7 Biopolymer-Biopolymer
Interactions at Interfaces
and Their Impact on
Food Colloid Stability
S. Damodaran
CONTENTS
Abstract
Introduction
Theoretical Background
Experimental
Adsorption
Epifluorescence Microscopy
Results and Discussion
Thermodynamic Incompatibility
Two-Dimensional Phase Separation
Conclusions
Acknowledgment
References
Abstract
Two-dimensional phase separation in several protein-1/protein-2/water ternary film
adsorbed at the air–water interface was studied using an epifluorescence microscopy
technique. The equilibrium composition of saturated mixed monolayer films of αs-
casein/β-casein, BSA/β-casein, and soy 11S/β-casein at the air–water interface at
various bulk concentration ratios did not follow a Langmuir-type competitive adsorp-
tion model. This was due to a change in the binding affinity of the proteins to the
interface as a result of incompatibility of mixing of the proteins in the mixed
monolayer. Among the systems studied, only the α-lactalbumin/β-casein/water ter-
nary film followed the ideal Langmuir behavior for competitive adsorption, indicat-
ing that these two proteins were thermodynamically compatible at the air–water
© 2002 by CRC Press LLC
interface. Fluorescence microscope images showed that the ternary films containing
incompatible proteins underwent two-dimensional phase separation. In these phase-
separated films, one of the proteins was present as dispersed phase and the other as
continuous phase. It is proposed that the interface between such phase-separated
regions of the protein film may act as “fault zones,” promoting instability in protein
stabilized foams, and possibly emulsions.
INTRODUCTION
A majority of processed foods are either foam-type (air-in-water) or emulsion-type
(oil-in-water) products. The stability of these dispersed colloidal systems depends
on the presence of an ampiphilic film at these interfaces. Proteins, which are
amphiphilic, exhibit a high propensity to migrate and bind to air–water and oil–water
interfaces and decrease the interfacial tension. In addition to lowering the interfacial
tension, the adsorbed protein can form a strong viscoelastic film via intermolecular
interactions that can withstand thermal and mechanical ,2 In addition,
conformational changes in proteins at interfaces allow them to form loops that
protrude from the interface into the bulk phase. The steric repulsion caused by
overlapping of the layer of protruding chains of emulsion particles when they
approach each other is considered to be the most important force stabilizing emul-
The latter property makes proteins more desirable than low-molecular-weight
ampiphiles such as lecithin and monoglycerides as surfactants in emulsion- and
foam-type food products. However, proteins differ significantly in their surface
activity and in their ability to stabilize colloidal systems. This can be attributed to
differences in their structural properties and their ability to undergo conformational
reorientation at
Notwithstanding such structure-dependent innate differences among proteins,
another important factor that might influence the stability of food emulsions and
foams is the thermodynamic incompatibility among proteins in the adsorbed protein
film at the interfaces. Typical food proteins, notably protein blends, used in food
industries are mixtures of several proteins. As a result, the protein film formed around
oil droplets of an emulsion usually consists of a mixture of proteins. The stability
of the food colloid will be dependent on the nature and intensity of protein-protein
interactions in the film. Generally, interactions between two dissimilar polymers are
thermodynamically incompatible. This is true of proteins as well. In concentrated
aqueous solutions (10–20% w/v), mixtures of two proteins exhibit thermodynamic
incompatibility of mixing and, as a result, undergo phase –8 Because the
local concentration of proteins in an adsorbed protein film is equivalent to 15–30%,
it is likely that two-dimensional phase separation of proteins might occur in the film.
If this occurs, then the “interface” between the phase-separated regions in a mixed
protein film might act as zones of high free energy. The stability and integrity of
the mixed protein film might be adversely affected by these high free energy “fault
zones,” which, in turn, might promote coalescence of oil droplets.
The thermodynamic incompatibility between two polymers arises from energet-
© 2002 by CRC Press LLC
ically unfavorable interactions between them. In ternary protein solutions, even if
the protein-1/protein-2 interaction parameter, χ12, is negligible, thermodynamic
incompatibility can still exist if there is a large difference between protein-1/solvent
(χ1s) and protein-2/solvent (χ2s) interaction ,9 The larger the value ofχ1s – χ2s, the higher the incompatibility. Conversely, the greater the difference in
hydrophilicities of proteins, the greater their incompatibility. However, one cannot
correctly predict incompatibility of two proteins at an interface based on their
behavior in solution, for two reasons. The thermodynamics at the asymmetrical force
field of an interface is not the same as that in the bulk phase, and the structural states
of proteins at an interface are not the same as that in bulk solution. Thus, proteins
that exhibit compatibility or incompatibility in solution may or may not exhibit the
same behavior in a mixed protein film at an interface.
Virtually no information currently exists in the literature on thermodynamic
incompatibility and two-dimensional phase separation of proteins in mixed protein
films at interfaces. This situation is partly due to lack of a theoretical framework
and experimental approach to study phase behavior of polymer mixture at an inter-
face. Recently, we have developed an experimental approach to study this phenom-
In this contribution, we provide experimental evidence of incompatibility
and phase separation in several mixed protein films at the air–water interface.
THEORETICAL BACKGROUND
Although theoretical models to study thermodynamic incompatibility of biopolymers
in solution are fairly well developed, no model currently exists for studying this
phenomenon at interfaces. We propose the following model as a theoretical basis
for studying thermodynamic incompatibility between biopolymers at interfaces.
Adsorption of proteins and other biopolymers at interfaces is generally assumed
to follow a Langmuir adsorption model. The Langmuir adsorption model for a
protein-1/solvent system is as follows:
()
where Γ = surface concentration
K = equilibrium binding constant
C = bulk phase protein concentration
a = area occupied per protein molecule at saturated monolayer coverage
For a protein-1/protein-2/water ternary system, the Langmuir adsorption model for
competitive adsorption is as follows:11
()
Γ KC
1 KaC+
---------------------=
Γ1
K 1C1
1 K1a1C1 K 2a2C2+ +
----------------------------------------------------=
© 2002 by CRC Press LLC
and
()
where K1 and K2 are the equilibrium binding constants of protein-1 and protein-2,
respectively, to the interface in protein-1/solvent and protein-2/solvent systems, C1
and C2 are their concentrations in bulk mixture, Γ1 and Γ2 are concentrations in the
mixed film at the interface and a1 and a2 are the area occupied by protein-1 and
protein-2, respectively, at saturated monolayer coverage in protein-1/solvent and
protein-2/solvent systems, respectively. This Langmuir model for competitive
adsorption in a protein-1/protein-2/water ternary system assumes that the surface
concentrations of protein-1 and protein-2 in the mixed protein film at the interface
are affected only by their relative binding affinities to the interface and their con-
centration ratio in the bulk phase. It assumes a priori that the adsorbed protein
molecules do not interact with each other and that their adsorption is dictated only
by their relative affinity to the interface and the availability of vacant sites at the
interface. Under noninteracting conditions, from Equations () and (), the
Langmuir adsorption model predicts the following:
()
and
()
Equations () and () suggest that, when K1 = K2, the mass ratio of protein-1 to
protein-2 in the mixed protein film will be exactly the same as the ratio of their
concentrations in the bulk phase. That is, a plot of Γi/Γtot versus Ci/Ctot will be a
straight line with a slope of 1. When K1 ≠ K2, the plot will be nonlinear, and it will
be a function of the ratio K1/K2. A family of theoretical Γi/Γtot versus Ci/Ctot curves
generated for various ratios of K1/K2 for protein-1/protein-2/water ternary systems
that obey Langmuir-type competitive adsorption is shown in Figure . Knowing
the ratio of the equilibrium binding constants of the proteins to an interface, deter-
mined from their adsorption isotherms in single protein systems [Equation ()],
we can predict the ideal competitive adsorption behavior in a protein-1/protein-
2/water ternary system.
If the proteins of a protein-1/protein-2/water ternary system exhibit thermody-
namic incompatibility of mixing at an interface, then the first impact of that would
be on the binding affinities of the proteins to the interface. The ratio of binding
affinities of the proteins in the ternary adsorption system would not be the same as
the ratio determined from single-protein adsorption , if thermodynamic
incompatibility between two proteins exists at an interface, then the experimental
Γ2
K 2C2
1 K1a1C1 K 2a2C2+ +
----------------------------------------------------=
Γ1
Γtot
--------
K1C1
K 1C1 K 2C2+
--------------------------------=
Γ2
Γtot
--------
K2C2
K 1C1 K 2C2+
--------------------------------=
© 2002 by CRC Press LLC
Γi/Γtot versus Ci/Ctot plot would not be the same as that predicted by the Langmuir
competitive adsorption model [Equations () and ()]. The extent of deviation
will be a direct measure of the extent of thermodynamic incompatibility between
the proteins.
The thermodynamic incompatibility between proteins at an interface can be
determined using another approach, as follows. From Equations () and (),
()
If K1´/K2´ is the ratio of the binding constants of the proteins in the protein-1/
protein-2/water ternary adsorption system, then the value of ln(K1´/K2´) for the
ternary system can be determined from the intercept of a plot of ln(Γ1/Γ2) versus ln
(C1/C2). If K1 and K2 are the binding constants of two proteins in protein-1/water
and protein-2/water adsorption systems, then the absolute difference between ln
(K1/K2) and ln (K1´/K2´), ., ∆ ln K can be regarded as a measure of thermody-
namic incompatibility among the proteins. Since ∆ ln K is a thermodynamic
quantity representing the intensity of interaction between protein-1 and protein-2
molecules at the interface, it might in fact be a measure of the Flory–Huggins
Figure Theoretical Γ1/Γtot vs. C1/Ctot profiles at various K1/K2 ratios for protein-1/protein-
2/water ternary systems that follow the Langmuir adsorption model according to Equation
().
Γ1/Γ2( )ln K 1/K2( )ln C1/C2( )ln+=
© 2002 by CRC Press LLC
interaction parameter χ12.
EXPERIMENTAL
ADSORPTION
The experimental parameters required to elucidate thermodynamic incompatibility
of proteins at an interface are the equilibrium binding constants (K) of individual
proteins in protein-solvent binary systems and equilibrium surface concentrations
(Γ1 and Γ2) of proteins 1 and 2 in a protein-1/protein-2/solvent ternary system at
various bulk concentration ratios (C1/C2). Adsorption of proteins from the aqueous
phase to the air/water interface was studied by using the surface radiotracer method
as described –14 The proteins were radiolabeled with [14C] nuclide by
reductive methylation of amino groups with [14C]-formaldehyde, as described else-
Equilibrium adsorption of radiolabeled proteins at the air–water (20 mM phos-
phate buffered saline solution, pH , I = ) interface was studied using the surface
radio tracer technique, as described ,15 Competitive adsorption of pro-
teins from a protein-1/protein-2/water ternary solution was studied as To
monitor adsorption of protein-1, stock solutions of [14C]-labeled protein-1 and unla-
beled protein-2 were mixed with the buffer to the required final bulk concentration
ratio. The solution was poured into a Teflon® trough (19 × × cm) and the
surface of the solution was swept using a capillary tube attached to an aspirator. The
proteins were then allowed to adsorb from the bulk phase to the air–water interface.
Although both protein-1 and protein-2 simultaneously adsorbed to the air–water
interface, the measured surface radioactivity corresponded only to the amount of
radiolabeled protein-1 at the interface. To determine the amount of protein-2 at the
interface, the experiment was repeated by mixing stock solutions of [14C]-labeled
protein-2 and unlabeled protein-1 with the buffer. Equilibrium surface concentration
(Γeq) values were obtained from surface radioactivity values after 24–30 h adsorption.
Adsorption isotherms for each protein were constructed by determining Γeq at various
bulk concentrations, Cb, in single-protein systems. These isotherms were used to
determine the binding affinity of proteins to the air–water interface using the Lang-
muir adsorption model [Equation ()].
EPIFLUORESCENCE MICROSCOPY
Phase separation in mixed protein films at the air–water interface was examined
using the epifluorescence microscopy technique. In this approach, protein-1 was
labeled with Fluorescin-5-EX succinimidyl ester and protein-2 was labeled with
Texas Red. In both cases, the fluorescence labeling was at the amino groups of the
proteins. Labeling with Fluorescin-5-EX succinimidyl ester, was carried out as
follows: The protein (~1 µmol) was dissolved in 5 ml of carbonate buffer at pH
and ionic strength M. A known amount of the fluorescent dye (~10 µmol) was
dissolved in ml anhydrous ethanol. The ethanol dye solution was added to the
protein solution, and the reaction mixture was stirred for h. After the reaction,
the mixture was applied to a Sephadex G-25-50 gel-filtration column and eluted
© 2002 by CRC Press LLC
with carbonate buffer (pH and ionic strength M). The fractions corresponding
to the first elution peak which contained the protein, as judged from absorption at
280 nm, were pooled and dialyzed extensively against phosphate-buffered saline
solution (pH and ionic strength M) and were then lyophilized. Under these
reaction conditions, about moles of Fluorescin-5-EX succinimidyl ester was
incorporated per mole of the protein.
Labeling of proteins with Texas Red was carried out by the method outlined by
Sigma Chemical Co. (St. Louis, MO). Briefly, the protein (40 mg) was dissolved in
8 ml of ice-cold sodium carbonate buffer at pH and ionic strength M, and
stirred continuously. Texas Red (4 mg) was dissolved in 1 ml of ethanol. At the
onset of reaction, ml of the Texas Red–ethanol solution were added to the protein
solution and stirred under ice-cold conditions for 10 min. This procedure was
repeated every 10 min by adding ml of the dye-ethanol solution to the protein-
buffer solution until all of the dye solution was consumed. After the last cycle, the
reaction mixture was stirred under ice-cold conditions for an additional hour. The
labeled protein was then purified as described above. Under the conditions employed,
typically, about moles of dye were incorporated per mole of protein.
For epifluorescence microscopy, first, the fluorescent-labeled proteins were
allowed to adsorb to the air–water interface for a period of four days from a solution
containing a mixture of the proteins in a Teflon trough. The mixed protein film
formed at the air–water interface was transferred to a clean microscope slide that
was precoated with 3-aminopropyl triethoxy silane (APTES).Transfer of the film
was done using a horizontal lifting method in which the microscope slide held
horizontally by a vacuum tube was gently lowered to make contact with the aqueous
surface for about 10 s. After lifting the film, excess water was allowed to drain off
for 1 min. The slide was then dipped in water to rinse loosely bound proteins from
the glass slide. The slide was then air dried. A drop of SlowFade Light Antifade
reagent in glycerol-water was added to the center of the protein film on the glass
slide and covered thereafter with a clean cover glass. This reagent prevented photo-
bleaching of the fluorophores during observation under the microscope. The slides,
thus prepared, were immediately observed under a computerized Olympus epifluo-
rescence microscope equipped with a wide excitation and band-pass emission Ore-
gon Green optical filter cube assembly selective for green (EX: 495 ± 15 nm, EM:
545 ± 25 nm), a narrow excitation and a long-pass emission Texas Red filter cube
assembly selective for red (EX: 545–550 nm, EM: 610 and above) and a wide
excitation and a long-pass emission Fluorescin filter cube assembly (EX: 460–490
nm, EM: 515 nm and above) for simultaneous detection of both dyes. The fluores-
cence images were digitized and analyzed with Olympus Image-Pro Analysis soft-
ware.
RESULTS AND DISCUSSION
THERMODYNAMIC INCOMPATIBILITY
The relative affinity of various proteins to the air–water interface was determined by
© 2002 by CRC Press LLC
measuring the surface concentration (Γ) of proteins at equilibrium at various bulk
concentrations (Cb) in single-protein systems. For all proteins listed in Table , the
adsorption isotherms, ., plots of Γ versus Cb, showed a well-defined plateau in the
bulk concentration range of – µg·ml–1. The saturated monolayer coverage
values, Γsat, for various proteins are given in Table . The equilibrium binding
constant K for the proteins, calculated using Equation (), is also shown in Table .
To determine thermodynamic incompatibility of proteins in a mixed protein film
at the air–water interface, the protein composition of the mixed protein film at
equilibrium was determined at various concentration ratios in the bulk. The protein-
1/protein-2/water ternary systems investigated are listed in Table . Figure
shows experimental Γi/Γtot versus Ci/Ctot curves for the ternary systems studied. It
should be pointed out that similar plots for the other protein component of the ternary
system would appear as inverse of the curves shown.
The theoretical curve predicted by Equation () or (), based on the Ki
values
obtained from single-component systems for each of the ternary systems, is also
shown in Figure (dotted lines). It should be noted that, except for the α-
lactalbumin/β-casein system, the experimental Γi/Γtot versus Ci/Ctot curves deviate
significantly from the predicted curves. Among the ternary systems shown in Figure
, only the α-lactalbumin/β-casein system showed somewhat ideal behavior at the
TABLE
Adsorption Isotherm Parameters of Various Proteins at the Air–Water
Interface
Protein K (cm) Γsat (mg/m–2)
Egg lysozyme (EL)
α-Lactalbumin (α-La)
Serum albumin (BSA)
Soy 11S
α-Casein (α-CN)
β-Casein (β-CN)
TABLE
Thermodynamic Incompatibility Parameters for Various Protein-1/Protein-
2/Water Ternary Systems at the Air-Water Interface
Protein-1/Protein-2 ln(K1/K2) ln(K1´/K2´) ∆ ln K X12
EL/β−CN − −
α-La/β-CN − − ~0
BSA/β-CN − −
α-CN/β-CN − −
11S/β-CN − −
© 2002 by CRC Press LLC
air–water interface. This is surprising because, in physicochemical terms, these two
Figure Plots of experimental Γi/Γtot vs. Ci/Ctot for various protein-1/protein-2/water ter-
nary systems at the air-water interface. The dotted lines represent the ideal curve predicted
© 2002 by CRC Press LLC
by Equation ().
proteins are very dissimilar. Whereas α-lactalbumin is a hydrophilic albumin-type
protein, β-casein is a hydrophobic disordered-type protein. A mixture of such pro-
teins is known to undergo phase separation in At the air–water interface,
however, it is quite likely that α-lactalbumin may undergo extensive denaturation.
Exposure of the interior hydrophobic groups may render it more hydrophobic in the
denatured state, thus enabling it to be more compatible with the hydrophobic β-
casein.
It is interesting to note that, although α-casein and β-casein belong to the same
class of proteins, they exhibit some degree of incompatibility at the air–water
interface. These two proteins are known to be highly compatible in This
contradictory behavior further emphasizes the suggestion that the thermodynamics
or protein-protein and protein-solvent interactions at an interface are different from
those in solution.
Compatibility of two proteins in a ternary system will be dictated by the delicate
interplay of four controlling parameters:
• protein-1-protein-2 interaction parameter (χ12)
• protein-1-solvent and protein-2-solvent interaction parameters (χ1s, χ2s)
• molecular weights of the two protein molecules
• conformation states of the two proteins7
From the Flory–Huggins theory, the critical interaction parameter, χc, for a binary
polymer-solvent system is and for a binary mixture of polymers is 0. The
significance of the critical interaction parameter is that two components will be
miscible if their interaction parameter is below χc but will separate into two phases
if the value of this parameter is above χc. Although intuitively two proteins in a
protein-1-protein-2-water ternary system should be totally compatible when the
interaction parameter χ12 is 0 (for instance, α-casein/β-casein), in reality, they may
still exhibit incompatibility in dilute solutions due to a difference in the protein-
solvent interaction parameters, χ1s − χ2s , for as little as The absolute
magnitude of the polymer–solvent interaction parameter has a considerably smaller
effect on polymer–polymer incompatibility than that due to the difference in the
polymer–solvent interaction parameters χ1s − χ2s for the two polymers of
The larger the difference in hydrophilicity (., χ1s − χ2s) of the two proteins, the
greater the incompatibility and the lower the threshold for phase separation in a
ternary system. On the other hand, if the solvent is equally good for each polymer
(., χ1s = χ2s), two polymers may yield a totally miscible solution in spite of a small
positive value of χ It should be pointed out that because the thermodynamic
state of water at the interface is different from that in bulk solution, the χ1s − χ2s
value for a ternary film at the air–water interface may not be the same as that
calculated for polymer solutions. Because the density of water at the air–water
interface is lower than its density in the bulk phase, a higher χ1s − χ2s value will
be expected for interactions at the air–water interface.
As discussed earlier, the deviation of the experimental Γi/Γtot versus Ci/Ctot curves
© 2002 by CRC Press LLC
from that predicted by the Langmuir equation is due primarily to incompatibility of
mixing of the proteins in the mixed protein film. The degree of incompatibility
between two proteins at an interface can be empirically determined from the extent
of deviation of the experimental curve from the predicted one. If two proteins are
totally incompatible with each other, then they cannot coexist at the interface. For
this situation, the total area above the concave surface of the ideal Langmuir curve
can be regarded as total incompatibility. Thus, the ratio of the area between the
experimental and predicted curves to the area representing total incompatibility can
be defined as the degree of incompatibility, X12, between the proteins at the air–water
interface. The X12 values for the five ternary systems are shown in Table . The
data suggest that the degree of incompatibility of β-casein with other proteins at the
air–water interface increases in the following order: α-lactalbumin < α-casein <
BSA < egg lysozyme < soy 11S.
The X12 values determined by the graphical method are only empirical in nature.
As discussed earlier, fundamentally, the deviation of the experimental Γi/Γtot versus
Ci/Ctot curve from the predicted one is due to a change in the binding affinity of a
protein to the interface in the presence of the other protein at the interface as a result
of incompatibility. The relative changes in the binding affinities of proteins in a
ternary system as compared to that in a protein/solvent binary system can be deter-
mined from Equation () by plotting ln (Γ1/Γ2) versus ln (C1/C2). The values of ln
(K1´/K2´) obtained from the intercept of such plots for the ternary systems at the
air–water interface are given in Table . The values of ln (K1/K2), determined from
protein/solvent binary systems at the air–water interface and the absolute difference
between ln (K1/K2) and ln (K1´/K2´), ., ∆ ln K , are also shown in Table . Since∆ ln K represents a net change in the ratio of the binding constant between the
ternary and binary systems at the interface, it represents the intensity of incompatible
interactions between proteins at the interface. A linear relationship between X12 and∆ ln K (Figure ) clearly demonstrates that these two are reflections of each other.
TWO-DIMENSIONAL PHASE SEPARATION
If ∆ ln K and X12 truly reflect the existence of thermodynamic incompatibility
between proteins in a mixed film at the air–water interface, then one should be able
to observe two-dimensional phase separation of proteins in the mixed film. This was
investigated using the epifluorescence microscopy technique.
In initial studies, it was observed that although equilibrium adsorption of the
proteins at the air–water interface took place within 24 h, the mixed protein films
transferred from the air–water interface to the αs-casein/β-casein mixed film trans-
ferred from the air–water interface to the APTES glass slide after just 24 h of
adsorption did not contain distinct phase-separated regions. However, when the film
at the air–water interface was aged for four days and then transferred, distinct phase-
separated regions could be observed under the fluorescence microscope. This sug-
gests that in the initial stages, the two proteins adsorb randomly to the air–water
interface and then undergo two-dimensional phase separation. Phase separation of
proteins at the interface seems to be kinetically limited, perhaps because of the
© 2002 by CRC Press LLC
viscosity barrier to lateral diffusion of the molecules. It should be pointed out that
the local protein concentration in a saturated monolayer film at the air–water inter-
face is almost equivalent to 20–30%, and the viscosity barrier for lateral diffusion
at this high concentration can indeed be very high.
Figure shows the fluorescence microscope image of αs-casein and β-casein
mixed film transferred from the air–water interface. In the image taken with the
Texas Red filter (Figure ), there are large black patches surrounded by a red
region corresponding to β-casein (light gray region in Figure ). These black
patches appear as intense green patches (light gray regions in Figure ) and the
red continuous region appears as green region (continuous medium-gray region in
Figure ) when the image is taken with the Oregon green filter. This clearly
suggests that, whereas the continuous phase contains an inhomogeneous mixture of
both αs- and β-caseins, the dispersed phase predominantly contains αs-casein with
very little or no β-casein. In all αs-casein to β-casein concentration ratios in the
mixed film, β-casein was present only in the continuous phase and αs-casein the
dispersed phase.
Figure shows fluorescence images of BSA/β-casein and α-lactalbumin/β-
casein mixed films transferred from the air–water interface. In this system, BSA was
labeled with Fluorescin-5-EX succinimidyl ester, and β-casein was labeled with
Texas Red. The image taken with the Texas Red filter shows a dark oval region
surrounded by a red region that corresponds to β-casein (light gray region in Figure
). When the same image was taken with the green filter, the dark region, which
corresponds to BSA, appeared light-green (light gray region in Figure ) sur-
Figure Relationship between the empirical incompatibility parameter, X12, and interaction
parameter, | ∆lnK |.
© 2002 by CRC Press LLC
rounded by a dark-green region (dark gray region in Figure ). When the same
image was taken with the mutual filter that allowed both the red and green fluores-
cence, the image contained a green patch (light gray patch in Figure ) surrounded
by a red continuous phase (continuous medium gray region in Figure ). It should
be pointed out that if the continuous phase contained a significant amount BSA, a
combination of green and red colors should have provided a yellow tinge to the
continuous phase. Thus, the orange-red color of the continuous phase under the
mutual filter (medium gray region in Figure ) suggests that it predominantly
contained β-casein and very little BSA. It should be noted that, as in the case of the
αs-casein/β-casein system, β-casein formed the continuous phase and BSA the
dispersed phase. This trend was observed at all BSA/β-casein concentration ratios
at the interface.
Figure shows a fluorescence image of a α-lactalbumin/β-casein film trans-
ferred from the air–water interface. In this system, α-lactalbumin was labeled with
Figure Fluorescence image of a mixed film of α-casein and β-casein at the air–water
interface. (a) Light and medium gray regions correspond to green regions of α-casein. (b)
Light gray regions correspond to the red regions of β-casein. Bar = 100 µm.
(a)
(b)
© 2002 by CRC Press LLC
Fluorescin-5-EX succinimidyl ester and β-casein with Texas Red. According to
Figure Fluorescence image of a mixed film of ΒSΑ and β-casein at the air–water interface.
Light gray regions in (a) and (c) correspond to BSA, and the dark gray region in (b) and the
© 2002 by CRC Press LLC
medium gray region in (c) correspond to β-casein. Bar = 100 µm.
the data presented in Figure , α-lactalbumin and β-casein appear to be thermo-
dynamically compatible and, therefore, two-dimensional phase separation would
not be expected in the film. However, the fluorescence image taken with the green
filter shows a dense, oval, green region (light gray region in Figure ) surrounded
by a dark-green continuous phase (black region in Figure ). The oval region
is apparently richer in α-lactalbumin than the continuous phase. The interior of
the oval region appears homogeneous with some dense internal structures. This
was not observed in other systems. When the same image was taken with the Texas
Red filter (Figure ), β-casein was found both in the continuous phase and in
the oval region. It seems that, whereas the continuous phase contains a homoge-
neous mixture of α-lactalbumin and β-casein, the dense oval-shaped region con-
tains aggregated complexes of α-lactalbumin and β-casein. These aggregates might
be complexes formed via hydrophobic interaction between β-casein and denatured
α-lactalbumin. Formation of complexes, existence of a homogeneous phase and
Figure Fluorescence micrograph of a mixed film of α-lactalbumin and β-casein at the
air–water interface. (a) Light gray and black regions correspond to the green regions of α-
casein. (b) Light gray regions correspond to the red regions of β-casein. Bar = 100 µm.
(a)
(b)
© 2002 by CRC Press LLC
in the film clearly suggests that α-lactalbumin and β-casein are compatible with
each other at the air–water interface. This agrees very well with the data shown
in Figure .
CONCLUSIONS
The fluorescence images of several mixed protein films at the air–water interface
clearly showed two-dimensional phase separation. The phase-separated regions
become more prevalent during aging of the film than at the initial stages. This is
probably due to hindered diffusion of the protein components in the highly viscous
film. In most cases, one of the proteins in the mixed film formed a dispersed phase
and the other the continuous phase. However, it can be assumed that, if sufficient
time has been provided, the film may eventually phase separate into protein-1 rich
and protein-2 rich phases, separated by a region of inhomogeneous mixture. The
phase-separation phenomenon is due to thermodynamic incompatibility of mixing
of the proteins in the two-dimensional film. The theoretical and the experimental
approaches described here seem to adequately predict the intensity of thermody-
namic incompatibility between proteins. The tendency of two-dimensional lateral
phase-separation in mixed protein films at interfaces may adversely impact the
stability of protein-stabilized foam and emulsions. Since the “interface” between the
phase-separated regions of the film around oil droplets or air bubbles is at a higher
free energy, they might act as zones of instability in emulsions and foam.
ACKNOWLEDGMENT
This research was supported by a grant from the National Science Foundation (Grant
#BES-9712197).
REFERENCES
1. Phillips, M. C. 1981. “Protein Conformation at Liquid Interfaces and Its Role in
Stabilizing Emulsions and Foams,” Food Technol., 35: 50–57.
2. Damodaran, S. 1997. “Protein-Stabilized Foams and Emulsions,” in Food Proteins
and Their Applications, S. Damodaran and A. Paraf, eds. New York, NY: Marcel
Dekker, pp. 57–110.
3. Liu, M. and S. Damodaran. 1999. “Effect of Transglutaminase-Catalysed Polymer-
ization of β-Casein on its Emulsifying Properties,” J. Agric. Food Chem., 47:
1514–1519.
4. Damodaran, S. and L. Razumovsky. 1998. “Molecular Bases for Surface Activity of
Proteins,” Amer. Chem. Soc. Symp. Ser., 708: 2–18.
5. Polyakov, V. I., V. Y. Grinberg, Y. A. Antonov, and V. B. Tolstoguzov. 1979. “Limited
Thermodynamic Compatibility of Proteins in Aqueous Solutions,” Polym. Bull., 1:
593–597.
6. Polyakov, V. I., I. A. Popello, V. Y. Grinberg, and V. B. Tolstoguzov. 1986. “Thermo-
© 2002 by CRC Press LLC
dynamic Compatibility of Proteins in Aqueous Medium,” Nahrung, 30: 365–368.
7. Polyakov, V. I., V. Y. Grinberg, and V. B. Tolstoguzov. 1997. “Thermodynamic Incom-
patibility of Proteins,” Food Hydrocolloids, 11: 171–180.
8. Ledward, D. A. 1994. “Protein/Polysaccharide Interactions,” in Protein Functionality
in Food Systems, N. S. Hettiarachchy and G. R. Ziegler, eds. New York, NY: Marcel
Dekker, pp. 225–259.
9. Zeman, L. and D. Patterson. 1972. “Effect of the Solvent on Polymer Incompatibility
in Solution,” Macromolecules, 5: 513–516.
10. Razumovsky, L. and S. Damodaran. 1999. “Thermodynamic Incompatibility of Pro-
teins at the Air- Water Interface,” Colloids Surf. B: Biointerfaces, 13: 251–261.
11. Hill, C. G. 1977. An Introduction to Chemical Engineering Kinetics and Reactor
Design. New York, NY: Wiley.
12. Xu, S. and S. Damodaran. 1993. “Calibration of Radiotracer Method to Study Protein
Adsorption at Interfaces,” J. Colloid Interface Sci., 157: 485–490.
13. Xu, S. and S. Damodaran. 1993. “Comparative Adsorption of Native and Denatured
Egg-White, Human, and T4 Phage Lysozymes at the Air–Water Interface,” J. Colloid
Interface Sci., 159: 124–133.
14. Xu, S. and S. Damodaran. 1992. “The Role of Chemical Potential in the Adsorption
of Lysozyme at the Air–Water Interface,” Langmuir, 8: 2021–2027.
15. Xu, S. and S. Damodaran. 1994. “Kinetics of Adsorption of Proteins at the Air–Water
Interface from a Binary Mixture,” Langmuir, 10: 472–480.
16. Hsu, C. C. and J. M. Prausnitz. 1974. “Thermodynamics of Polymer Compatibility
in Ternary Systems,” Macromolecules, 7: 320–324.
© 2002 by CRC Press LLC
Engineering and Food for the 21st Century
Contents
Chapter 7: Biopolymer-Biopolymer Interactions at Interfaces and Their Impact on Food Colloid Stability
Abstract
Introduction
Theoretical Background
Experimental
Adsorption
Epifluorescence Microscopy
Results and Discussion
Thermodynamic Incompatibility
Two-Dimensional Phase Separation
Conclusions
Acknowledgment
References
