reveal a substantial change in α-helix content as a result of unfolding of BSA from N-form to F- and E-forms due to binding interactions with surfactants.34,35 The content of αhelix of around 50 for 29:1 surfactant-to-BSA ratio (oxyC2)

Figure 3. Ellipticity vs wavelength of the incident light beam for (a) BSA/oxyC2, (b) BSA/oxyC12, and (c) BSA/C4 samples (the insets inside the figure denote concentrations of gemini surfactants in mM). D

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Figure 5. Experimental points and fits of log I(s2) for oxyC2 SR-SAXS data. The radius of gyration is proportional to the slope.

complex acquires positive net charge and regains solubility.37 This change (transition) of optical properties and change of Rg occurs for a surfactant-to-BSA ratio of 17−33:1 (oxy C2 and oxyC12) or even smaller than 9:1 for C4 surfactant. After the transition, SAXS curves reveal the formation of complex structures in solution, presumably micelle-like ones, since many surfactant concentrations are presumably much above the cmc, which is supported by a characteristic “broad peak” (about 150−300:1) detected in the middle of the SR-SAXS plots. This is especially visible in Figure 4c, but it is also present in Figure 4a as well as in Figure 4b. The solution is likely to comprise unfolded protein chains with surfactant micelles attached to them and free surfactant micelles in the solution.37−40 The gyration radius of BSA in this region is very high, compared to the BSA without surfactant and is not further changing distinctly with surfactant concentration, which suggests that the maximum number of surfactant molecules is bound to the BSA molecule. Any higher surfactant concentrations will increase the aggregation number of surfactant micelles without affecting BSA. In similar systemsalkylimidazolium-based ionic liquids with BSAZhu et al. observed that one type of binding between protein and surfactant is caused by electrostatic interaction of the cation with negatively charged sites on BSA molecules.41 Another type of interactions (low affinity binding) is related to the hydrophobic interactions of imidazole rings and the hydrophobic chains of the ionic liquids (or surfactants) with the hydrophobic cavities of the protein molecules.41 Nuclear Magnetic Resonance Diffusion. The DOSY spectra obtained for 8 mg/mL BSA and 10 mM oxyC2 in D2O/ HEPES solution are presented in parts a and b of Figure 6,

Figure 4. SAXS scattering curves as obtained for 4 mg/mL BSA solutions with (a) oxyC2, (b) oxyC12, and (c) C4 gemini surfactants at different concentrations (mM). The intensity log(I) is plotted versus scattering vector s.

appearance of precipitation were observed (“polydispersity region”); therefore, SAXS curves referring to this region have a much lower signal-to-noise ratio (Figure 4). The solutions became clear again at higher surfactant concentrations. The initial growth of Rg values can be attributed to surfactant binding and unfolding of BSA, as processes such as these involve protein expansion and therefore imply an increase of radius of gyration. The latter growth could be due to an aggregation at low values of surfactant-to-protein molar ratios, occurring when the protein−surfactant complex becomes insoluble. Namely, as positively charged gemini molecules attach to BSA, the macromolecule gradually loses its initial net charge, and eventually the complex solubility decreases.37 This property of the BSA−surfactant complex persists until more gemini molecules are attached to the protein. Then, the E

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Figure 6. 2D DOSY spectra as obtained by CONTIN and TopSpin from FT PGSTE data. (a) 8 mg/mL BSA in D2O (5 mM HEPES). The strongest BSA signal intensity is recorded for 0−2 ppm. (b) 10 mM oxyC2 gemini surfactant in D2O (50 mM HEPES). The diffusion of surfactant molecules can be analyzed for two chemical shift regions: (1) 0−2 ppm and (2) 6−8 ppm.

Figure 7. Signal decays for the 0−2 ppm region: (a) oxyC2, (b) oxyC12, and (c) C4 samples with and without BSA. The diffusion coefficients for the 0−2 ppm region refer to the surfactant and BSA diffusion. For concentrations higher than 4−5 mM, the diffusion coefficients are considerably slower as a result of BSA/surfactant aggregation.

respectively. The remaining water signal (HDO) is visible at around 4.8 ppm and log D at around −8.7, while the resonance from HEPES is visible between 2 and 4 ppm and log D around −9.4. The highest intensity NMR resonance lines corresponding to both BSA and surfactant are visible in the ppm range 0− 2 ppm. Therefore, this ppm region is further analyzed for BSA/ gemini surfactant diffusion. The diffusion coefficient obtained from DOSY analysis for 8 mg/mL BSA in D2O at 21 °C is estimated to be DBSA = 3.9 × 10−11 m2/s (Figure 6a). The exponential analysis of Fourier transformed PGSTE data for BSA (eq 1) yields a diffusion coefficient of DBSA = 4.1 × 10−11 m2/s, which is in good agreement with DOSY analysis. The average value of the BSA diffusion coefficient of 4 × 10−11 m2/s corresponds to the hydrodynamic radius Rh of around 5 nm. This value corresponds to SAXS results where the radius of gyration Rg is estimated to ca. 3 nm; however, this size is obtained for BSA solution of a smaller concentration of 4 mg/mL (data in Table 2). The DOSY analysis for oxyC2 gemini surfactant is shown in Figure 6b. Water and HEPES resonance lines are within the same region as in Figure 6a. The oxyC2 surfactants give a stronger signal than BSA, especially in the range 5−8 ppm. The region 0−2 ppm was selected for detailed analysis of BSA/surfactant diffusion. Accordingly, parts a−c of Figure 7 show signal decays for a set of various compositions of BSA/

oxyC2, BSA/oxyC12, and BSA/C4. The BSA signal was repeated in each figure for easier comparison. The plots of the fits were not incorporated in Figure 7 for the sake of clarity. Summarized parameters from the fits to eqs 1 and 2 are shown in Table 3. Figure 8 shows a comparison between the simple Stejskal−Tanner model for isotropic and unrestricted diffusion and the model which accounts for the distribution of particle size (eq 2). Diffusion data was collected for 8 mg/mL native-BSA as well as 10 and 20 mM oxyC2 gemini surfactants and a set of various oxyC2/BSA compositions varied from 0.05 mM up to 20 mM. The diffusion coefficient for native-BSA (8 mg/mL) obtained from the fit by the Stejskal−Tanner equation gives DBSA = 4.1 × 10−11 m2/s, which corresponds to a hydrodynamic radius of 5 nm. The diffusion coefficients for native oxyC2 of 10 and 20 mM concentrations are comparable with those for native-BSA. In NMRD measurements, similarly to SRSAXS, the precipitate formation was observed for certain gemini surfactant concentrations, which for NMRD were around 1−5 mM. In the F

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Table 3. Fit Parameters for Selected Data Curves (eqs 2 and 3) as Obtained from Diffusion NMRa sample (c (mM)) BSA (110 μM) oxyC2 (0.05 mM), BSA (110 μM) oxyC2 (10 mM) oxyC2 (10 mM), BSA (110 μM) oxyC2 (20 mM) oxyC2 (20 mM), BSA (110 μM) oxyC12 (0.05 mM), BSA (110 μM) oxyC2 (10 mM) oxyC12 (5 mM), BSA (110 μM) oxyC2 (20 mM) oxyC12 (20 mM), BSA (110 μM) C4 (1 mM) C4 (0.1 mM), BSA (110 μM) C4 (4 mM), BSA (110 μM) C4 (8 mM), BSA (110 μM) C4 (15 mM), BSA (110 μM) a

D (m2/s) --2.5 --2.5 6.6 4.1 4.4

× 10−10 × × × ×

−10

10 10−11 10−11 10−10 −11

3.2 × 10 ----2.0 × 10−10 1.4 × 10−10 8.0 × 10−11 35 × 10−11 2.0 × 10−10

Ddist (m2/s) 4.1 4.2 4.8 2.5 3.3 1.9 3.8 4.4 1.3 3.5 2.6 5.1 3.5 2.4 2.8 2.5

× × × × × × × × × × × × × × × ×

AF

−11

10 10−11 10−11 10−11 10−11 10−11 10−11 10−11 10−11 10−11 10−11 10−11 10−11 10−11 10−11 10−11

0.45 --0.13 0.48 0.52 0.24 0.62 ----0.20 0.59 0.38 0.14 0.10

1 − AF 1 0.55 0.93 0.87 0.52 0.48 0.76 0.95 0.39 0.99 0.99 0.80 0.41 0.62 0.86 0.90

σ (m2/s) ----1.3 7.4 2.6 4.1 --1.2 --1.0 0.8 ----5.9 1.0 0.8

× × × ×

10−11 10−12 10−12 10−12

× 10−11 × 10−11 × 10−11

× 10−11 × 10−11 × 10−11

Rh (nm) (min−max) 5.0 0.8−5.0 4.5 0.8−9.0 3.0−6.0 5.0−11.0 0.5−5.6 4.9 7.0−16.0 6.0 8.4 1.0−4.3 1.5−6.2 2.7−9.0 0.6−7.7 1.0−8.6

For multiple component diffusion, the hydrodynamic radius is given as “min” and “max”.

radius is higher than the “raw” gyration radius dependent on the electron density contrast as obtained from X-ray scattering. Also, a higher BSA concentration was used for NMR experiments which also influences aggregation. For csurf:cBSA from around 25 to 125 for oxyC2, a transition (reduction of protein aggregation level) occurs (denoted in Figure 9 as a dotted line and SR-SAXS polydispersity region) where we could not resolve a gyration ratio from SAXS data. In this region, the percentage of α-helix in the BSA molecule drops down to 24−31% which also correlates with the decrease of the hydrodynamic radius down to 9 nm, which indicates that the degree of aggregation for BSA is lowered. Then, for higher csurf:cBSA ratios, a semi-plateau is observed where the gyration radius Rg is less affected by surfactant concentrations. A similar transition is observed for NMR diffusion experiments and for similar molar ratios of around 10−45:1, which is in good agreement with data from SR-SAXS, considering that for NMR experiments a higher BSA concentration was used. The gyration radius obtained for oxyC12 of csurf:cBSA = 25 (Figure 9b) is comparable to the gyration radius obtained for oxyC2. Although for higher molar ratios we were unable to derive Rg for oxyC12 from SR-SAXS, we succeeded with diffusion NMR. The hydrodynamic radii obtained from NMR for oxyC12 were similar to those obtained for oxyC2, which indicates that a similar BSA aggregation level is achieved. For C4 surfactant, we were able to derive both gyration and hydrodynamic radii (Figure 9c). Despite that the α-helix content drop is not so pronounced as for the other two surfactants, the gyration radii are much smaller despite the higher C4 surfactant concentration. Again, we were unable to obtain Rg for 0.5 and 1 molar ratios, so the polydispersity region is reached for smaller C4 concentrations which suggests that interaction of C4 with BSA indeed starts for lower concentrations than for the other two gemini surfactants. Summarizing, according to data shown in Figure 9c and Table 2, the C4 surfactant prevents BSA aggregation at smaller ratios than oxyC2 and oxyC12 surfactants. The Impact of Surfactant Type on the BSA Structure. Three different surfactants in the study can be characterized by different structural properties. The oxyC2 is similar to oxyC12 and to C4. Gemini surfactant oxyC2 differs from oxyC12 by spacer length while keeping similar chemical characteristics of

Figure 8. Comparison between the two-exponential fit by the Stejskal−Tanner equation (eq 1) and that by the combined Stejskal−Tanner equation with a Gaussian distribution of diffusion coefficients (eq 3) for 15 mM C4 gemini surfactant and BSA.

same concentration range, a significant change in diffusion took place, as evident from Figure 7 (change of slopes). This distinct system transition occurred at about 10−45:1 for oxyC2 and oxyC12 and at about 20−35:1 for C4. The hydrodynamic radius rises from about 5.0 to 8.4−11 nm for different surfactants, as a result of attachment of gemini molecules, which very well coincides with the results obtained from SRSAXS (initial and final values of radii of gyration are comparable with initial and final values of hydrodynamic radii, respectively). Comparative Analysis of Results from SR-SAXS and NMR Diffusion. A comparison of changes in the radii of gyration, hydrodynamic radii, and percentage of α-helicity with increasing surfactant-to-BSA (csurf:cBSA) molar ratios for all surfactants in the study is shown in Figure 9. The results for oxyC2, oxyC12, and C4 are shown in parts a, b, and c of Figure 9, respectively. The gyration radius Rg obtained from SR-SAXS and hydrodynamic radius Rh are well correlated in general (e.g., the gyration radius is proportional to the hydrodynamic radius obtained from diffusion NMR). Starting from zero concentration of oxyC2 surfactant (Figure 9a), both Rg and Rh start to grow gradually with an increasing molar ratio csurf:cBSA. The hydrodynamic radius obtained from NMR is slightly higher than that obtained from SR-SAXS, since the hydrodynamic G

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

polydisperse and gyration radii cannot be derived from SRSAXS experiments. At the same time, the α-helix content in BSA is dropping down to 24% (BSA in extended form) for oxyC2 and down to 35% for oxyC12, respectively, which indicates that an interaction between surfactant molecules and α-helices is more effective for shorter spacer, albeit in that particular case we were unable to interpret some CD results. This does not exclude the possibility that the impact of oxyC12 on α-helix content is even stronger so that CD cannot detect a signal from α-helices anymore. The reason for that is that the shorter spacer of oxyC2 molecule is more rigid due to repulsion forces between positively charged heads localized on the opposite ends of the spacer. There are some papers that indicate that a short-spacer surfactant might lower its ability to bind with proteins, as compared with that of a surfactant having a longer spacer, due to inflexibility.42 On the other hand, the long spacer of oxyC12 surfactant renders the connection of polar heads too elastic to bind to the protein, which can also weaken the surfactant binding strength.42 The system of micelle-like clusters of surfactant molecules bound among tightly packed BSA chains is complex, and the description of detailed factors governing this process may be difficult to disclose.43 Both surfactants oxyC2 and C4 have similar spacer lengths. Surfactant C4 has a hydrophobic (butyl) spacer group, while the oxyC2 spacer group (1,2-dimetoxyethyl) contains two oxygen atoms. The presence of two oxygen atoms in the spacer group of oxyC2 results in a much higher loss of α-helix content in the system if compared to C4. The content of α-helix dropped from about 55 to 24% for oxyC2 and to 37% for C4, respectively. This stronger influence of oxyC2 on α-helix content may result from an increased distance between polar heads, the flexibility of the chain, and also the presence of two additional oxygen atoms. For an alkyl-type spacer (−(CH2)n−), Wang et al. discovered that for n = 4 the interaction with BSA was the strongest.42 They reasoned that, for n = 4, the surfactant’s spacer is not that rigid as for n = 3 and not so flexible as for n = 6. In our case, however, it was found that the −CH2−O−(CH2)2−O−CH2− spacer, approximately six carbon atoms long, brought about a stronger binding ability of the surfactant than the surfactant with a −(CH2)4− spacer. Nevertheless, the conclusion that there is an optimal length of spacer for a particular gemini surfactant remains relevant. On the other hand, the more polar character of the spacer of oxyC2 is likely to contribute to a stronger interaction of surfactant and α-helices, which has a tremendous impact on the unfolding behavior and transitions from N to E BSA forms. Without surfactant, BSA solution is stabilized by the negative charge of I and II domains.33 The binding of a surfactant with BSA decreases the overall charge, and steric stabilization is weakened due to smaller repulsion interactions. As a result, BSA starts to aggregate without a pronounced change in α-helix for all surfactants in the study (CD data). At the same time, BSA stays in its native normal form. Then, for higher surfactant concentrations, the N−E transition occurs and α-helices are turned into β-sheets as well as other forms (according to CD data).33,44 Single surfactant molecules bind through the electrostatic interactions between the positive charge of gemini polar heads and negatively charged BSA residues. At a pH of 7.5−7.7, BSA’s overall charge is negative, as the isoelectric point of BSA is 4.7. At small concentrations, double-chained gemini surfactants may contribute to stabilization of the secondary structure of BSA, as

Figure 9. Comparison of the character of changes in the gyration radii, hydrodynamic radii, and α-helicity caused by increasing surfactant-toBSA concentration ratio for all three examined surfactants: (a) oxyC2, (b) oxyC12, and (c) C4. Inability of radius of gyration determination was denoted as “SAXS polydispersity region”.

hydrogen bonding hydrophilic groups. At the same time, oxyC2 is similar to C4 surfactant, since the spacer group length is slightly longer, while C4 does not have additional hydrophilic oxygen groups. The role of the length and chemical composition of the spacer group on the interaction strength is quite complex, although it is apparent that a longer spacer group is not affecting BSA structure as effectively as shorter spacers in the case of oxyC2 (Figure 9a,b). Nevertheless, if we compare the spacer length and its impact on the gyration radii, the oxyC2 surfactant is preventing BSA from aggregating at lower concentrations than oxyC12, since the gyration radii are smaller for comparable concentrations. Despite that, at higher concentrations, both oxyC2 and oxyC12 make aggregates H

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

(3) Pietralik, Z.; Krzyszton, R.; Kida, W.; Andrzejewska, W.; Kozak, M. Structure and Conformational Dynamics of DMPC/Dicationic Surfactant and DMPC/Dicationic Surfactant/DNA Systems. Int. J. Mol. Sci. 2013, 14 (4), 7642−7659. (4) Krejci, J. Interaction of mixture of anionic surfactants with collagen. Int. J. Cosmet. Sci. 2007, 29 (2), 121−9. (5) Hu, M.; Wang, X.; Wang, H.; Chai, Y.; He, Y.; Song, G. Fluorescence spectroscopic studies on the interaction of Gemini surfactant 14−6-14 with bovine serum albumin. Luminescence 2012, 27 (3), 204−210. (6) Tadros, T. F. Surfactants in the Food Industry; Wiley-VCH Verlag GmbH & Co. KGaA: Wienheim, FRG, 2005. (7) Rosen, M. J.; Kunjappu, J. T. Surfactants in Biology, 4th ed.; John Wiley & Sons, Inc.: Hoboken, NJ, 2012. (8) Cserhati, T.; Forgacs, E.; Deyl, Z.; Miksik, I.; Eckhardt, A. Interaction of surfactants with homologous series of peptides studied by reversed-phase thin-layer chromatography. J. Chromatogr. A 2001, 910 (1), 137−145. (9) Zhang, Q.; Tao, H.; Hong, W.-X. New amphiphiles for membrane protein structural biology. Methods 2011, 55 (4), 318−323. (10) Otzen, D. Protein-surfactant interactions: A tale of many states. Biochim. Biophys. Acta, Proteins Proteomics 2011, 1814 (5), 562−591. (11) Cao, M.; Han, Y.; Wang, J.; Wang, Y. Modulation of fibrillogenesis of amyloid beta(1−40) peptide with cationic gemini surfactant. J. Phys. Chem. B 2007, 111 (47), 13436−13443. (12) Han, Y.; He, C.; Cao, M.; Huang, X.; Wang, Y.; Li, Z. Facile Disassembly of Amyloid Fibrils Using Gemini Surfactant Micelles. Langmuir 2010, 26 (3), 1583−1587. (13) Zana, R. Dimeric (gemini) surfactants: Effect of the spacer group on the association behavior in aqueous solution. J. Colloid Interface Sci. 2002, 248 (2), 203−220. (14) Rosen, M. J. Gemini surfactants; Elsevier: 1999; pp 151−161. (15) Sharma, V. D.; Ilies, M. A. Heterocyclic Cationic Gemini Surfactants: A Comparative Overview of Their Synthesis, Selfassembling, Physicochemical, and Biological Properties. Med. Res. Rev. 2014, 34 (1), 1−44. (16) Perez, L.; Pinazo, A.; Pons, R.; Infate, M. Gemini surfactants from natural amino acids. Adv. Colloid Interface Sci. 2014, 205C, 134− 155. (17) Sakai, K.; Sakai, H.; Abe, M. Recent Advances in Gemini Surfactants: Oleic Acid-Based Gemini Surfactants and Polymerizable Gemini Surfactants. J. Oleo Sci. 2011, 60 (4), 159−163. (18) Zana, R. Dimeric and oligomeric surfactants. Behavior at interfaces and in aqueous solution: a review. Adv. Colloid Interface Sci. 2002, 97 (1−3), 205−253. (19) Quinlan, G. J.; Martin, G. S.; Evans, T. W. Albumin: Biochemical properties and therapeutic potential. Hepatology 2005, 41 (6), 1211−1219. (20) Vlasova, I. M.; Zhuravleva, V. V.; Vlasov, A. A.; Saletsky, A. M. Interaction of cationic surfactant cethyltrimethylammonium bromide with bovine serum albumin in dependence on pH: A study of tryptophan fluorescence. J. Mol. Struct. 2013, 1034, 89−94. (21) Pietralik, Z.; Taube, M.; Skrzypczak, A.; Kozak, M. SAXS Study of Influence of Gemini Surfactant, 1,1′-(1,4-butanediyl)bis 3-cyclododecyloxymethylimidazolium di-chloride, on the Fully Hydrated DMPC. Acta Phys. Pol., A 2010, 117 (2), 311−314. (22) Whitmore, L.; Wallace, B. A. DICHROWEB, an online server for protein secondary structure analyses from circular dichroism spectroscopic data. Nucleic Acids Res. 2004, 32, W668−W673. (23) Compton, L. A.; Johnson, W. C. Analysis of protein circular dichroism spectra for secondary structure using a simple matrix multiplication. Anal. Biochem. 1986, 155, 155−167. (24) Lees, J. G.; Miles, A. J.; Wien, F.; Wallace, B. A. A reference database for circular dichroism spectroscopy covering fold and secondary structure space. Bioinformatics 2006, 22 (16), 1955−1962. (25) Pernot, P.; Round, A.; Barrett, R.; Antolinos, A. D. M.; Gobbo, A.; Gordon, E.; Huet, J.; Kieffer, J.; Lentini, M.; Mattenet, M.; Morawe, C.; Mueller-Dieckmann, C.; Ohlsson, S.; Schmid, W.; Surr, J.; Theveneau, P.; Zerrad, L.; McSweeney, S. Upgraded ESRF BM29

gemini molecules attach their hydrophobic tails to different, adjacent hydrophobic sites at the BSA surface, so that its secondary structure is preserved (N-form). As the surfactant concentration increases, the tails of already bound to BSA surfactant molecules can interact not only with hydrophobic regions of the protein but also with tails of other bound surfactant molecules as well. This leads to formation of micellelike clusters of gemini surfactant; a surfactant interacts with BSA cooperatively. Surfactant molecules start to enter the hydrophobic interior of the protein, leading to exposure of the inner residues out to the solvent. The way of binding of gemini surfactants to BSA, therefore, promotes a decrease in the αhelix content as an indicator of unfolding of BSA. As the surfactant concentration increases further, an increasing number of surfactant molecules bind with the protein. The micellar structures grow in extent and start to repel each other strongly, which brings a further expansion of BSA chain. At a certain stage, BSA is fully denaturated and the surfactant, in the form of micelles, is bound to the unfolded chain.



CONCLUSIONS The binding of a surfactant with BSA is decreasing its overall negative charge; therefore, steric stabilization is weakened due to weaker repulsion interactions. This results in BSA aggregation and increased gyration and hydrodynamic radii, albeit without a pronounced change in α-helix content for all surfactants in the study (CD data). During initial aggregation, BSA is preserved at its native normal form at pH 7.7. Then, for higher surfactant concentrations, the interaction of α-helices with surfactants induces the N−E transition and a dramatic decrease of α-helix content. All three methods used different protein concentrations, although the same surfactant-to-protein molar ratios were kept. Some differences between hydrodynamic and gyration radii were observed due to different absolute BSA concentrations. This indicates that both absolute concentration values and molar ratios play a significant role in determining the system properties. The drop in α-helix content versus surfactant concentration is more pronounced for oxyC2 and oxyC12 gemini surfactants. The C4 gemini surfactant requires higher concentrations than oxyC2 and oxyC12 gemini surfactants in order to induce the N−E transition, although it better prevents BSA aggregation at smaller ratios.



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS The research was supported by a research grant (UMO-2011/ 01/B/ST5/00846) from National Science Centre (Poland). REFERENCES

(1) Bombelli, C.; Giansanti, L.; Luciani, P.; Mancini, G. Gemini Surfactant Based Carriers in Gene and Drug Delivery. Curr. Med. Chem. 2009, 16 (2), 171−183. (2) Hamrang, Z.; Zindy, E.; Clarke, D.; Pluen, A. Real-time evaluation of aggregation using confocal imaging and image analysis tools. Analyst 2014, 139 (3), 564−568. I

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

beamline for SAXS on macromolecules in solution. J. Synchrotron Radiat. 2013, 20, 660−664. (26) Huang, T. C.; Toraya, H.; Blanton, T. N.; Wu, Y. X-Ray-Powder diffraction analysis of silver behenate, a possible low-angle diffraction standard. J. Appl. Crystallogr. 1993, 26, 180−184. (27) Kozak, M. Glucose isomerase from Streptomyces rubiginosus potential molecular weight standard for small-angle X-ray scattering. J. Appl. Crystallogr. 2005, 38, 555−558. (28) Konarev, P. V.; Volkov, V. V.; Sokolova, A. V.; Koch, M. H. J.; Svergun, D. I. PRIMUS - a Windows-PC based system for small-angle scattering data analysis. J. Appl. Crystallogr. 2003, 36, 1277−1282. (29) Stilbs, P. Fourier transform NMR pulsed-gradient spin echo (FT-PGSE) self-diffusion measurements of solubilization equilibria in SDS solutions. J. Colloid Interface Sci. 1982, 87 (2), 385−394. (30) Stejskal, E. O.; Tanner, J. E. Spin diffusion measurements: spin echoes in the presence of a time-dependent field gradient. J. Chem. Phys. 1965, 42 (1), 288. (31) Huang, B. X.; Kim, H. Y.; Dass, C. Probing three-dimensional structure of bovine serum albumin by chemical cross-linking and mass spectrometry. J. Am. Soc. Mass Spectrom. 2004, 15 (8), 1237−1247. (32) Peters, T. SERUM-ALBUMIN. Adv. Protein Chem. 1985, 37, 161−245. (33) Foster, J. F. Albumin Structure, Function and Uses; Pergamon: Oxford, U.K., 1977; pp 53−84. (34) Zhou, T.; Ao, M.; Xu, G.; Liu, T.; Zhang, J. Interactions of bovine serum albumin with cationic imidazolium and quaternary ammonium gemini surfactants: Effects of surfactant architecture. J. Colloid Interface Sci. 2013, 389, 175−181. (35) Mir, M. A.; Khan, J. M.; Khan, R. H.; Rather, G. M.; Dar, A. A. Effect of spacer length of alkanediyl-alpha,omega-bis(dimethylcetylammonium bromide) gemini homologues on the interfacial and physicochemical properties of BSA. Colloids Surf., B 2010, 77 (1), 54−59. (36) Carter, D. C.; Ho, J. X. Structure of Serum-Albumin. Adv. Protein Chem. 1994, 45, 153−203. (37) Li, Y.; Wang, X.; Wang, Y. Comparative Studies on Interactions of Bovine Serum Albumin with Cationic Gemini and Single-Chain Surfactants. J. Phys. Chem. B 2006, 110, 8499−8505. (38) Pi, Y.; Shang, Y.; Peng, C.; Liu, H.; Hu, Y.; Jiang, J. Interactions between bovine serum albumin and gemini surfactant alkanediylbis(dimethyldodecyl-ammonium bromide). Biopolymers 2006, 83 (3), 243−249. (39) Tai, S.; Liu, X.; Chen, W.; Gao, Z.; Niu, F. Spectroscopic studies on the interactions of bovine serum albumin with alkyl sulfate gemini surfactants. Colloids Surf., A 2014, 441, 532−538. (40) Mir, M. A.; Gull, N.; Khan, J. M.; Khan, R. H.; Dar, A. A.; Rather, G. M. Interaction of Bovine Serum Albumin with Cationic Single Chain plus Nonionic and Cationic Gemini plus Nonionic Binary Surfactant Mixtures. J. Phys. Chem. B 2010, 114 (9), 3197− 3204. (41) Zhu, L.-Y.; Li, G.-Q.; Zheng, F.-Y. Interaction of bovine serum albumin with two alkylimidazolium-based ionic liquids investigated by microcalorimetry and circular dichroism. J. Biophys. Chem. 2011, 2 (2), 146−151. (42) Wang, H.; Jiang, X.; Zhou, L.; Cheng, Z.; Yin, W.; Duan, M.; Liu, P. Interaction of NAEn-s-n gemini surfactants with bovine serum albumin: A structure−activity probe. J. Lumin. 2013, 134, 138−147. (43) Ge, Y.-S.; Tai, S.-X.; Xu, Z.-Q.; Lai, L.; Tian, F.-F.; Li, D.-W.; Jiang, F.-L.; Liu, Y.; Gao, Z.-N. Synthesis of Three Novel Anionic Gemini Surfactants and Comparative Studies of Their Assemble Behavior in the Presence of Bovine Serum Albumin. Langmuir 2012, 28, 5913−5920. (44) Khan, M. Y. Direct evidence for the involvement of domain-III in the N-F transition of bovine serum-albumin. Biochem. J. 1986, 236 (1), 307−310.

J

dx.doi.org/10.1021/jp5047485 | J. Phys. Chem. B XXXX, XXX, XXX−XXX