Langmuir 2008, 24, 4989-4998 4989

Characterization of Secondary and Tertiary Conformational Changes

of β-Lactoglobulin Adsorbed on Silica Nanoparticle Surfaces
Xiaoyu Wu and Ganesan Narsimhan*
Biochemical and Food Process Engineering, Department of Agricultural and Biological Engineering,
Purdue UniVersity, West Lafayette, Indiana 47907

ReceiVed October 26, 2007. In Final Form: January 11, 2008

Nanoparticles possess unique properties as a result of their large surface area per unit volume and therefore can
be functionalized by the immobilization of enzymes for a variety of biosensing applications. Changes in the tertiary
conformation of β-lactoglobulin adsorbed on 90 nm silica nanoparticles with time were inferred using tryptophan
fluorescence and Fourier transform infrared spectroscopy (FTIR) for different surface concentrations, temperature,
pH, ionic strength, and 2,2,2-trifluoroethanol (TFE) and dithiothreitol (DTT) concentrations. Rapid initial unfolding
followed by a much slower rate at longer times was observed, with the extent of unfolding being higher at lower surface
concentrations, higher ionic strengths, higher temperature, higher TFE and DTT concentrations, and pI. The effect
of temperature on the unfolding of adsorbed protein on the nanoparticle surface was similar to that in the bulk even
though the extent of unfolding was higher for adsorbed protein molecules. The results of the extent of change in tertiary
conformation using FTIR as indicated by the change in the ratio of amide II′/amide I were consistent with those
obtained by tryptophan fluorescence whereas the rates of conformational changes given by FTIR were found to be
much faster. Circular dichroism (CD) spectra showed that altering the surface concentration by itself did not change
the secondary structure of β-lactoglobulin on the surface. TFE was found to increase the R helix content at the expense
of the fraction of the β sheet, whereas the β sheet was converted to an unordered conformation in the presence of
DTT.

Introduction of protein or drug molecules. The large interfacial area might

Proteins tend to adsorb at interfaces because they contain both
hydrophobic and hydrophilic functional groups. Upon adsorption,
protein molecules may unfold or may interact with each other
or with the support surfaces to form a new system that provides
various avenues of applications in food processing, biotechnology,
and medical technology.1-3 The interaction between proteins
and solid interfaces is exploited in a variety of applications
such as proteins (enzymes) immobilized onto bioreactors or
biosensors.4-6 The most commonly employed techniques for
immobilizing proteins on solid surfaces are entrapment,7 adsorp-
tion, and covalent bonding.8,9 Among these, physical adsorption
is the simplest immobilization method and requires the mildest
experimental conditions for obtaining an enzyme-solid complex.
Moreover, another advantage of physical adsorption compared
to incorporating proteins within a matrix or covalent bonding is
the reduction in diffusional resistance, which results in a much
faster response and provides an ideal system for kinetic study.
The use of nanoparticles as a solid support for adsorption is
attractive because of the extremely large interfacial area per unit
volume, which may make nanoparticles more efficient carriers
* To whom correspondence should be addressed. Tel: (765) 494-1199.
Fax: (765) 496-1115. E-mail: narsimha@purdue.edu.
(1) Wang, A. J.; Arnold, M. A. Anal. Chem. 1992, 64, 1051-1055.
(2) Vega, C.; Roos, Y. H. J. Dairy Sci. 2006, 89, 383-401.
(3) Field, S.; Udalova, I.; Ragoussis, J. AdV. Biochem. Eng. Biotechnol. 2007,
104, 87-110.
(4) Zhao, C.; Jiang, H.; Smith, D. R.; Bruckenstein, S.; Wood, T. D. Anal.
Biochem. 2006, 359, 167-175.
(5) Qu, H.; Wang, H.; Huang, Y.; Zhong, W.; Lu, H.; Kong, J.; Yang, P.; Liu,
B. Anal. Chem. 2004, 76, 6426-6433.
(6) Meij, P.; Vervoort, M. B.; de Gooijer, K.; Bloemena, E.; Meijer, C. J.;
Middeldorp, J. M. Protein Expr. Purif. 2000, 20, 324-333.
(7) Janes, K. A.; Calvo, P.; Alonso, M. J. AdV. Drug DeliV. ReV. 2001, 47,
83-97.
(8) Falb, R. D.; Grode, G. A. Fed. Proc. 1971, 30, 1688-1697.
(9) Chevolot, Y.; Martins, J.; Milosevic, N.; Leonard, D.; Zeng, S.; Malissard,
M.; Berger, E. G.; Maier, P.; Mathieu, H. J.; Crout, D. H.; Sigrist, H. Bioorg. Med.
Chem. 2001, 9, 2943-2953.
also make it possible to miniaturize biosensors and bioreactors
for microarrays for the identification of biomolecules. Several
reports have revealed the successful use of nanoparticles as solid
supports that did not lead to much loss in the activity of the
absorbed enzyme. For example, silica nanoparticles have been
biochemically functionalized with enzymes for cell membrane
staining and other applications.10-12 The secondary conformation
of human carbonic anhydrase enzymes (wild type and mutants)
adsorbed on silica nanoparticle as monitored using circular
dichroism has been shown to exhibit surface-induced unfolding
with features similar to those of the molten globule intermediate.10
Studies on secondary conformational changes of proteins adsorbed
on silica nanoparticle13 and Teflon particles14 indicated a strong
relationship between conformational changes and affinity to the
surface and a decrease in the R helix with a corresponding increase
in the unordered structure, respectively. Although many studies
have revealed the behaviors of proteins on flat surfaces, including
the rate of adsorption, conformational change, loss of enzymatic
activity and interactions,15-18 the behavior of proteins, especially
the unfolding kinetics on the nanoparticle surfaces, is not known
and may be different from that at a flat surface because of the
high curvature (small radius) of the nanoparticle.
(10) Billsten, P.; Freskgard, P. O.; Carlsson, U.; Jonsson, B. H.; Elwing, H.
FEBS Lett. 1997, 402, 67-72.
(11) Kneuer, C.; Sameti, M.; Haltner, E. G.; Schiestel, T.; Schirra, H.; Schmidt,
H.; Lehr, C. M. Int. J. Pharm. 2000, 196, 257-261.
(12) Qhobosheane, M.; Santra, S.; Zhang, P.; Tan, W. Analyst 2001, 126,
1274-1278.
(13) Kondo, A.; Mihara, J. J. Colloid Interface Sci. 1996, 177, 214-221.
(14) Mollmann, S. H.; Jorgensen, L.; Bukrinsky, J. T.; Elofsson, U.; Norde,
W.; Frokjaer, S. Eur. J. Pharm. Sci. 2006, 27, 194-204.
(15) Xie, Q.; Xiang, C.; Yuan, Y.; Zhang, Y.; Nie, L.; Yao, S. J. Colloid
Interface Sci. 2003, 262, 107-115.
(16) Krivov, S. V.; Karplus, M. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 14766-
14770.
(17) Agnihotri, A.; Siedlecki, C. A. Langmuir 2004, 20, 8846-8852.
(18) Pereira, L. G. C.; Hickel, A.; Radke, C. J.; Blanch, H. W. Biotechnol.
Bioeng. 2002, 78, 595-605.

10.1021/la703349c CCC: $40.75 © 2008 American Chemical Society

Published on Web 03/27/2008
4990 Langmuir, Vol. 24, No. 9, 2008
Table 1. Surface Concentration of β-Lactoglobulin on Nanoparticle Surfaces under Different Conditions
pH 7 pH 5
conc of Np
(w/w) 30 60 240 30 60
0.75% 0.83 0.83 0.84 0.86 0.86
1.5% 0.46 0.45 0.46 0.48 0.46
3.75% 0.18 0.18 0.18 0.18 0.17
The extent of unfolding of globular proteins depends on various
factors such as surface pressure, surface potential, and hydro-
phobicity of the surface. Because of the lack of tertiary structure,
a highly flexible protein such as β-casein adsorbed much faster
and spread at the interface to a greater extent than did a very
rigid, compact protein such as lysozyme.19,20 Because the latter
with its tertiary structure has to undergo conformational changes
in order to spread at the interface, the extent of unfolding of
globular proteins and enzymes at an interface depends on the
available surface area and the surface concentration.21This is
because unfolding takes time, and before it can happen, the surface
may already be fully covered with other proteins/enzymes. In
such a case, there is little driving force for unfolding because the
interfacial free energy cannot be lowered any further.22 The rate
of conformational change depends on the energy barrier to such
changes, which may be influenced by steric and electrostatic
interactions of the neighboring protein molecules.23,24 The
adsorption of seven structural intermediates of bovine serum
albumin25 and the stability mutants of the T4 lysozyme26 indicated
that the extent of unfolding and the area occupied at the interface
depended on the conformational stability of the protein molecule.
Tryptophan fluorescence was employed27 to monitor the change
in the tertiary conformation of β lactoglobulin. Fourier transform
infrared spectroscopy (FTIR)28-30 and Circular dichroism
(CD)10,21 were employed to monitor the secondary structure of
adsorbed proteins and enzymes. The change in the secondary
structure of R-chymotrypsin was found to be more pronounced
when it was adsorbed onto a hydrophobic surface29 and was
pH-dependent28 as a result of strong hydrophobic and electrostatic
interactions between the enzyme and the solid surface. A greater
loss of R helix content of the lysozyme was observed21 upon
adsorption on larger silica nanoparticles as well as for larger
surface coverage. Human carbonic anhydrase was shown to attain
a molten globulelike state after interaction with silica nanopar-
ticles.10 A two-phase sequential dynamic change in the secondary
structure of lysozyme adsorbed onto a solid substrate was
observed,30 with the first phase involving the fast conversion of
the R helix to the β sheet within a few minutes and the second
phase involving a much slower conversion. The kinetics of
(19) Graham, D. E.; Phillips, M. C. J. Colloid Interface Sci. 1979, 79, 403-
414.
(20) Graham, D. E.; Phillips, M. C. J. Colloid Interface Sci. 1979, 79, 415-
426.
(21) Vertegel, A. A.; Siegel, R. W.; Dordick, J. S. Langmuir 2004, 20, 6800-
6807.
(22) Walstra, P.; Deroos, A. L. Food ReV. Int. 1993, 9, 503-525.
(23) Arai, T.; Norde, W. Colloids Surf. 1990, 51, 1-15.
(24) Arai, T.; Norde, W. Colloids Surf. 1990, 51, 17-28.
(25) Damodaran, S.; Song, K. B. Biochim. Biophys. Acta 1988, 954, 253-264.
(26) Wang, J.; McGuire, J. J. Colloid Interface Sci. 1997, 185, 317-323.
(27) Sharma, V. K.; Kalonia, D. S. J. Pharm.l Sci. 2003, 92, 890-899.
(28) Baron, M. H.; Revault, M.; Servagent-Noinville, S.; Abadie, J.; Qui-
quampoix, H. J. Colloid Interface Sci. 1999, 214, 319-332.
(29) Noinville, S.; Revault, M.; Baron, M. H. Biopolymers 2002, 67, 323-
326.
(30) Sethuraman, A.; Vedantham, G.; Imoto, T.; Przybycien, T.; Belfort, G.
Proteins: Struct., Funct. Biointerfaces 2004, 56, 669-678.
Wu and Narsimhan
Γ (mg/m2)
salt TFE DTT
time (min)
240 30 60 240 240 240
0.85 0.81 0.82 0.82 0.83 0.85
0.48 0.44 0.44 0.44 0.43 0.48
0.18 0.17 0.18 0.18 0.17 0.17
enzymatic production of benzaldehyde by prunus amylgdalus
hydroxynitrile lyase has been described18 by a mathematical model
that accounts for the loss of enzymatic activity in the first adsorbed
layer due to its unfolding.
Bovine β-lactoglobulin is one of the major components of the
whey of cow’s milk. Its structure has been determined by X-ray
crystallography31-33 and NMR spectroscopy,34,35 showing that
β-lactoglobulin is a globular protein stabilized by two disulfide
bonds, with nine antiparallel β strands and one short and one
long R-helix at the carboxyl terminus. Moreover, the unfolding
kinetics of β-lactoglobulin in solution is well known,36,37 which
also makes it a good candidate for the investigation of unfolding
kinetics on surfaces. During unfolding, the protein molecule has
the ability to form extended hydrophobic patches exposed to
solvent, which can bind to hydrophobic dyes such as 1-anilino-
8-naphthalene sulfonate (ANS).38,39 The changes in the confor-
mation of this protein in response to changes in pH, temperature,
and ionic strengths have also been characterized.40-43
Systematic research on the kinetics of tertiary conformational
changes of proteins/enzymes adsorbed on nanoparticles surfaces,
which are important for developing nanoscale biotechnology,
has not been reported so far. Even though experimental studies
on conformational changes of commercially significant enzymes
and the resulting changes in enzymatic activity are more relevant
for biosensing applications, as a first step, we undertook studies
on well-characterized globular protein such as β-lactoglobulin.
In this article, we report the results of the changes in secondary
and tertiary structure as well as the evolution of tertiary structures
of β-lactoglobulin adsorbed onto nanoparticle surfaces under
different surface concentration, ionic strength, temperature, pH,
and TFE and DTT concentrations.
Materials and Methods
Materials. β-Lactoglobulin was purchased from Sigma. Silica
nanoparticles of 90 nm diameter were purchased from Interfacial
(31) Brownlow, S.; Morais Cabral, J. H.; Cooper, R.; Flower, D. R.; Yewdall,
S. J.; Polikarpov, I.; North, A. C.; Sawyer, L. Structure 1997, 5, 481-495.
(32) Qin, B. Y.; Bewley, M. C.; Creamer, L. K.; Baker, H. M.; Baker, E. N.;
Jameson, G. B. Biochemistry 1998, 37, 14014-23.
(33) Qin, B. Y.; Bewley, M. C.; Creamer, L. K.; Baker, E. N.; Jameson, G.
B. Protein Sci. 1999, 8, 75-83.
(34) Kuwata, K.; Hoshino, M.; Forge, V.; Era, S.; Batt, C. A.; Goto, Y. Protein
Sci. 1999, 8, 2541-2545.
(35) Uhrinova, S.; Smith, M. H.; Jameson, G. B.; Uhrin, D.; Sawyer, L.; Barlow,
P. N. Biochemistry 2000, 39, 3565-3574.
(36) Invernizzi, G.; Grandori, R. Rapid Commun. Mass Spectrom. 2007, 21,
1049-52.
(37) Bhattacharjee, C.; Saha, S.; Biswas, A.; Kundu, M.; Ghosh, L.; Das, K.
P. Protein J. 2005, 24, 27-35.
(38) Dodin, G.; Andrieux, M.; al Kabbani, H. Eur. J. Biochem. 1990, 193,
697-700.
(39) Dufour, E.; Marden, M. C.; Haertle, T. FEBS Lett. 1990, 277, 223-226.
(40) Fink, A. L.; Calciano, L. J.; Goto, Y.; Kurotsu, T.; Palleros, D. R.
Biochemistry 1994, 33, 12504-12511.
(41) Hughson, F. M.; Wright, P. E.; Baldwin, R. L. Science 1990, 249, 1544-
1548.
(42) Tayyab, S.; Ahmad, B.; Kumar, Y.; Khan, M. M. Int. J. Biol. Macromol.
2002, 30, 17-22.
(43) Timasheff, S. N. Methods Mol. Biol. 1995, 40, 253-69.
Conformational Changes of β-Lactoglobulin
Figure 1. Tryptophan fluorescence spectra of β-lactoglobulin
subjected to GdmCl-induced unfolding at neutral pH. (a) Fluorescence
spectra and (b) normalized fluorescence intensity at 380 nm. The
measurements were carried out at low ionic strength: 10 mM
phosphate at pH 7.0 with a protein concentration of 0.5 mg/mL. The
incubation time at high temperature was 8 min, and the cooling time
to room temperature was 24 h. Similar calibration curves were
obtained for other bulk protein concentrations (not shown here).
Dynamics Corporation, Japan. The nanoparticles are monodisperse
with a very small standard deviation (10 nm) in the particle size.
Ultra GdmCl, 1-anilino-8-naphthalene sulfonate (ANS), 2,2,2-
trifluoroethanol (TFE), Dithiothreitol (DTT), deuterium hydrogen,
hydrogen chloride (HCl), and other chemicals were purchased from
Sigma.
Methods. Protein Unfolding in the Bulk. β-Lactoglobulin solution
in 10 mM phosphate buffer at pH 7 was subjected to unfolding either
by heating at different temperatures (25, 45, 60, 70, 80, and 90 °C)
for 8 min or by the addition of different amounts of GdmCl and
allowing the reaction to occur for 2 h. It was found that the
fluorescence spectra of heated β-lactoglobulin solution did not change
upon cooling to room temperature, thus indicating that heat-induced
unfolding of β-lactoglobulin is irreversible. Therefore, the fluores-
cence spectra of all heated samples were monitored after they were
cooled to room temperature.
The unfolded fraction was inferred by comparing the fluorescence
intensities with the fluorescence intensity of samples subjected to
GdmCl-induced unfolding.
Fluorescence Spectra. Tertiary structural changes of β-lactoglo-
bulin on the surfaces of nanoparticles were measured using a
tryptophan fluorescence emission technique and an ANS binding
assay. In the first method, measurements were carried out on a Flux
Station II fluorescence spectrophotometer with a standard 96-well
plate. The excitation wavelength at 288 nm was chosen because
almost all the fluorescence emission signal excited at this wavelength
Langmuir, Vol. 24, No. 9, 2008 4991
Figure 2. Fluorescence intensity at 380 nm in different solutions.
The experimental conditions are the same as in Figure 1.
Figure 3. Effect of surface concentration on the unfolding kinetics
of β-lactoglobulin adsorbed on the surface of silica nanoparticles at
neutral pH and a low concentration of salts. Normalized data of
fluorescence intensity of β-lactoglobulin at 380 nm for different
surface concentrations expressed as the unfolded fraction. The buffer
was 5 mM phosphate saline at pH 7.0.
is derived from tryptophan. The tryptophan emission fluorescence
spectra were collected in the wavelength range of 360 to 500 nm.
The fluorescence intensity at 380 nm, the wavelength for maximum
intensity, was monitored as a function of time to characterize the
unfolding kinetics because it gave the largest difference between the
native and unfolded states of proteins. All data were collected at
25 °C except in the temperature-induced unfolding experiment.
An ANS binding assay was carried out at an excitation wavelength
of 360 nm, and an emission spectrum was collected from 450 to 550
nm. Other conditions were the same as for tryptophan fluorescence
emission.
Fourier Transform Infrared Spectra (FTIR). Fourier transform
infrared (FTIR)-attenuated total reflectance (ATR) spectra were
recorded using a ThermoNicolet Nexus 670 FTIR spectrometer
equipped with a mercury cadmium telluride A (MCTA) detector
and KBr optics cooled with liquid nitrogen. Aliquots were taken at
different times from the sample of β-lactoglobulin solution in the
presence of nanoparticles and mixed with equal volumes of 10 mM
phosphate buffer prepared in 99.9% D2O at pH 7 at room temperature
for 30 min, and FTIR spectra were analyzed to infer the exchange
of deuterium with hydrogen inside the protein molecule. The
spectrophotometer was continuously purged with dry N2 to remove
water vapor. The spectra were obtained after a lag time of 20 s and
were determined by collecting 256 scans per sample at a resolution
of 2 cm-1. The infrared spectrum of buffer was subtracted. The
amide I absorbance maxima of all spectra were normalized to the
same intensity for comparison. Curve-fitting analysis was performed
4992 Langmuir, Vol. 24, No. 9, 2008
Figure 4. Unfolding kinetics of β-lactoglobulin on the surfaces of
the silica nanoparticles at two different pH values. Other experimental
conditions were the same as in Figure 3.
Figure 5. Unfolding kinetics of β-lactoglobulin on the surfaces of
the silica nanoparticles at different salt concentrations. Other
experimental conditions were the same as in Figure 3.
using classical multivariate procedures including discriminant
analysis (DA) and partial least-squares (PLS) analysis with TQ
software to estimate quantitatively the spectral parameters of
overlapping components of the amide I band (around 1650 cm-1)
and amide II′ band (1450 cm-1) in the infrared spectra. Band positions
were locked, but the width and the height were varied to obtain the
best fit.
Circular Dichroism (CD) Spectra. CD spectra were measured
from 190 to 300 nm at room temperature on a Jasco J-810
spectrometer (Jasco Spectroscopic Co., Hachioji, Japan) using a cell
with a path length of 0.1 mm. Data were collected every 0.2 nm with
2 nm bandwidth, at a scan speed of 50 nm/min, by averaging three
scans. Molar ellipticity (deg cm-1 dmol-1) is expressed on a mean
residue concentration basis in the far-UV. Spectra were analyzed for
secondary structure content by using the program Contin.44
Surface Concentration of β-Lactoglobulin. The amount of
β-lactoglobulin adsorbed on the silica nanoparticle surface under
different condition for different exposure times was measured. A
0.5 mg/mL protein solution was mixed with nanoparticles of different
particles concentrations (0.75 to 3.75 wt %). After exposure times
of 30 min, 1 h, and 4 h, this mixture was centrifuged at 4500 rpm
(centrifuge) in a centrifuge tube equipped with a membrane filter
with a 30 000 Da molecular weight cutoff so as to exclude
nanoparticles from the supernatant. The concentration of protein in
the supernatant was inferred from UV absorbance at 290 nm using
a Cole Palmer UV2100 spectrometer.
By material balance, the surface concentration Γ(c) was estimated
from
(44) Lees, J. G.; Miles, A. J.; Wien, F.; Wallace, B. A. Bioinformatics 2006,
22, 1955-1962.
Wu and Narsimhan
a
cf + Γ(cf) ) ci (1)
V
where ci and cf are the initial protein concentration and final protein
concentration after equilibration with nanoparticles, respectively, a
is the total surface area of nanoparticles, and V is the volume of
protein solution.
In eq 1, the total area a of nanoparticles is given by
3wφ
a) (2)
FR
where w is the mass of the sample and φ is the mass fraction of
nanoparticles of radius R and density F.
EValuation of the Unfolded Fraction of Adsorbed β-Lactoglobulin
from Fluorescence Measurements. The normalized fluorescence
intensity Is from the adsorbed protein on the nanoparticle surface
was obtained by subtracting the fluorescence Ib(cb) for the bulk of
protein concentration cb as given by
I - Ib(cb) I - Ib(cb)
Is ) ) (3)
Γa (ci - cb)
where ci and cb refer to the initial and final bulk protein concentrations,
respectively. The unfolded fraction of protein on the nanoparticle
surface was inferred from Is/{IDb (ci - cb) - INb (ci - cb)} (D and N
refer to fully denatured and native protein solutions, respectively)
using the calibration of fluorescence intensities of protein solution
with different extents of denaturation as obtained from GdmCl-
induced denaturation (Results section).
Results
Adsorption Kinetics of β-Lactoglobulin on Nanoparticle
Surfaces under Different Conditions. Table 1 shows the surface
concentration of β-lactoglobulin on nanoparticles under different
conditions for three different mixing times (30, 60, and 240
min). As can be seen from the Table, the surface concentration
did not increase significantly from 30 to 240 min, thereby
indicating that protein adsorption approached equilibrium around
30 min. In addition, the surface concentration of β-lactoglobulin
was found to be insensitive to variations in pH, ionic strength,
and TFE and DTT concentrations. As will be shown later, the
evolution of tertiary structures of adsorbed protein is found to
be much slower than the rate of protein adsorption.
Effect of Surface Concentration, pH, Ionic Strength, and
Temperature on the Unfolding Kinetics of β-Lactoglobulin
on Nanoparticle Surfaces. Typical tryptophan fluorescence
spectra of β-lactoglobulin solution (0.5 mg/mL) in the presence
of different concentrations of GdmCl at pH 7 are shown in Figure
1a. As expected, the fluorescence intensity was found to increase
at higher GdmCl concentrations because of the exposure of more
tryptophan residues as a result of unfolding. Interestingly, there
was no significant change in the spectra up to a GdmCl
concentration of 1.5 M, thus indicating that β-lactoglobulin retains
its native conformation at lower GdmCl concentrations. β-Lac-
toglobulin was fully denatured at a GdmCl concentration above
3.5 M as evidenced by no further change in the fluorescence
spectra at higher concentrations (Figure 1a). Following the
procedure of Sontoro and Bolen,45 the normalized fluorescence
intensity at 380 nm is shown in Figure 1b for different GdmCl
concentrations. For a two-state model, this fractional increase
can be interpreted as the unfolded fraction. As shown in Figure
1b, the unfolded fraction was zero for GdmCl concentrations
below 1.5 M and was one for concentrations above 3.5 M.
Consistent with results reported in the literature, the variation of
(45) Santoro, M. M.; Bolen, D. W. Biochemistry 1988, 27, 8063-8068.
Conformational Changes of β-Lactoglobulin
Figure 6. Unfolding kinetics of β-lactoglobulin on the surfaces of
the silica nanoparticles at different temperatures. Other experimental
conditions were the same as in Figure 3.
the unfolded fraction with GdmCl in the intermediate concentra-
tions is linear. Similar calibration curves were obtained for
different bulk protein concentrations (not shown here). Figure
2 shows the fluorescence intensity at 380 nm for different
solutions. As expected, buffer gave very small fluorescence signal
and was subtracted as a background reference in the data analysis.
It is important to note that the addition of nanoparticles of different
concentrations (in the range of 0.75 to 3.75 wt %) to the buffer
did not result in a significant increase in fluorescence intensity.
The addition of 0.5 mg/mL native β-lactoglobulin resulted in an
increase in fluorescence intensity corresponding to that of protein
containing four tryptophan amino acids. As expected, the addition
of 5 M GdmCl to β-lactoglobulin solution further increased the
fluorescence intensity as a result of the complete denaturation
of protein. Further addition of nanoparticles of different
concentrations did not change the fluorescence intensity, thereby
indicating that nanoparticles did not interfere with tryptophan
fluorescence. Consequently, tryptophan fluorescence can be
employed to monitor the unfolding of β-lactoglobulin on
nanoparticle surfaces. Interestingly, the fluorescence intensity
of adsorbed β-lactoglobulin on nanoparticle surfaces in the
absence of GdmCl was found to fall in between the two levels
of native and fully denatured proteins. The calibration in Figure
1b was employed in the inference of the extent of unfolding of
adsorbed β-lactoglobulin on nanoparticle surfaces from the
normalized fluorescence intensity at 380 nm as explained in the
Methods section.
No quantitative relationship between the extent of protein
unfolding and its available area on the surface has been reported
so far, even though this relationship may be important to the
design of biosensor or bioreactor systems. The evolution of
unfolding of β-lactoglobulin on silica nanoparticle surfaces for
different nanoparticle concentrations is shown in Figure 3. The
surface concentration of β-lactoglobulin on silica nanoparticle
surfaces was calculated from the initial protein bulk concentration
and the nanoparticle surface area as explained in the Methods
section. It is clear from Figure 3 that β-lactoglobulin at low
surface concentration was unfolded to a greater extent in the
equilibrium state and unfolded faster during the conformation
change process compared with β-lactoglobulin at a higher surface
concentration. And the shape of the kinetics curves at different
surface concentrations was quite different. For example, at the
lowest surface concentration of around 0.18 mg/m2, β-lacto-
globulin reached an extremely stretched equilibrium state with
about 80% extent of unfolding after 500 min, whereas at the
highest surface concentration β-lactoglobulin unfolded only about
Langmuir, Vol. 24, No. 9, 2008 4993
Figure 7. (a) Fluorescence spectra of ANS bound β-lactoglobulin
on the surfaces of nanoparticles at different surface concentrations.
(b) Fluorescence spectra of ANS bound β-lactoglobulin on the
surfaces of nanoparticles at different temperatures.
20% after a much longer (1400 min) time. This result also suggests
that at higher surface concentrations the system may not reach
equilibrium even after 1400 min. This finding indicates the
existence of an extremely high energy barrier for unfolding in
a crowed surface environment, which is also evidenced by our
observation that no fluorescence signal change was detected when
β-lactoglobulin was exposed to a much higher surface concen-
tration of 1.66 mg/m2 (data not shown). Such a dramatic difference
in unfolding behavior at different surface concentrations has
interesting implications for the activity of protein/enzyme on the
nanoparticle surfaces. A possible explanation is that at low surface
concentration the predominant molecular interaction is between
proteins and the hydrophobic surfaces of silica nanoparticles,
which induces the unfolding of protein molecule without energy
barriers as a result of the freely available space. However, at
higher surface concentration (i.e., in a crowded environment),
the unfolding behavior of protein molecules is strictly limited
by the interaction between protein molecules (i.e., in such a
crowded environment, the extension of protein structure is limited
by the physical space and the energy barriers that arise from the
interaction with other nearby protein molecules). Apparently,
this energy barrier not only decreases the extent of unfolding but
also decreases the unfolding rate at high surface concentration.
It is also evident from Figure 3 that at lower surface
concentration, there was a rapid increase in the unfolded fraction
at short times, followed by a much slower increase at longer
times. Therefore, one can attribute two phases for protein
unfolding on the surface of nanoparticles. The initial phase occurs
within minutes, which is believed to involve the reorientation as
well as the unfolding of the protein molecule on the surface after
its adsorption. Because of the lower energy barrier due to the
4994 Langmuir, Vol. 24, No. 9, 2008
Figure 8. (a) Spectra of H-D exchange of β-lactoglobulin on
nanoparticle surfaces at different mixing times with D2O. Before
D2O is added to the protein solution, 4 h of incubation is necessary
for protein to equilibrate with nanoparticles. The surface concentration
is 0.64 mg/m2, and the final buffer solution contains 75% D2O. (b)
Effect of the duration of D2O exposure on the unfolding kinetics of
β-lactoglobulin on the nanoparticle surfaces at neutral pH and at a
low concentration of salts expressed as ratio of II′/I for a surface
concentration of 0.4 mg/m2. The buffer was 10 mM phosphate saline
at pH 7.0.
larger surface area available to the protein molecule at lower
surface concentration, such a reorientation process takes a very
short time as indicated by an initial rapid increase in the unfolded
fraction. Because the initial time scale was on the order of minutes,
experimental measurements of unfolding kinetics within that
time scale could not be made. Therefore, the reported experimental
data observed an initial rapid jump in the unfolded fraction. This
initial jump was found to be larger for smaller protein surface
concentration. Figure 3 also shows that as the surface concentra-
tion increased, the extent of the initial jump decreased, reaching
an almost insignificant level at the highest surface concentration.
Similarly, the unfolding rate after the initial jump became smaller
with increasing surface concentration. The slower and smaller
unfolding at higher surface concentrations could be attributed to
protein-protein interactions (crowded environment) at the
nanoparticle surface. At lower surface concentrations, however,
this interaction may become significant only after the adsorbed
protein molecules have unfolded sufficiently, thus increasing
their effective molecular areas (and thereby decreasing the
distance between neighboring molecules).
The evolution of unfolding of β-lactoglobulin on silica
nanoparticle surfaces at two different pH values (7 and 5) is
shown for different surface concentrations in Figure 4. The net
charge of protein molecule at pH 5, near pI, is much smaller than
that at pH 7, indicating that the electrostatic interactions between
adsorbed protein molecules were more pronounced at pH 7, which
may lead to an inhibition of the unfolding of protein molecules.
Such a notion is supported by the data in Figure 4: the extent
Wu and Narsimhan
Figure 9. (a) Spectra of H-D exchange of β-lactoglobulin on
nanoparticle surfaces at different mixing times with nanoparticles.
Before nanoparticles are added to the protein solution, a half hour
of incubation is necessary for protein to equilbrate with D2O. The
surface concentration is 0.83 mg/m2, and the final buffer solution
contains 75% D2O. (b) Effect of surface concentration on the
unfolding kinetics of β-lactoglobulin adsorbed on the surface of
silica nanoparticles at neutral pH and a low concentration of salts
expressed as ratio of II′/I for different surface concentrations.
of unfolding was higher at pH 5 than at pH 7 at lower surface
concentration but not significantly different at higher surface
concentrations. A possible explanation is that electrostatic
interactions between protein molecules that inhibit unfolding
are dramatically different at different pH values at lower surface
concentration, whereas this difference is masked by stronger
protein-protein interaction in a crowded environment at higher
surface concentration.
The evolution of unfolding of β-lactoglobulin on the silica
nanoparticle surface for two different ionic strengths under the
conditions of pH 7 and three different nanoparticle concentrations
is shown in Figure 5. Higher ionic strength was found to induce
a much higher extent of unfolding at higher surface concentration.
In general, at higher ionic strength, the electrical double layer
in the vicinity of the adsorbed protein molecule is compressed,
and as a result, the electrostatic interaction between adsorbed
protein molecules is reduced, thus leading to more unfolding.
Interestingly, such an effect of ionic strength on protein unfolding
on nanoparticle surfaces was not significant at lower surface
concentration, where the average distance between adsorbed
protein molecules is far greater than the interaction range of the
double layers of different protein molecules. Thus a further
compressed double layer caused by increasing the ionic strength
will not reduce the interaction of protein molecules. The diameter
of the sphere of influence of adsorbed proteins can be estimated
to be 1/Γ, where Γ is the surface number concentration of protein
on the nanoparticle surface. Assuming the radius of adsorbed
protein to be that of native rnative (3 nm), one can estimate the
Conformational Changes of β-Lactoglobulin
Figure 10. Effect of TFE concentration on the unfolding kinetics
of β-lactoglobulin adsorbed on the surface of silica nanoparticles at
neutral pH and a low concentration of salts. Normalized data of the
fluorescence intensity of lysozyme at 380 nm at different surface
concentrations expressed as the unfolded fraction. (a) 0.83 mg/m2
and (b) 0.18 mg/m2. The buffer was 10 mM phosphate saline at pH
7.0.
surface-to-surface distance between neighboring adsorbed protein
molecules to be s ) (4π/Γ)1/2 - 2rnative. For example, for a high
surface concentration of 0.828 mg/m2, s ) 0.854 nm, which is
comparable to a double-layer thickness of 0.78 nm. However,
for a low surface concentration of 0.184 mg/m2, s ) 8.54 nm.
As adsorbed protein molecules unfold with time, the average
distance between adsorbed protein molecules decreases, thus
increasing protein-protein interactions. As a result, one can
expect the effect of ionic strength to be more pronounced for
smaller distances between adsorbed protein molecules at longer
times.
The measured unfolded fraction of adsorbed protein on silica
nanoparticles when exposed to different temperatures for 8 min
is shown in Figure 6. Control refers to protein in the bulk in the
absence of nanoparticles. As expected, the extent of unfolding
was higher in the presence of nanoparticles, especially at lower
surface concentration. Also, there was very little change in the
unfolded fraction at up to 60 °C for control samples. At higher
temperatures, however, the change in the unfolded fraction with
temperature was larger. Interestingly, the plot of the unfolded
fraction versus temperature was shifted vertically in the presence
of nanoparticles, with this shift being larger at lower surface
concentrations. In other words, the adsorbed protein on nano-
particle surfaces at very high surface concentration behaved more
like a protein molecule in the bulk.
ANS Binding Assay. ANS binding is used as a probe to detect
exposed hydrophobic patches during protein unfolding. Because
of higher fluorescence intensity as a result of ANS binding, this
method is more sensitive than tryptophan fluorescence to detect
protein unfolding qualitatively. However, ANS binding is not
used to quantify the extent of unfolding because (1) the quantity
of hydrophobic patches bound to ANS is not identical to the
Langmuir, Vol. 24, No. 9, 2008 4995
Figure 11. Effect of DTT on the unfolding kinetics of β-lactoglobulin
adsorbed on the surface of silica nanoparticles at neutral pH and a
low concentration of salts. Normalized data of the fluorescence
intensity of β-lactoglobulin at 380 nm at different surface concentra-
tions expressed as the unfolded fraction. The buffer was 10 mM
phosphate saline at pH 7.0.
extent of unfolding though they are related and (2) no reference
can be set to define the fully denatured state on the basis of ANS
binding. Therefore, ANS binding was employed to confirm the
observations made by tryptophan fluorescence. The fluorescence
spectra of an adsorbed protein molecule on a nanoparticle surface
when exposed to ANS binding was shown for different surface
concentrations and temperatures in Figure 7a,b, respectively.
The fluorescence intensity was found to be higher for lower
surface concentrations and higher temperatures, thereby indicating
more unfolding under these conditions, which is consistent with
the observed results using tryptophan fluorescence.
Characterization of Tertiary Conformation using FTIR
Spectroscopy. To further characterize the tertiary structure change
of protein on nanoparticle surfaces, we used an FTIR technique.
An advantage of FTIR is that it enables the determination of
spectra of adsorbed proteins on nanoparticle surfaces because
the nanoparticles do not scatter light in the wavelength range of
FTIR (the mid-infrared region of 4000-650 cm-1) because the
size of the nanoparticles is much smaller. FTIR spectra of proteins
in deuterium solution are used to monitor the tertiary structure
change based on the extent and the rate of exchange of deuterium
with hydrogen inside protein molecules, providing information
on solvent accessibility and the hydrogen bond stability of amide
bonds.
Typical H-D exchange of β-lactoglobulin adsorbed onto the
nanoparticles is shown at different times of exposure to D2O in
Figure 8a. The area under amide I did not change significantly
whereas the area under amide II (1550 cm-1) decreased with a
corresponding increase in the area under amide II′ (1450 cm-1).
The changes in amide II and amide II′ areas were significant
only up to an exposure time of 30 min as can be seen from Figure
8b, which shows the effect of D2O exposure for different durations
on the extent of exchange of deuterium inside the protein molecule
that was adsorbed onto the nanoparticle surface. A longer exposure
time of protein to D2O led to higher ratio of the amide band II′
area to the amide band I area as a result of more deuterium
exchange. However, this ratio reached a plateau when the exposure
time reached a critical value, which is found to be around 30 min
(Figure 8) Thus in the following experiments, D2O exposure to
nanoparticles with adsorbed protein was maintained at 30 min
for the measurement of the degree of protein unfolding.
Typical H-D exchange of β-lactoglobulin adsorbed onto the
nanoparticles is shown in Figure 9a for different times when
4996 Langmuir, Vol. 24, No. 9, 2008 Wu and Narsimhan

Figure 12. Effect of TFE on the secondary structure of β-lactoglobulin adsorbed on the surface of silica nanoparticles at neutral pH and
a low concentration of salts. (a) CD spectra of β-lactoglobulin at different surface concentrations without TFE, (b) percentage of secondary
structure of β-lactoglobulin at different surface concentrations without TFE, (c) CD spectra of β-lactoglobulin at different TFE concentrations
without nanoparticles, and (d) percentage of secondary structure of β-lactoglobulin at different TFE concentrations without nanoparticles.

incubated with D2O for 30 min. The area under amide I remains
almost the same whereas the area under amide II (1550 cm-1)
decreased with a corresponding increase in the area under amide
II′ (1450 cm-1). This indicates more accessibility of deuterium
with hydrogen inside the protein molecule, thereby implying
tertiary conformational changes due to unfolding. The evolution
of the unfolding of β-lactoglobulin onto the silica nanoparticle
surface for different nanoparticle concentrations is shown in
Figure 9b as the ratio of the areas under amide II′ and amide I.
As shown in the Figure, a lower surface concentration leads to
a higher amide II′/amide I ratio, indicating more unfolding.
However, no significant difference in the unfolding rates was
found between different surface concentrations, and the equi-
librium time was around 60 min, which was much faster than
those obtained by tryptophan fluorescence. This difference might
indicate that mixing protein in D2O could accelerate the exposure
of the hydrophobic core inside of the protein to solvent and thus
also accelerate the kinetics of the interaction of protein and
hydrophobic surfaces. The unfolded fraction of β-lactoglobulin
on silica nanoparticle surfaces could not be determined by FTIR
in our study because the calibration for protein molecules at
different extents of unfolding could not be obtained using GdmCl
in D2O.
Both FTIR and tryptophan fluorescence revealed similar trends
with respect to the extent of protein unfolding on surfaces (i.e.,
protein at lower surface concentration unfolds to a greater extent,
possibly as a result of less interaction between protein molecules).
However, the unfolding rates of β-lactoglobulin on nanoparticle
surfaces determined by fluorescence and FTIR were different.
This may be caused by the presence of D2O in the system or by
the intrinsic difference between the two methods. Fluorescence
detects the microenvironmental change in tryptophan, whereas
FTIR focuses on the N-H bending vibrations of backbones.
Thus, using these two techniques for the same process gave
seemingly different unfolding rates.
Effect of TFE and DTT on the Tertiary and Secondary
Structures of Protein on the Surface. 2,2,2-Trifluoroethanol
(TFE) is reported to induce a change in tertiary structure by
destabilizing hydrophobic interactions while stabilizing hydrogen
bonding. Therefore, TFE has been widely used in the study of
partially folded states of intact proteins, in unfolding transition
investigations, and in the kinetics of protein folding. Dithiothreitol
(DTT) is an unusually strong reducing agent, owing to its high
conformational propensity to form a six-membered ring with an
internal disulfide bond. Therefore, DTT is frequently used to
reduce the disulfide bonds of proteins and, more generally, to
prevent intramolecular and intermolecular disulfide bond forma-
tion between cysteine residues, thus leading to the loss of tertiary
structure.
The evolution of the unfolding of β-lactoglobulin on silica
nanoparticle surfaces at different TFE concentrations is shown
for two different surface concentrations in Figure 10. At higher
surface concentration (0.83 mg/m2), the extent of unfolding and
the unfolding rate increased as the TFE concentration increased.
However, at lower surface concentration (0.18 mg/m2), the
increasing TFE concentration accelerated the unfolding rate but
had no significant effect on the extent of unfolding. This is because
at low surface concentration the protein molecule almost fully
extends its tertiary structure, thus leading to the exposure of its
hydrophobic groups to the binding surface. As a result, the addition
of TFE had no significant effect on the extent of unfolding.
The evolution of unfolding of β-lactoglobulin on silica
nanoparticle surfaces with or without DTT is shown in Figure
11 at three different surface concentrations. At higher surface
concentration, the extent of unfolding as well as the unfolding
rate increased in the presence of DTT. However, similar to what
Conformational Changes of β-Lactoglobulin Langmuir, Vol. 24, No. 9, 2008 4997

Table 2. Secondary Structural Content of β-Lactoglobulin

Adsorbed onto the Nanoparticle Surface at Different TFE
Concentrations
surface
conc TFE%
(mg/m2) (v/v) R helix β sheet others
0 0 0.151 0.351 0.498
5 0.163 0.317 0.520
15 0.196 0.283 0.521
30 0.551 0.055 0.394
0.83 0 0.170 0.325 0.505
5 0.163 0.305 0.532
15 0.203 0.268 0.529
30 0.553 0.050 0.397
0.46 0 0.151 0.332 0.517
5 0.158 0.313 0.529
15 0.192 0.260 0.548
30 0.554 0.046 0.400
0.18 0 0.155 0.326 0.519
5 0.154 0.307 0.539
15 0.189 0.264 0.547
30 0.539 0.053 0.408

Table 3. Secondary Structural Content of β-Lactoglobulin

Adsorbed onto the Nanoparticle Surface in the Presence and
Absence of DTT
surface
conc DTT
(mg/m2) (mg/mL) R helix β sheet others
0 0 0.151 0.351 0.498
0.5 0.148 0.328 0.524 Figure 13. Effect of DTT on the secondary structure of β-lacto-
0.83 0.5 0.151 0.315 0.534 globulin adsorbed on the surface of silica nanoparticles at neutral
0.46 0.5 0.151 0.291 0.558 pH and a low concentration of salts. (a) CD spectra of β-lactoglobulin
0.18 0.5 0.137 0.265 0.598 at different surface concentrations in 0.5 mg/mL DTT and (b)
percentage of secondary structure of β-lactoglobulin at different
we found for TFE, at lower surface concentration, DTT surface concentration in 0.5 mg/mL DTT. No Np NoDTT refers to
accelerated the unfolding rate without any significant effect on the solution without nanoparticles and DTT. No Np DTT refers to
the solution without nanoparticle in the presence of DTT. The other
the extent of unfolding. conditions are at different surface concentrations in the presence of
CD spectra were employed here to detect the secondary DTT.
structure of β-lactoglobulin on the nanoparticle surface under
different conditions. Figure 12a compares the CD spectra of These results indicate that TFE has a much stronger influence
β-lactoglobulin in solution with those of protein adsorbed on the on the interaction with protein molecules than on the interaction
nanoparticle surface at different surface concentrations, which between protein and the nanoparticle surface.
reveals that β-lactoglobulin retained its secondary structure almost The effect of DTT concentration on secondary structure is
intact upon adsorption to the nanoparticle surface, even at low shown in Figure 13. The CD spectra (Figure 13a) show that the
surface concentration. A similar observation was made for the addition of DTT to the solution significantly changed the
behavior of other proteins adsorbed on flat surfaces by previous secondary structure of β-lactoglobulin in solution or on the
investigators.10,21 Further analysis of the percentage of different nanoparticle surfaces, and this effect was enhanced by increasing
secondary structure by the CONTIN program is shown in Figure the nanoparticle concentration (lowering the protein surface
12b, confirming that all of the secondary structure of β-lacto- concentration). This suggests that disulfide bonds are very
globulin remained intact whereas the tertiary structure was lost. important to keeping the secondary structure stable because DTT
We also measured the CD spectra of β-lactoglobulin exposed to mainly broke down the disulfide bonds and therefore changed
different concentrations of TFE (0, 5, 15, and 30%) at a specific the secondary structure. The enhanced effect of DTT on secondary
surface concentration. Figure 12c shows the CD spectra of structure at lower surface concentration also indicates a strong
β-lactoglobulin in TFE solution without any particles, revealing relationship between disulfide bonds and secondary structure:
that the secondary structure of β-lactoglobulin is dramatically after the disulfide bonds are broken, the β-lactoglobulin molecule
affected by increasing the concentration of TFE in the solution. becomes more flexible, thus making its secondary structure prone
The secondary structure content of the protein is shown in Figure to be affected by the interaction with the surface. Such behavior
12d, indicating an increase in R-helix content at the expense of is similar to that with respect to the tertiary structure (i.e., more
β sheets as the concentration of TFE was increased. The effects change at lower surface concentration). The secondary structure
of TFE and surface concentration on the secondary structure of analysis (Figure 13b) shows that there is a loss of β-sheet content
adsorbed β-lactoglobulin on the nanoparticle surface are shown at higher DTT concentration and lower protein surface concen-
in Table 2. At a fixed TFE concentration, changing the surface tration with a corresponding increase in the content of random
concentration did not alter the secondary structure of β-lacto- coils and turns, whereas the R-helix content stays almost the
globulin in the TFE concentration range of 0-30%. Different same. Such an exchange of secondary structure is mostly due to
TFE concentrations did have a significant influence on the the fact that the two disulfide bonds of β-lactoglobulin link two
secondary structure irrespective of the surface concentration. cystein amino acids located in two different β sheets, whereas
4998 Langmuir, Vol. 24, No. 9, 2008
the other one is between one β sheet and one random coil.
Therefore, DTT had a significant effect on β sheets but an
insignificant effect on the R helix. Table 3 shows the secondary
structure of β-lactoglobulin at different surface concentrations
in the presence or absence of DTT.
Conclusions
To understand better the factors that influence the conformation
of proteins adsorbed on nanoparticle surfaces, the kinetics of
tertiary conformational changes as well as the extent of secondary
conformational changes of a well-characterized protein such as
β-lactoglobulin adsorbed on silica nanoparticles is reported. A
rapid initial unfolding followed by much slower rates at longer
Wu and Narsimhan
times was observed after a time lag, with this time lag being
shorter at lower surface concentrations. The effects of surface
concentration, pH, and ionic strength on the unfolding kinetics
indicated that protein-protein interactions on nanoparticle
surfaces are important. The variation of the surface concentration
of β-lactoglobulin as well as the presence of TFE did not influence
the secondary conformational change on the surface. DTT,
however, was found to result in a decrease in the β-sheet fraction
with a corresponding increase in the random coil fraction.
Acknowledgment. We acknowledge the Purdue Research
Foundation for providing financial support to X.W.
LA703349C