GM1-Induced Structural Changes of Bovine Serum Albumin after

(10, 11) GM1 possesses only one sialic acid attached to the second galactose ... For the CD experiments with the urea-denatured protein, a 20 mM ... f...
0 downloads 0 Views 217KB Size
974

Biomacromolecules 2008, 9, 974–983

GM1-Induced Structural Changes of Bovine Serum Albumin after Chemical and Thermal Disruption of the Secondary Structure: A Spectroscopic Comparison Anindita Gayen, Chiradip Chatterjee,† and Chaitali Mukhopadhyay* Department of Chemistry, University of Calcutta, India Received March 28, 2007; Revised Manuscript Received December 10, 2007

GM1-induced structural transitions of native and unfolded conformers of bovine serum albumin (BSA) have been studied where in the unfolded conformers, the secondary structures were disrupted either chemically by 8 M urea or thermally by heating at 65 °C. With decreasing protein:ganglioside ratio at pH 7.0, the native BSA partially unfolds and expands, while the urea-denatured BSA forms an R-helical structural pattern with shrinking in the conformational space. However, a continuous loss of R-helicity with minor increase in size was observed for the thermally altered protein in the presence of the GM1 micelle. The changes in the secondary structural content were followed by far-UV circular dichroism (CD) analysis. The dynamic light scattering (DLS) experiments were used to study the variation of the size of the protein-GM1 complexes with increasing concentration of the GM1. Fluorescence experiments show that tryptophan residues of BSA experience a more hydrophobic environment in the presence of the GM1 micelle with a decreasing protein:ganglioside ratio at pH 7.0. The present study shows that GM1 has a strong effect on the conformation of BSA depending on the conformational states of the protein that would relate to a physiological function of GM1 such as acting as the receptor of proteins in the cell membrane.

1. Introduction Gangliosides, the most complex of glycosphingolipids, are abundant in the plasma membrane of nerve cells (making up 5–10% of the total lipid mass) and are found widely in most vertebrate cell types.1 Gangliosides are acidic glycolipids showing strong amphiphilic character with the hydrophobic ceramide moiety inserted into the external leaflet of the cell membrane and the hydrophilic oligosaccharide head group with one or more N-acetyl neuraminic acids (sialic acid) faced toward the extracellular space. Protein can cross-link gangliosides on the cell surface, which may cause redistribution of these gangliosides in the segregated regions, resulting in modification of the membrane fluidity.2 This redistribution and receptor property of gangliosides support the idea that interaction between protein and membrane gangliosides are determinants not only of membrane structure but also function.2 Gangliosides are believed to be involved in cell-cell/substratum interactions, modulation of the transmembrane signaling, calcium homeostasis, and synaptic transmission.3 Previously, Katagiri et al. found an interesting phenomenon in the interaction of gangliosides with proteins, where the solubility of gangliosides to biphasic solvent (the upper hydrophilic phase and lower hydrophobic phase) changes drastically with the addition of methylated bovine serum albumin (BSA), which is used as a model chemical for basic proteins in synaptic membranes.4–7 Then, Takizawa et al. and Hirai et al. showed that this phenomenon results from the heterogeneous patchlike binding of ganglioside to an albumin molecule.8,9 Recently, Hayashi et al., using X-ray and neutron scattering techniques, found * Corresponding author. E-mail: [email protected]. Telephone: 91033-2351-8386. Fax: 91-033-2351-9755. GM1 induced Structural Changes of BSA, Dr. Chaitali Mukhopadhyay, Department of Chemistry, University of Calcutta, 92, APC Road, Kolkata 700 009, India. † Present address: Department of Biological Sciences, Structural Biology Division, National University of Singapore, Singapore.

Figure 1. Chemical structure of ganglioside GM1.

differences in the interaction process of three different types of ganglioside samples with various bovine serum albumins whose surface was chemically modified by different saccharides.3 The gangliosides that are the major species in nerve cells are ganglioside monosialo type I (GM1) and ganglioside disialo type I (GD1a).1 GM1 acts as a receptor of cholera toxin, plays the role of a seed for formation of amyloidal β-fibrils, elicits the down-modulation of CD4, which is the binding site of HIV, and was recently found to specifically interact with R-synuclein to inhibit fibrillation, the key step of Parkinson’s disease.10,11 GM1 possesses only one sialic acid attached to the second galactose residue (gal) of the neutral tetrasaccharide, which defines the glycolipid (Galβ1-3GalNacβ1-4Galβ1-4Glcβ1-1ceramide) (Figure 1). The binding of serum albumin with the GM1 micelle was previously studied by absorption and fluorescence

10.1021/bm701144k CCC: $40.75  2008 American Chemical Society Published on Web 01/19/2008

GM1-Induced Structural Changes of BSA

Biomacromolecules, Vol. 9, No. 3, 2008 975

spectroscopy by Tomashi and Roda.12 The binding of ganglioside GM1 to bovine serum albumin was reported by the authors as a phenomenon where the native protein formed 1:1 or 1:2 association complexes with the GM1 micelle. In this present study, we report the process of complexation of GM1 as a micelle with native and structurally disrupted bovine serum albumins. Here we report a differential interaction of ganglioside GM1 with native, urea-unfolded, and thermally disrupted bovine serum albumin conformers through fluorescence, circular dichroism (CD), and dynamic light scattering approaches. We have also used the red edge excitation shift (REES) to monitor the organization and dynamics of tryptophan residues of native and disrupted BSA in presence of the GM1 micelle. The results clearly indicate that GM1 at neutral pH induces native protein to partially unfold and expand, while urea-denatured BSA forms some degree of secondary structure with R-helical character. Again, BSA after thermal denaturation at 65 °C with retained 44% of the R-helicity suffers continuous loss of R-helicity with minor increase in size in the presence of the GM1 micelle. Finally, the comparison of the results of fluorescence, CD, and DLS characterizes the possible BSA-GM1 complexes with different [BSA]:[GM1] ratios under native and non-native conditions.

where ν ) average number of GM1 molecules bound per BSA, R ) fraction of protein bound to GM1, [Ts] ) total GM1 concentration, [Tp] ) total protein concentration, and R is determined from the intensities of Trp fluorescence as

2. Experimental Section

r ) (Ivv - GIvh) ⁄ (Ivv + 2GIvh)

2.1. Materials. BSA was purchased from Sigma Aldrich with 98% purity in the monomeric form (A4378) and used without further purification. Ganglioside GM1 was isolated and purified from goat brain following our published protocol.13–15 Urea and Tris were from SRL, and urea was used after recrystallization from alcohol. A 10 mM TrisHCl buffer of pH 7.0 was prepared in double-distilled water for all the experiments. For the CD experiments with the urea-denatured protein, a 20 mM phosphate buffer of pH 7.0 was exclusively prepared in deionized water. Protein concentration was determined by spectrometric measurement using 280 ) 44720 M-1 cm-1 for BSA at neutral pH. The final concentration of protein in solution was about 5 µM for each experiment except where mentioned. Throughout the experiments, the concentration of GM1 was maintained well above its CMC (3.32 × 10-8 to 3.2 × 10-6 M), confirming the micellar character of GM1.16 2.2. Methods. Denaturation by Urea (8 M). Stock solution of ∼8.5 M urea in Tris buffer (pH 7.0) was prepared just before use. To the weighed urea, Tris-HCl buffer of pH 7.0 was added, vortexed, and the volume was made up to the level to make 8.5 M urea in Tris buffer (pH 7.0). Then 2–3 µL of protein from concentrated stock was added to the urea-Tris buffer and the solution was stored overnight so that the concentration of urea in solution remained above 8 M. GM1 from a 20 mM concentrated stock solution was added in the denatured protein solution so that the dilution effect was minimized and the protein was maintained in its completely denatured form during all the experiments. Denaturation by Heat (65 °C). The protein solution was kept in the cell of the respective instruments, and the temperature was increased gradually using the temperature controller. The protein solution was incubated at 65 °C for at least 15 min prior to addition of GM1. GM1 was added to the protein solution at 65 °C. Then the protein-GM1 solution was kept for 5 min or more at 65 °C to ensure the achieved temperature of the medium. No turbidity or precipitation was observed during the experimental conditions.17–19 Steady-State Fluorescence Spectra. Steady-state fluorescence spectra were performed on a Perkin-Elmer LS-50B spectrophotometer using FL WinLab software. Experiments were done at room temperature except the thermal experiments, which were performed at 65 °C. The excitation and emission slits with a band pass of 5 nm were used for all the measurements. Background intensities of the buffer blanks in which BSA was omitted were subtracted from each sample spectrum to cancel out any contribution due to the solvent. To measure the changes of the intrinsic fluorescence of tryptophan upon GM1 binding,

small-volume aliquots from a GM1 stock solution were added to the protein solution in a cell. The average number of GM1 molecules bound per BSA molecule at different GM1 concentrations can be determined by monitoring the intrinsic tryptophan fluorescence (λexc 280 nm). The following formula was used for the calculation of νGM1, the number of GM1 molecules bound per protein molecule:20

ν)R

[Ts] [Tp]

R ) [Iobs - Ifree] ⁄ [Imin - Ifree]

(1)

(2)

where Iobs ) Trp fluorescence intensity at any GM1 concentration, Imin ) Trp fluorescence intensity at the saturation region of plot of ∆λmax against the concentration of GM1, and Ifree ) Trp fluorescence intensity at the zero concentration of GM1. Fluorescence Anisotropy. The fluorescence anisotropy measurements were carried out on the Perkin-Elmer LS-50B spectrophotometer using FL WinLab software. Steady-state anisotropy (r) values are calculated in this paper as a ratio of the intensity (I) of vertically (v) and horizontally (h) polarized emission, according to the equation

(3)

where G is the grating factor measured as Ihv/Ihh and the double subscripts refer to the direction of excitation and the emission polarization, respectively. Experiments were conducted using excitation and emission wavelengths of 295 and 350 nm for tryptophan of the protein. The concentration of BSA was 10 µM in fluorescence anisotropy experiments. The excitation and emission slits used for fluorescence anisotropy were 4 and 8 nm, respectively. Fluorescence Quenching. For the acrylamide quenching measurements, the excitation wavelength used was 295 nm and the emission was monitored at 340 nm. Quenching data were analyzed by fitting them to the classical Stern-Volmer equation

F0 ⁄ FC ) 1 + KSV[Q] ) 1 + kqι0[Q]

(4)

where F0 and FC are the fluorescence intensities in presence and absence of the quencher acrylamide, [Q] is the molar concentration of the quencher, KSV is the collisional quenching constant (M-1), kq is the bimolecular quenching constant, and ι0 is the unquenched excited-state average fluorescence lifetime. The fluorescence intensities were corrected for absorption of exciting light and reabsorption of the emitted light to decrease the inner filter effect using the relationship

Fcor ) Fobs × e(Aex+Aem) ⁄ 2

(5)

where Fcor and Fobs are the fluorescence intensities corrected and observed, respectively, and Aex and Aem are the absorption of the systems at the excitation and the emission wavelength, respectively. The intensity of fluorescence used in this paper is the corrected fluorescence intensity. The steady-state fluorescence quenching experiments were performed with the Perkin-Elmer LS-50B spectrophotometer using polar uncharged acrylamide. Experiments were done with the protein solution in the absence and presence of the GM1 micelle. The concentration of the protein was maintained at 5 µM throughout the experiment. The fluorescence lifetimes were determined from the time-resolved fluorescence intensity decays using the spectrophotometer (LifeSpecps) from Edinburgh Instrument, UK, with a 460 ps instrument response function at the excitation wavelength of 299 nm to trigger the tryptophan fluorescence. The observed fluorescence transients were fitted by using a nonlinear least-squares fitting procedure to a function (X(t) ) ∫t0 E(t′)

976 Biomacromolecules, Vol. 9, No. 3, 2008

Gayen et al.

R(t - t′) dt) comprising the convolution of the IRF (E(t)) with a sum N of exponentials (R(t) ) A + ∑i)1 Biet/τ1) with pre-exponential factors (Bi), characteristic lifetimes (ιi), and a background (A). Relative contribution of each component was obtained using a triexponential fitting and finally was expressed by the following equation: N

⁄∑

an ) Bn

Bi

(6)

i)1

The mean lifetimes for the decay curves were calculated from the decay times and the relative contribution of the components using the following relation

〈τ0 〉 )

∑ Riτi2 i

∑ Riτi

(7)

i

where 〈ι0〉 is the mean lifetime of tryptophan. Circular Dichroism Measurements. CD spectra were recorded with a Jasco J-720 spectropolarimeter using a 10 cm path length demountable silica quartz cell. Measurements were taken at wavelengths between 190 and 240 nm with 0.1 nm step resolution and averaged over five scans recorded as a speed of 20 nm/min. The measurements at 65 °C were carried out using a Jasco J-600 spectropolarimeter at a scan speed of 50 nm/min with other parameters unaltered. The protein solution along with GM1 was put into a 0.1 cm path length cuvette attached with an incubation jacket, through which water was circulated at the desired temperature. All observed spectra were baseline subtracted for buffer and the helicity was calculated on the basis of change of meas [θ222] using the following equation:21,22 meas

helicity )

[θ222]

max

[θ222]

(8)

where max[θ222] stands for 100% helicity and is estimated using the formula

[θ222] ) -40000 × [1 - (2.5 ⁄ n)]

max

(9)

Here n implies the total number of amino acid residues in protein, and [θ222] represents the molar residue ellipticity in degrees per inverse square centimeter per decimole. meas [θ222] can be measured with help of the following equation:23 meas

[θ222] )

(θobs) × M × 100 C × l × NA

(10)

In the above equation, (θ222) is the experimental ellipticity measured in degrees, M the molecular mass in Daltons, C the concentration of protein in solution in mg/mL, l is the optical path length of the cuvette in cm, and NA is the number of residues in the protein. Quantitative estimation of the secondary structure contents from the far-UV CD spectra were made by CDSSTR software.24 Results reported are from the 43 proteins data set except the denatured protein state, where 43 proteins along with the 5 denatured proteins data set were used as reference. Other reference data sets were 29, 37, and 42, which were also tested and gave comparable results. The CD spectra using the CDSSTR software are in good agreement with the value of R2, which is the ratio of the intensity of the minimum near 222 nm and the intensity of the minimum between 200 and 210 nm. R2 is independent of inaccuracies in the determination of peptide concentration as well as those caused by small shifts in wavelengths.25 For a random structure, R2 is close to zero, and for a helical conformation, R2 will approach 1. Light Scattering Measurements. Light scattering measurements on native BSA were performed at 25 °C on a DYNA PRO version VI instrument with an argon ion laser operating at 800 nm. The

denaturation studies were performed on a Nano S from Malvern Instruments employing a 4 mW He-Ne laser (λ ) 632.8 nm) equipped with a thermostat chamber. All the scattered photons are collected at a 173° scattering angle. The scattering data were processed using the instrumental software to obtain the hydrodynamic diameter (dH) and the size distribution of the scatterer in each sample. The instrument measures the time-dependent fluctuation in the intensity of the light scattered from the particles in solution at the fixed scattered angle. The hydrodynamic diameter (dH) of the protein, the GM1 micelle, and the protein-GM1 micelle complexes were estimated from the autocorrelation function of the time-dependent fluctuation in the intensity. dH is defined as

dH )

kBT 3πηD

(11)

where, kB is the Boltzman constant, η is the viscosity, and D is the translational diffusion coefficient. The viscosity parameters that were set were the viscosity of media from the system operating processor (SOP) library of the Malvern instrument. In a typical size distribution graph, the X axis shows the distribution of the size classes in nanometers and the Y axis is either the relative intensity of the scattered light or the relative number of the particles scattering the light. The thermal denaturation was done at 65 °C, and the measurements on urea-denatured BSA solution were taken by adding small aliquots of GM1 to the solution of BSA, maintaining the 8 M urea concentration in solution unaltered. The protein and GM1 solutions in buffer were passed through 0.2 µm membrane filters prior to the measurements. A protein concentration of 5 µM was used in all the light scattering experiments.

3. Results 3.1. Fluorescence Experiments. Fluorescence experiments show that GM1 can markedly influence the emission properties of BSA. At an excitation wavelength of 280 nm, the emission maximum was found at 349 nm for native BSA in absence of GM1. As can be seen from Figure 2a, increasing concentrations of GM1 caused a progressive reduction of fluorescence intensity accompanied by a blue-shift in λmax by 6 nm. Compared to this, addition of GM1 to BSA in 8 M urea gave rise to a larger blueshift of λmax from 354 to 338.5 nm (Figure 2b). In a similar experiment, BSA after incubation at 65 °C experienced a 8.5 nm blue-shift of the emission maximum with decreasing intensity, as shown in Figure 2c. This blue-shift of emission maximum along with decrease in intensity indicates that interaction of BSA with GM1 changes the environment of tryptophan with probable increase in hydrophobicity in its vicinity and this phenomenon is seen when the protein is in either native or structurally disrupted states. The nature of the emission spectra of urea- and heat-treated BSA predicts that there is an equilibrium between different conformational states present in the solution. It should be noted here that about 2:1 molar ratio of GM1:BSA caused a detectable change in the emission peak position, as observed in the inset curves of parts a, b, and c of Figure 2. It is possible to estimate the quenching of tryptophan fluorescence in the presence of the GM1 micelle using the Stern-Volmer equation. The resultant Stern-Volmer constant (KSV) values obtained using acrylamide as a quencher are shown in Table 1, and the Stern-Volmer plots are shown in Figure S1 of the Supporting Information. The decrease in the KSV values for the BSA-GM1 complex relative to the BSA, native or structurally perturbed, is in excellent agreement to the observed blue-shift of the emission maximum. However, a KSV, although twice the KSV of the native BSA, cannot account for

GM1-Induced Structural Changes of BSA

Biomacromolecules, Vol. 9, No. 3, 2008 977

Figure 3. Time-resolved fluorescence intensity decay curves of native (dotted line), chemically disrupted (solid line), and chemically disrupted BSA in the presence of the GM1 micelle (dashed line). Excitation wavelength was 299 nm, and the emission wavelengths were chosen according to the steady-state fluorescence data. The sharp peak at the extreme left is the lamp profile. The relatively broad peaks on the right are the decay profiles, which were fitted to a triexponential function. The concentration of BSA was 5 µM, and the GM1/BSA used was 100:1 (mol/mol). The solutions were made in 10 mM TrisHCl buffer of pH 7.

Figure 2. Fluorescence titration spectra of (a) native (b) chemically disrupted and (c) thermally disrupted BSA (5 µM) in presence of the GM1 micelle in 10 mM neutral Tris-HCl buffer. Rightmost spectrum in each figure is for the free protein, and successive spectra were obtained with consecutive GM1 addition to (a), (b), and (c), respectively. The inset curves in (a), (b), and (c) show the shifts of the emission maximums (in nm) that occurred during the above titrations against different GM1/BSA in the solution. Table 1. Stern-Volmer Constants from Fluorescence Spectra by Applying Stern-Volmer Equation in absence of GM1 λmax(nm) native in 8 M urea at 65 °C

349 355 340.5

-1

KSV (M

7.49 22.5 6.48

in presence of GM1 )

λmax(nm)

KSV (M-1)

343 338.5 332

3.15 13.2 5.41

the large blue-shift in the emission spectrum. To explain this, we have done the time-resolved fluorescence experiment with the urea-denatured sample in the absence and presence of GM1. The fluorescence lifetime decay profiles of the native BSA, ureadenatured BSA, and urea-denatured BSA containing GM1 samples are shown in Figure 3. As seen from Figure 3, the fluorescence decay curves could be well fitted with the triexponential function

〈τ〉 ) A + B1e-t ⁄τ1 + B2e-t ⁄τ2 + B3e-t ⁄τ3

(12)

The fluorescence lifetimes and the various statistical parameters used to check the goodness of fit are given in Table 2.

The lifetime of the tryptophan of the native BSA was found to be 3.16, 6.69, and 0.14 ns. The mean fluorescence lifetime calculated was 6.119 ns. In presence of urea, the contribution of the component with the longer lifetime decreases, and two components with lifetimes 0.25 and 2.45 ns appear in almost equal fraction (Table 2). The mean fluorescence lifetime calculated for the tryptophan in the urea-denatured condition was 3.87 ns. With GM1 present in the solution, although the lifetime of the long-lived component increased to 6.94 ns, its contribution was almost the same as in the denatured protein. The mean lifetime of tryptophan for the urea-denatured BSA in the presence of GM1 was 4.119 ns, which is still relatively small when compared to the 6.119 ns mean lifetime of the tryptophan in the native BSA. Hence, when GM1 is the quencher, the fluorescence of tryptophan was not sufficiently quenched, which results in the unexpected higher value of the KSV using acrylamide as the quencher for tryptophan. The possible explanation for the overall decrease in the Stern-Volmer quenching constant would be the tryptophan residue of BSA being more shielded in presence of the GM1 micelle. We estimated the steady-state fluorescence anisotropy of BSA in the presence of GM1 in the native condition after denaturation by 8 M urea and after application of heat (65 °C). Anisotropy can indicate the microenvironment of the protein fluorophore. The anisotropy values as can be seen in Figure 4 indicate that tryptophan experiences motional restriction when bound to the GM1 micelle. It was interesting to note that the highest change of anisotropy was produced when urea-denatured BSA interacted with the GM1 micelle. The increase in restriction in the environment around the protein tryptophan is also shown in Figure 5, where emission maximum is plotted as a function of the excitation wavelength. At the ultimate saturation condition in our experiment, where the protein:lipid molar ratio was 1:100, the emission maximum for native BSA shifted from 343.4 to 352.9 nm (Figure 5) in the presence of the GM1 micelle when the excitation wavelength was changed from 280 to 305 nm. This accounts for a red-shift with change in excitation wavelength, namely a REES of 9.5 nm. Figure 5 shows that structurally perturbed states of BSA also used to display a REES of comparable magnitude when interacting with the GM1 micelle. As the excitation wavelength

978 Biomacromolecules, Vol. 9, No. 3, 2008

Gayen et al.

Table 2. Fluorescence Lifetimes and Fitting Parameters for the Native, Urea-Denatured, and the Urea-Denatured BSA in Presence of GM1 Obtained by Triexponential Fittinga TRP environment

R1

τ1(ns)

R2

τ2 (ns)

R3

τ3 (ns)

A

χ2

native BSA urea (8 M) denatured BSA urea (8 M) denatured BSA + GM1

20 38.23 42.42

3.1648 0.2466 2.5382

54.29 41.18 12.12

6.696 2.4629 6.9383

0.86 20.59 45.45

0.1470 5.462 0.3945

0.165 0.225 0.327

0.926 0.915 1.025

a

See Materials and Methods for details.

Figure 4. Effect of GM1 addition to the anisotropy (r) of (a) native, (b) chemically disrupted, and (c) thermally disrupted BSA (10 µM in 10 mM Tris buffer of pH 7.0). The data points shown here are the ( standard error (S.E.) of three independent measurements. See Materials and Methods for other experimental details.

Figure 5. Effect of changing excitation wavelength on the emission maxima (with sigmoidal fit) of BSA in 10 mM Tris buffer of pH 7.0 in the saturation condition of the steady-state fluorescence spectra. The molar ratio of protein to lipid was chosen 1:100, according to the results obtained in the emission spectra of (a) native BSA (9), (b) chemically disrupted BSA (b), and (c) thermally disrupted BSA at 65 °C (2) with the GM1 micelle.

was changed from 280 to 300 nm (beyond λex 300 nm, the peak became broad and therefore determination of λem was erroneous for the urea-denatured BSA), the emission maximum of ureadenatured BSA changed from 338.7 to 350.6 nm and the emission maximum of the thermally disrupted BSA changed

from 332.6 to 340 nm. So, the largest shift (12 nm) was again obtained for the urea-denatured BSA-GM1 complex. The curves in Figure 5 are characteristic curves of REES. From this figure, it can be concluded that the GM1 micelle offers maximum restrictions to the native and urea-denatured BSA and the effect becomes weaker when the interaction takes place between the thermally treated BSA and GM1 micelle.26 The greater sensitivity to REES of the urea-denatured BSA could be related to the exposed charged residues of the protein, which may promote the formation of a stable hydration shell of oriented water molecules characterized by the time scale sensitive to REES measurements.26 Here, the shell around the protein is considered to be caused by the charged micellar ganglioside molecules, which form clusters at different hydrophilic regions of the protein. This strong charge–charge interaction yields a higher REES for the urea-denatured BSA in the presence of GM1. On the basis of on the fluorescence measurements above, the average number of GM1 molecules bound per BSA was roughly calculated (data not shown) as described in the Materials and Methods sections following the procedure adopted by De et al.20 The tryptophan fluorescence was monitored, and eq 1 was used to obtain γGM1, i.e., the average number of bound GM1 molecules per BSA. The γGM1 obtained were about 60, 80, and 75 for the native, urea-disrupted, and heat-disrupted BSA, respectively, at the saturation region (>1:60 BSA:GM1 molar ratio) found from the plots of shift in emission maxima of BSA against concentration of GM1. 3.2. CD Spectroscopy. Far-UV CD spectroscopy gives quantitative information on the change in secondary structure of BSA when titrated with the GM1 micelle. The far-UV CD spectral region in Figure 6a shows that the native state of BSA mostly preserves its helical secondary structure in the presence of the GM1 micelle. In contrast, GM1 was found to induce a R-helical structure to the urea-denatured BSA, which had negligible secondary structural motif in the absence of the GM1 micelle (Figure 6b). The formation of secondary structure was also found when we used neutral phosphate buffer (pH 7.0) to alleviate the effect of chloride ion having high far-UV absorption used in the Tris-HCl buffer. The spectra obtained for the ureadenatured BSA and the urea-denatured BSA in presence of GM1 (BSA:GM1 1:100 molar ratio) in phosphate buffer are really convincing and are shown in Figure S2 of the Supporting Information. The near-UV spectral region is shown in Figure S3 of the Supporting Information, which provides information about the tertiary structure of the protein. A negative band in the region 265–280 nm along with the broad positive band near 298 nm indicates that urea-denatured BSA had lost its tertiary structural elements present in the native condition. The addition of GM1 to the protein in the molar ratio 1:100 failed to change the nature of the spectra but succeeded in increasing the ellipticity. A protein conformation like BSA in 8 M urea with GM1 reported here with lack of any tertiary structure but with a secondary structure can be stated as a “molten globule state” or “incorrectly folded state”. For the thermally treated BSA (10 mM Tris buffer, pH 7.0 at 65 °C), GM1 enhances β-turn and

GM1-Induced Structural Changes of BSA

Figure 6. Selected far-UV CD spectral region showing the effect of titration of BSA (5 µM) with GM1 on ellipticity. Different spectra of (a) native, (b) chemically disrupted, and (c) thermally disrupted BSA in the figure represent various molar ratios of GM1/BSA, (2) without addition of GM1, (b) GM1 added, and (9) more GM1 added. See Table 3 for the BSA:GM1 ratios used in the CD experiment.

unstructured conformation with decreasing helicity, as shown in Figure 6c. All the structural analysis about the protein secondary structure as obtained from far-UV CD data is summarized in Table 3. It was observed that the analysis using CDSSTR software was in good agreement with R2 parameters and the changes of [θ222]. 3.3. DLS Measurements. The results of analysis of the dynamic light scattering spectra for BSA in different conditions in the absence and presence of GM1 are shown in Figure 7. Figure 7 displays the change in hydrodynamic radius of the protein in solution when titrated with the GM1 micelles. The hydrodynamic radii used in the figures are of the maximum (99.9–100%) number of particles in solution. As observed in Figure 7a, the size of BSA-GM1 complex increases with increasing GM1 concentration. The Rh in neutral Tris buffer was found as 3.4 nm (literature value 3.4 ( 0.3 nm) for the native BSA, and at the saturated condition (GM1:BSA 100:1 molar ratio), the Rh was found as 5.85 nm for the BSA-GM1 complex (Figure 7a).27,28 Addition of a very small amount of GM1 to the urea-denatured BSA (molar ratio GM1:BSA 4:1) caused a surprisingly large decrease in hydrodynamic radius from 46.13 to 8.09 nm. The Rh finally dropped to 4.27 nm when 500 µM of GM1 was present in the solution (Figure 7b). However, the change in Rh for the thermally treated BSA in the presence of the GM1 micelle was small enough, as shown in Figure 7c, and the hydrodynamic radius increased from 3.4 to 4.86 nm only when 500 µM of GM1 was added to the solution.

Biomacromolecules, Vol. 9, No. 3, 2008 979

For native BSA in Tris buffer (10 mM) of neutral pH, two peaks were observed in the intensity distribution plot (Figure 8a) corresponding to the effective diameters 7.91 and 77.91 nm. According to earlier reports, here the component having smaller size can be assigned to the native protein and the component having larger size can be assigned to the protein in self-aggregate form. However, size of the globules increased in the presence of GM1 exhibits almost twice the diameter of the native protein. Urea (8 M) denatured BSA in 10 mM Tris-HCl buffer also exhibits polydispersity of the medium. As shown in Figure 8b, two peaks of diameter 11.49 and 118.02 nm are detected in the 8 M urea-BSA system. Here, on addition of GM1 to the protein in the molar ratio GM1/BSA 100:1, along with the increase in the diameter of the globules, an additional peak appeared in the intensity distribution plot with an effective diameter of 5222 nm. The change in size was found to be relatively small for the thermally treated protein after incubation for the experimental time at 65 °C. This is similar to the result reported in the article by Wetzel et al., where the authors have found that albumin samples with a concentration of e7 mg/mL when heated to 65 °C formed no oligomers.29 However, the extent of aggregation depends not only on the temperature and the concentration but also on the time. In fact, we have found another small aggregate in the intensity distribution for the solution of BSA after heat treatment at 65 °C, which is supported by the article by Palma et al.,30 where it was reported that small-scale aggregates of average diameter 20–40 nm are formed in solution when BSA is heated to 65 °C. With added 500 µM GM1, the dH of the small size component due to the protein increased to 11.28 nm from 9.14 nm, and the diameter of the large size component due to the oligomer increased from 43.96 to 58.08 nm only. No extremely large aggregate was found during the course of thermal denaturation. 3.4. Effect of GM1 on Different Spectroscopic Results. To test the effect of GM1, we have done control with the GM1 micelle to test the effect of GM1 in fluorescence and CD spectra of the protein. GM1 is fluorescence and CD silent. Figure S4 of the Supporting Information represents far-UV CD region of the spectrum with the GM1 micelle in neutral 20 mM phosphate buffer. However, in DLS, a pure GM1 micelle shows a dH of 11.6 nm, which is consistent with the previously reported data in literature. The aggregation number of a GM1 micelle having this diameter can be considered to be 350, as the earlier reports suggest.16 For the GM1 micelle at 65 °C, the dH decreases to 9.8 nm and therefore the aggregation number also decreases to 295 relative to that at the room temperature.16

4. Discussion Bovine serum albumin of MW 66411 Da, calculated from amino acid composition, consists of 583 amino acids in a single polypeptide chain. The protein is known to have a heart-shaped structure (N form) in solution with a net charge of -18 on its surface.31 The structure of BSA consists of nine loops resulting in three domains (I, II, III), each having a large double loop, a short helical segment, a small double loop, a long connecting segment or hinge, again a large double loop, and another connecting segment to the next domain.32,33 These large loops are composed mostly of the helices resulting from about twothird of the molecule (66%) exhibiting helical structure.34 As stated in the results, the fact that is common in all the interactions of BSA whether native or disturbed with the GM1 micelle is that the tryptophan residue moves to a more shielded environment in the presence of the GM1 micelle. However, the

980 Biomacromolecules, Vol. 9, No. 3, 2008

Gayen et al.

Table 3. Quantitative Analysis of the Change of Secondary Structural Content in BSA in Presence of the GM1 Micelle Using CDSSTR Software with Basis Set SP43 for Native and SDP48 for Denatured Protein secondary structure (%) environment

helix

sheet

turn

unstructured

RMSD

R2

[θ222]

native BSA native BSA:GM1 (1:24) native BSA:GM1 (1:100) in 8 M urea BSA in 8 M urea BSA:GM1 (1:50) in 8 M urea BSA:GM1 (1:160) at 65 °C BSA at 65 °C BSA:GM1 (1:60) at 65 °C BSA:GM1 (1:120)

66.8 59.8 58.7 21.1 58.2 63.4 43.8 42.7 37.5

7.7 6.5 8.4 18.8 14.5 7.5 23.5 13.1 13.4

9.5 12.5 12.6 16.4 11.6 11.8 16.9 17.9 18.9

16.6 21.6 20.5 43.8 15.5 17.3 25.8 26.4 30.3

0.119 0.059 0.071 0.103 0.115 0.104 0.094 0.086 0.094

0.916 0.907 0.899 0.489 0.655 0.727 0.813 0.818 0.803

-2.46 -2.16 -2.16 -1.98 -1.5 -0.67 -1.67 -1.60 -1.36

extent to which the micelle shields the Trp residue differs from the native to chemically and thermally disrupted BSA. There are two tryptophans in BSA molecule: Trp134 is in domain I in the eighth helix of D129-R144, which is well exposed, and Trp213 is in domain II, in the second helix of E206-F221, and is buried.35 The changes in the fluorescence parameters (Figures 2-5) show that the environment of tryptophan is highly affected during the interaction with the GM1 micelle. The blue-shift of λmax, the decrease in KSV, the increase in anisotropy, and the nature of the REES curve indicate that the strongest interaction

Figure 7. Change in hydrodynamic radius, Rh (9) for 5 µM BSA in 10 mM neutral Tris buffer with increasing concentration of GM1 where the plots represent (a) native, (b) chemically disrupted, and (c) thermally disrupted conditions. The data points shown here are the means of at least two independent measurements. All other conditions regarding the experiment are stated in the Materials and Methods sections.

is that between the urea-denatured BSA and the GM1 micelle and the weakest interaction takes place between the thermally disrupted BSA and the GM1 micelle.26 Action of GM1 on Native BSA. As observed from the number distribution plot (Figure 7a) obtained from dynamic light scattering, there is an indefinite increase in the size of native BSA in the presence of GM1 forming the BSA-GM1 complex, which is supported by eq 1 (data not shown) discussed in the Materials and Methods sections. With this equation, the number of GM1 molecules increases indefinitely with increase of GM1/ BSA in the medium. At a molar ratio of protein to lipid 1:100, the emission maximum exhibits no further blue-shift and no lower intensity (Figure 2a), which suggests that the environment of the tryptophan is almost unchanged. From fluorescence measurements (at the above-mentioned BSA/GM1 ratio), it is calculated that about 100 GM1 molecules are bound per BSA molecule. The increase in size of the protein-GM1 complex might be either due to the bound GM1 molecules to the protein or the effect of altered protein conformation or the protein aggregation in presence of GM1. The intensity distribution graph (Figure 8a) against the diameter of the globules in the native BSA-GM1 solution shows that the hydrodynamic diameter increases almost twice that of the native BSA (the component with smaller dH) on addition of GM1 in the molar ratio GM1/ BSA 100:1. Although the diameter of the particles here in solution match well with the hydrodynamic diameter of the pure GM1 micelles (11.6 nm),16 fluorescence and far-UV CD results discard the possibility of the protein to be in fully aggregated form. Hence, the globule with 11.6 nm dH found in solution in the presence of a GM1 micelle (Figure 7a) might be either the BSA-GM1 complex or the protein–protein complex in stoichiometric composition 1:1. The aggregate in the native BSA in solution also increases in size on addition of GM1, as seen in Figure 8a. According to the fluorescence data using eq 1, about 100 GM1 molecules that were considered to be present in the native BSA-GM1 complex at the saturating concentration can lead to an increase in mass of approximately 236% when compared to the native BSA. It is known that, for globular protein of spherical shape, hydrodynamic radius changes with MW1/3. Hence, hydrodynamic radius is expected to increase by approximately 6.18% only, which is much less than the observed 52% increase in Rh (∼1.8 nm) here. Therefore, the observed increase in size of the BSA results not only because of the binding of GM1 molecules to the BSA surface.32 Rather, it could be due to the small change in protein conformation upon interaction with GM1, a fact that is also supported from the fluorescence and CD spectra (Figure 6a). The far-UV CD results suggest that there may be a partial unfolding of BSA in the presence of the GM1 micelle, which was reflected in decreasing R-helicity and the small change in the parameter R2. It was

GM1-Induced Structural Changes of BSA

Biomacromolecules, Vol. 9, No. 3, 2008 981

Figure 8. The intensity distribution graphs showing the various BSA-GM1 complexes (with 500 µM GM1) shown in bottom panel in solutions of (a) native, (b) chemically disrupted, and (c) thermally disrupted BSA in 10 mM Tris-HCl buffer at pH 7.0. The Y axis represents the relative intensity of the scattered light, and the X axis denotes the size classes present in the solution.

observed that here, GM1 induces partially unfolded states of protein in the protein-GM1 complexes that still possess high R-helicity monitored by two well-defined minima at 222 nm and between 202 and 210 nm. Action of GM1 on Urea-Denatured BSA. In comparison, in the presence of GM1, the urea-denatured BSA showed a large increase in the ellipticity in the far-UV CD spectral region (Figure 6b). With 250 µM GM1 (BSA:GM1 1:50), the ureadenatured BSA showed a 58% helix content as determined by the software CDSSTR, and when 800 µM GM1 (BSA:GM1 1:160) was added to the protein solution, the helicity increased to 63%. However, the Cl- ion that was used in the 10 mM Tris-HCl buffer is one of the ions that should be avoided because of high absorbance in the far-UV spectral region. So, we have rechecked the ellipticity in 20 mM phosphate buffer with BSA: GM1 molar ratio 1:100 and observed satisfying similar results indicating increasing R-helicity. The absorbance of GM1 in the far-UV spectral region was also tested, and GM1 was found to be CD silent in the region 210–240 nm (Figure S4 of the Supporting Information). We also found a large change in fluorescence parameters such as intensity, emission maximum, anisotropy, and REES, which suggests a structure formation in the urea-denatured BSA in presence of the GM1 micelle. As found from DLS, the Rh decreases from 46.13 to 4.27 nm (from number distribution Figure 7b) when GM1:BSA molar ratio is 100:1, which indicates a probable process of folding. Although the number distribution graph (Figure 7b) displays monodisperisity in the solution, the intensity distribution graph (Figure 8b) shows polydispersity even in 8 M urea solution without GM1 containing only the denatured protein. Although urea should solubilize all the aggregates stabilized by hydrogen and hydrophobic bonds and it is hard to imagine any significant population of protein aggregates, yet the data presented here show possible evidence of aggregates with dsph 118.02 nm (Figure 8b). According to Cavagnero et al.,36 two possible explanations can elucidate this unusual experimental fact. If the self-aggregates formed by a particular protein are extremely thermodynamically stable over the urea-stabilized unfolded state, an overall larger population of aggregate can persist in solution

at equilibrium in presence of a high concentration of urea. Alternatively, if the self-aggregates that are present in the native condition are very compact so that urea is not able to penetrate it, then even at high concentration of urea (8 M), the aggregates may still persist in solution. In our study, however, the first explanation seems to be true because the aggregates in native condition were found to be smaller than the aggregates present in the 8 M urea solution. Taking into account the observations by Cavagnero et al.36 on R-helical globular protein apomyoglobin in 8 M urea, which is quite similar to our findings, a conformational equilibrium between the unfolded and aggregated forms of protein in the presence of 8 M urea is suggested here. In presence of GM1, another peak of diameter ∼5 µm appears in the intensity distribution graph, and this is also similar to the findings by Cavagnero et al.,36 where upon refolding apomyoglobin, very large aggregates with hydrodynamic diameter of ∼2 µm were found in solution along with the partially refolded and self-associated aggregates. So, there may be an equilibrium among loosely structured, aggregated, and extremely large aggregated forms of protein in presence of GM1 (Figure 8b bottom panel). The formation of heavily aggregated species observed here has already been interpreted as due to noncovalent rearrangement of preexisting large species, which may here form a complex with GM1 (Figure 8b).36 From the far-UV CD (Figure 6b), it can be suggested that GM1 induces some degree of structure to the urea-denatured BSA. However, the structure of the protein in 8 M urea in the presence of GM1 micelle had no tertiary elements, as can be seen in Figure S3 of the Supporting Information. This kind of protein structure, where some degree of secondary structure is present but the tertiary structure signals are absent, can be considered as a “molten globule state” or “incorrectly folded state”. The mean lifetime of the denatured protein in the presence of GM1 micelle increased from 3.87 to 4.119 ns, whereas the average lifetime of native BSA was found to be 6.119 ns (Table 2). This time-resolved fluorescence result also suggests that the environment in the presence of GM1 around the fluorophore somehow experiences a restriction, but the situation is not the same as the situation of the native protein. In 1948, it was found

982 Biomacromolecules, Vol. 9, No. 3, 2008

by Duggan and Luck that addition of SDS prevents the viscosity rise of serum albumin in urea solution.37 Then Moriyama and Takeda in 1993 studied the refolding effect of SDS on helices of urea-denatured BSA by applying a detailed CD study.38 It was found by them that the helicity that was lost in 8 M urea mostly recovered by 40% at 4–10 mM concentration of SDS. Our experimental results suggest that GM1 can be a better helicity-recovering agent for the urea-denatured BSA than SDS when 8 M urea is the denaturizing agent. Action of GM1 on Thermally Disrupted BSA. The heatinduced conformational change of BSA is a complex process: BSA does not denature below 40 °C, around 57 °C, the shortsegment chains connecting the R-helical segments get collapsed, above 75 °C, the R-helix and turn structures get cooperatively denatured with formation of intermolecular β-sheet structure, and the disordered structures become much stronger with further rise in temperature.39 Still, we have chosen 65 °C as the denaturizing temperature because of pathogenic interest. The possible cytotoxicity of the early stage oligomers of the nondiseaseassociated proteins has added interest in the study of the heat treatment of BSA at this temperature because small-scale oligomers are already reported to be formed around 65 °C.30 Prior to the spectral studies done on BSA in the presence of GM1 at 65 °C, the thermal denaturation of the protein itself at 65 °C was examined in detail. After a long incubation of the protein at 65 °C, the conformational change of BSA was found to be irreversible.17–19,40 It is reported by Moriyama et al. that, after heating the BSA sample at 65 °C, when the temperature is cooled down from 65 to 25 °C, the helicity does not attain the original value before heating up to 65 °C but increases only to 53%.40 In our experiments with 5 µM protein in 10 mM TrisHCl buffer, cooling of the protein solution after incubation for the experimental time at 80 °C led to precipitation and the solution became turbid. On the other hand, after heating the protein in buffer at 65 °C for the experimental time, the solution was clear and the irreversibility in the structure of the protein, which was our interest, was achieved.40 It was reported by Moriyama et al. that the helicity that decreases to 44% at 65 °C recovers incompletely upon the decrease of temperature from 65 to 25 °C. In their article, it was proved that SDS as monomer can protect 58% helicity at 65 °C, and after cooling to 25 °C, the protected helicity increases to 64%.40 So, it was interesting to see whether GM1 has any induced effect on BSA similar to SDS or not. The process of thermal denaturation was adopted from Valeria Militello et al.17 Unlike urea, a blue-shift was produced by BSA (Figure 2c) when thermally denatured at 65 °C. The Stern-Volmer quenching constant as determined from the Stern-Volmer equation was 6.48 M-1 for BSA at 65 °C (Table 1) and the corresponding anisotropy as measured was 0.063 at 65 °C (Figure 4). The emission maximum shifted from 340.5 to 332 nm in the presence of the GM1 micelle. The KSV further decreased to 5.41 M-1 and the anisotropy increased to 0.091. However, the extents of change in the magnitudes of the fluorescence parameters show that Trp residues are located in a less-restricted environment in the thermally disrupted state when compared to the other two situations discussed. The relative nature of the REES curve, too, suggests that interaction with GM1 has a minimal effect on the thermally disrupted BSA. The far-UV CD spectra in the absence and presence of two different concentrations of GM1 (Figure 6c) were recorded at 65 °C. GM1, however, does not increase the helicity at all in the thermally treated BSA although it has pronounced folding effect on the urea-denatured BSA. The CD spectra clearly indicate that R-helicity decreases on subsequent addition of GM1

Gayen et al.

to the protein solution. In the absence of GM1, 44% helicity was intact at 65 °C. Then at a molar ratio of BSA/GM1 1/60 ratio of protein to lipid, the helicity was found to be 43% from CDSSTR. The helicity further decreased to 38% with increased ratio of GM1/BSA. It was previously reported by Moriyama et al., that 44% helicity of BSA can be retained only when the helices in the large loops remain intact and the other helices at the C terminal and at the connecting segments between the domains and between the loops are disrupted.40 So, 38% helicity that was finally achieved suggests that helices in the large loops might be getting disrupted at the 65 °C with high concentrations of GM1 micelle (BSA:GM1 1:60). The DLS study also revealed a very small increase in size of the particles in the intensity distribution plot and the number distribution plot with thermally disrupted BSA when complexed with GM1. The thermal BSA-GM1 complex had maximum dH 4.86 nm in the number distribution (Figure 7c), i.e., only a 43% increase in hydrodynamic radius has occurred. With added 500 µM GM1, the dH of the small size component due to the protein increased to 11.28 nm from 9.14 nm and the diameter of the large size component due to the oligomer increased from 43.96 to 58.08 nm only in the intensity distribution (Figure 8c). No extremely large aggregate was found during the course of thermal denaturation, which suggests that no precipitation occurred during the course of the experiment. How GM1 Interacts: A Suggestion. The difference in the way of interaction of the GM1 micelle with the albumin molecule is not surprising because it is expected that the protein-surfactant interaction will be strongly influenced by the amphiphilic nature of the ionic surfactant, that is, the hydrophobic chain length and the charge on the hydrophilic group.41 GM1 is a double-chain surfactant micelle having a very low cmc (3 × 10-8 to 3 × 10-6 M)16 and a strong hydrophobic microdomain, but it also has a pentasaccharide part with a sialic acid and therefore the head group is strongly hydrophilic. It is believed that, due to the repulsion with net negative charge on the protein surface, the GM1 micelle is disintegrated in small micelles in the presence of bovine serum albumin in solution, where the head group of the GM1 micelles can interact with the protein, resulting ultimately in perturbation of the interchain hydrophobic interaction of the protein, making it partially unfolded.8,9 This effect of charged head group would be stronger for the thermally treated BSA when the aggregation number of GM1 becomes too low due to the increase in temperature as well as the presence of the protein with conformation intact in the large loops.42,43 It is the cause why GM1 has no protective effect on the helical structure of the thermally disrupted BSA. As reviewed by Keiderling et al. and Deep et al., in the thermal denaturation process, the small anionic micelles may occupy the hydrophilic binding sites for polar anionic surfactants on the polypeptide chain of protein, specifically to the lysyl, histydyl, and arginyl amino acid side chains, allowing hydrophobic sites to be exposed.44,45 These reports support the small increase in hydrodynamic diameter in dynamic light scattering along with small changes in fluorescence parameters in our experiments, suggesting the idea that the main contributing interaction may be a site-specific local interaction between the small GM1 micelles and the thermally treated BSA. But, for the urea-denatured BSA, the GM1 micelle is not disintegrated because the protein is now almost unfolded, with helices disrupted including those in the large loops and therefore because of the strong hydrophobic domain due to the two hydrophobic chains, the GM1 micelle exhibits more protection to the helical structure than other single-chain surfactants like

GM1-Induced Structural Changes of BSA

SDS, SDes, STS, etc.40,41 As soon as the urea-denatured protein molecule is considered to be arrested within the hydrophobic core of GM1, the unfolding process is prevented and helicity increases suddenly, leading to a drop in hydrodynamic radius. The large changes in the fluorescence parameters too indicate the probable structure formation in the unfolded protein when titrated with the GM1 micelle. The complex-formation processes in solution are clearly evidenced for the chemically (8 M urea) and thermally disrupted (at 65 °C) BSA, as the steady-state fluorescence spectra in parts b and c of Figure 2 suggest. All the GM1 present in solution were found to be bound with BSA forming the BSA-GM1 complexes, as evidenced from the intensity distribution graphs of Figure 8.

5. Conclusion Although majority of the gangliosides reside in the outer leaflet of the cell surface, due to the strong amphiphillic character of GM1, there is always equilibrium between plasma membrane molecules and free monomers present in the extracellular environment like blood serum and other body fluids. The amount of these free monomers in the extracellular solution increases when a favorable equilibrium leads to the formation of stable protein-bound complexes.46 Because a significant amount of gangliosides is detectable in serum, the interaction between serum albumin and GM1 as reported here by us is relevant for the living organisms. Again, sphingolipid and cholesterol-enriched domains or rafts in the cell membrane are involved in many membrane-associated events like cell signaling, cell adhesion, and protein sorting, which further justifies choice of our system, where ganglioside GM1 micelle behaves as a membrane mimic and interacts with the protein. So, this can be a model system for the interaction between gangliosideenriched rafts and protein–ligands. Acknowledgment. This work was supported by the Council for Scientific and Industrial Research, Government of India (no. 01(1790)/02/EMR-II) and University Grant Commission (UGC/ 320/Fellow (University)). We acknowledge the instrumental facility of Saha Institute of Nuclear Physics, Biophysics Division, for CD and DLS experiments, and the Instrumental Facility of Satyendra Nath Bose National Centre for Basic Sciences for the DLS experiments. Supporting Information Available. Stern-Volmer plots using acrylamide as external quencher, the far-UV CD and the near-UV CD spectra of urea-denatured BSA in neutral phosphate (20 mM) buffer, and the far-UV CD spectra of control GM1 micelle. This material is available free of charge via the Internet at http://www.pubs.acs.org.

References and Notes (1) Ledeen, R. W.; Yu, R. K. Methods Enzymol. 1982, 83, 139–191. (2) Revesz, T.; Greaves, M. Nature (London) 1975, 257, 103–106. (3) Hayashi, K.; Iwase, H.; Arai, S.; Takizawa, T.; Hirai, M. Biophys. J. 1998, 74, 1380–1387. (4) Katagiri, A.; Hayashi, K. Biochim. Biophys. Acta 1974, 337, 107– 117. (5) Abreu, M. S. C.; Estronoca, L. M. B. B.; Moreno, M. J.; Vaz, W. L. C. Biophys. J. 2003, 84, 386–389. (6) Gelamo, E. L.; Tabak, M. Spectrochim. Acta, Part A 2000, 56, 2255– 2271.

Biomacromolecules, Vol. 9, No. 3, 2008 983 (7) Turnbull, W. B.; Precious, B. L. J. Am. Chem. Soc. 2004, 126, 1047– 1054. (8) Takizawa, T.; Hirai, M.; Yabuki, S.; Nakata, Y.; Takahashi, K. Thermochim. Acta 1995, 267, 355–364. (9) Hirai, M.; Takizawa, T.; Yabuki, S.; Nakata, Y.; Mitomo, H.; Hirai, T.; Shimizu, S.; Furusaka, M.; Kobayashi, K.; Hayashi, K. Physica B 1995, 214, 751–753. (10) Khalil, M. B.; Kates, M.; Carrier, D. Biochemistry 2000, 39, 2980– 2988. (11) Martinez, Z.; Zhu, M.; Han, S.; Fink, A. L. Biochemistry 2007, 46, 1868–1877. (12) Tomasi, M.; Roda, L. G.; Aussiello, C.; D’Agnolo, G.; Venerandp, D.; Ghidoni, R.; Sonnino, S.; Tettamanti, G. Eur. J. Biochem. 1980, 111, 315–324. (13) Chatterjee, C.; Mukhopadhyay, C. Biochem. Biophys. Res. Commun. 2002, 292, 579–585. (14) Chatterjee, C.; Majumder, B.; Mukhopadhyay, C. J. Phys. Chem. B 2004, 108, 7430–7436. (15) Chatterjee, C.; Mukhopadhyay, C. Biochem. Biophys. Res. Commun. 2004, 315, 866–871. (16) Sonnino, S.; Cantu, L.; Corti, M.; Acquotti, D.; Venerando, B. Chem. Phys. Lipids 1994, 71, 21–45. (17) Valeria, M.; Vetri, V.; Leone, M. Biophys. Chem. 2003, 105, 133– 141. (18) Kosa, T.; Maruyama, T.; Otagiri, M. Pharm. Res. 1998, 15, 449–454. (19) Biagio, S. P. L.; Bulone, D. Biophys. J. 1996, 70, 494–499. (20) De, S.; Girigoswami, A.; Das, S. J. Colloid Interface Sci. 2005, 285, 562–573. (21) Forood, B.; Feliciano, E. J.; Nambiar, K. P. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 838–842. (22) Scholtz, J. M.; York, E. J.; Stewart, J. M.; Baldwin, R. L. Biopolymers 1991, 31, 1463–1470. (23) Rahaman, A.; Srinivasan, N. Biochemistry 2003, 42, 922–931. (24) Sreerama, N.; Woody, R. W. Anal. Biochem. 2000, 282, 252–260. (25) Kallick, D. A.; Tessmer, M. R.; Watts, C. R.; Li, C. J. Magn. Reson., Ser. B. 1995, 106, 60–65. (26) Raghuraman, H.; Chattopadhyay, A. Eur. Biophys. J. 2004, 33, 611– 622. (27) Takeda, K.; Ssaoka, H.; Sasa, S.; Hirai, H.; Hachiya, K.; Moriyama, Y. J. Colloid Interface Sci. 1992, 154, 385–392. (28) Oh, Y. S.; Johnson, J. C. S. J. Chem. Phys. 1981, 74, 2717–2720. (29) Wetzel, R.; Becker, M.; Behlke, J.; Billwitz, H.; Bohm, S.; Ebert, B.; Hamann, H.; Krumiegel, J.; Lassmann, G. Eur. J. Biochem. 1980, 104, 469–478. (30) Vaina, S. M.; Emanuele, A.; Palma-Vittorelli, M. B.; Palma, M. U. Proteins: Struct., Funct., Bioinform. 2004, 55, 1053–1062. (31) Peters, T., Jr. Serum albumin. In AdVances in Protein Chemistry; Academic Press: New York, 1985; Vol. 37, pp 161–245. (32) Lee, C. T.; Smith, K. A. Biochemistry 2005, 44, 524–536. (33) Brown, J. R.; Rosenoer, V. M.; Ortaz, M.; Rothschild, M. A. Albumin: Structure, Function, and Uses; Pergamon: Oxford, U.K, 1977; pp 27– 51. (34) Takeda, K.; Shigeta, M.; Aoki, K. J. Colloid Interface Sci. 1987, 117, 120–126. (35) Sun, C.; Yang, J. Biophys. J. 2005, 88, 3518–3524. (36) Chow, C.; Kurt, N.; Murphy, R. M.; Cavagnero, S. Biophys. J. 2006, 90, 289–309. (37) Duggan, E. L.; Luck, J. M J. Biol. Chem. 1948, 172, 205–220. (38) Moriyama, Y.; Sato, Y.; Takeda, K. J. Colloid Interface Sci. 1993, 117, 420–424. (39) Murayama, K.; Tomida, M. Biochemistry. 2004, 43, 11526–11532. (40) Moriyama, Y.; Takeda, K. Langmuir 2005, 21, 5524–5528. (41) Wu, D.; Xu, G.; Sun, Y.; Zhang, H.; Mao, H.; Feng, Y. Biomacromolecules 2007, 8, 708–712. (42) Cantu, L.; Corti, M.; Sonnino, S.; Tettamanti, G. Chem. Phys. Lipids 1986, 41, 315–328. (43) Cantu, L.; Corti, M; Favero, E. D.; Raudino, A. J. Phys.: Condens. Matter 1997, 9, 5033–5055. (44) Keiderling, A. T.; Xu, Q. I. Protein Sci. 2004, 13, 2949–2959. (45) Deep, S.; Ahluwalia, J. C. Phys. Chem. Chem. Phys. 2001, 3, 4583– 4591. (46) Chigorno, V.; Giannotta, C.; Ottico, E.; Sciannamblo, M.; Mikulak, J.; Prinetti, A.; Sonnino, S. J. Biochem. 2005, 280, 2668–2675.

BM701144K