Full Paper
Morphology and Phase Separation of
Hydrophobic Clusters of Soy
Globular Protein Polymers
Xiuzhi Susan Sun,* Donghai Wang, Lu Zhang, Xiaoqun Mo, Li Zhu,
Dan Bolye
Protein hydrophobic interaction has been considered the most important factor dominating
protein folding, aggregation, gelling, self-assembly, adhesion, and cohesion properties. In this
paper, morphology and phase separation of hydrophobic clusters, networks, and aggregates of
soy globular protein polymers, induced by using a reducing agent (NaHSO3), are studied using
microscopic instruments. The morphology and phase separation of these hydrophobic clusters
are sensitive to protein structure and composition, pH, and ionic-strength (Im). Most of the
clusters are in spherical-shape architecture and mainly consist of hydrophobic polypeptides.
Rod-shape clusters were also observed at higher ionic strength, and mainly consist of
hydrophilic polypeptides. The ratio of hydrophobic/hydrophilic (HB/HL) polypeptides is
important to facilitate the formation of clusters in an environment with a certain pH value
and ionic strength. At HB/HL 0.8, uniform spherical clusters were observed and diameters
ranged from 30 to 70 nm. At HB/HL <0.8, large spherical
clusters were formed with diameters ranging from 100
to 1 000 nm, and at HB/HL 1.8, large hydrophobic
aggregates formed, and size of aggregates can be up
to 2 500 nm. When solid content increased from 3% to
38%, at Im 0.058 mol L1, pH ¼ 4.8, HB/HL ratio <0.8,
the networks turned into a continuous substance
that has strong adhesion to various surfaces;
at Im 0.115 mol L1, HB/HL ratio 1.8, the large aggregates became very cohesive and viscoelastic. Clear phase
separation was observed during curing between hydrophobic and hydrophilic protein polymers. Phaseseparation degree increased as HB/HL ratio increased.
X. S. Sun, L. Zhang, X. Mo, L. Zhu
Bio-Materials and Technology Laboratory, Department of Grain
Science and Industry, Kansas State University, Manhattan, KS
66506, USA
Fax: þ1-785-532-7193; E-mail: [email protected]
D. Wang
Department of Biological and Agriculture Engineering, Kansas
State University, Manhattan, KS 66506, USA
Macromol. Biosci. 2008, 8, 295–303
ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
D. Bolye
Transmission Electron Microscope Lab, Division of Biology,
Kansas State University, Manhattan, KS 66506, USA
DOI: 10.1002/mabi.200700235
295
X. S. Sun, D. Wang, L. Zhang, X. Mo, L. Zhu, D. Bolye
Introduction
It has long been known that the driving force for protein
folding is provided by water, so that the hydrophobic
clusters exclude water and are tightly packed together and
buried inside of the protein. Therefore, the surface of most
proteins in water is hydrophilic, and it is impossible for
such macro proteins with a hydrophilic surface to interact
with each other, forming a hydrophobic cluster. Once the
protein is unfolded or denatured and turned inside out,
the buried hydrophobic groups are exposed toward the
surface of the protein, which would promote proteinprotein interactions and form hydrophobic clusters.
Milk-protein, casein-colloid, aggregates can be induced
by ionic strength, pH, and temperature. They have been
studied using various techniques including FT-IR spectroscopy,[1] thermodynamic linkage analysis,[2] scattering
and turbidimetric techniques,[3] and transmission electron
microscopy and scanning electron microscopy.[4] Spherical
micelles (up to 300 nm) form at low ionic strength.[2,4]
Caseins are heterogeneous proteins containing mainly
aS1, aS2, b, and k polypeptides consisting of both
hydrophobic and hydrophilic amino acids. Caseins have
been widely used as emulsifiers in food products by adding
sodium caseinate as an emulsifying agent at neutral pH.
As pH approaches the isoelectric point, caseins become an
aggregated emulsion gel.[5] Micelle structure was also
observed in a model monomeric protein with 101 amino
acids, a mixture of a-helix and b-sheet from a ribosomal
subunit of prokaryote.[6] The monomeric protein was
unfolded using detergent sodium dodecyl sulfate and
stimulated by NaCl. The protein formed a spherical micelle
structure at NaCl below the critical pH and a cylindrical
structure at NaCl above the critical pH. Temperature also
promoted protein micelle structure formation.[4,6]
Soy proteins have been recently considered as potential
alternative polymers for environmentally friendly adhesives.[7] Like many globular proteins, soy proteins are made
up of the 20 various amino acids forming primary,
secondary, tertiary, and quaternary structures. Soy proteins consist of many polypeptide chains. Glycinin and
conglycinin proteins are major subunits of soy proteins.[8,9]
Scheme 1. A schematic description of hydrophobic clusters and aggregates of soy globular proteins. (A) Native soy protein, conglycinin
contains a and a0 hydrophilic polypeptides and b hydrophobic polypeptide, which is mainly associated by hydrogen and hydrophobic
bonding; glycinin contains acidic (A) hydrophilic and basic (B) hydrophobic polypeptides, which are associated by disulfide bonds as well as
hydrogen and hydrophobic bonding; and all these polypeptides are arranged based on their structure and biological function during
synthesis in the soy seed. Once reducing agent is introduced, those bonds are broken forming individual polypeptides. Several scenario
could occur depending on the force difference (DF) between hydrophobic force (HBF) and electrostatic force (ESFrepell or ESFattract). (B)
HBF < ESFrepell; (C) HBF ¼ ESFrepell; (D) HBF > ESFrepell; (E) HBF ESFrepell; and (F) ESFrepell > 0, and ESFattract > 0.
296
Macromol. Biosci. 2008, 8, 295–303
ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/mabi.200700235
Morphology and Phase Separation of Hydrophobic Clusters of Soy Globular Protein Polymers
Glycinin protein, with a molecular mass of 300–380 kDa,
has six major polypeptides, and each containing a pair of
acidic and basic polypeptides that are alternatively linked
by a disulfide bond, forming a hexamer.[10] Conglycinin
protein, with a molecular mass of 150–200 kDa, contains
three major polypeptides, including a, a0 , and b.[11,12] The
polypeptides of conglycinin protein are held together
primarily by hydrophobic forces and hydrogen bonding[13]
(Scheme 1A). Structure, surface properties, and conformation of soy globular protein are sensitively influenced by
environmental conditions such as pH, ionic strength, and
temperature.[14–23] Analogues of globular protein aggregates could exist but may require a different stimulation
environment. The resulting proteins may present different
properties because of their differences in structure,
composition, and molecular size.
This paper reports the morphology and phase separation of hydrophobic clusters caused by hydrophobic
globular polypeptides of soy protein induced by ionic
strength in an aqueous system. The hydrophobic cluster is
defined, in this paper, as the cluster formed by interprotein-protein interactions.
Then the pH of the solution was adjusted to 5.4 and 4.8, and
protein samples were collected at three pH levels, respectively.
The leftover SPI solution at 5.4 pH, at each Im level, was
centrifuged at 10 000 g force and 4 8C for 10 min to remove some
glycinin proteins (precipitates) that were not reduced. The supernatant at pH 5.4 was collected as sample, and then the pH of the
supernatant was adjusted to 4.8 and collected as another set of
samples. These samples were studied as comparisons with SPI
samples. The supernatant at pH 4.8, at each Im level, was
centrifuged at 10 000 g force and 4 8C for 10 min. The precipitate
of the centrifuge was collected as the protein sample with 38%
solid content.
Ionic-Strength Estimation
P
Ionic strength was estimated using the equation (Im ¼ 1/2 MB Z 2B )
based on a molality basis,[19] where S is the sum of all B ions used
in the system. ZB is the charge number of ion B, and M is the molar
concentration of ion B. In this paper, NaHSO3 was used to prepare
protein-water solutions in three different ionic strengths (0.029,
0.058, and 0.115 mol L1). The ions generated by adjusting the pH
of the system, using either sodium hydroxide (NaOH) or
hydrochloride (HCl), were neglected.
Transmission Electron Microscopy (TEM)
Experimental Part
Materials
Protein samples were isolated from defatted soy flour using the
acidic precipitation method. Soy flour with a protein dispersion
index of 90 was provided by Cargill (Cedar Rapids, Iowa, USA),
containing about 50% protein and 10% moisture, and 98% particle
size of the soy flour was <150 mm. The soy flour was dissolved in
distilled water at a 1:16 (soy flour: water) ratio at room temperature. pH of the slurry was adjusted to 9.5, and the slurry was
stirred for about 1 h at room temperature, then centrifuged at
12 000 g force at 4 8C for 10 min. The precipitates were mainly
carbohydrate-based materials and were discarded; the supernatant was saved as the starting material named as soy protein
isolates (SPI) for protein sample preparation. The protein contained about 55% glycinin and 27% conglycinin, and 18% other
polypeptides components, as estimated using the sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) method.
The SPI had 90% purity determined using the SDS-PAGE method,
and 3% solid content determined using oven drying method.
A.C.S.-certified sodium bisulfite (NaHSO3) was used as a reducing
agent and ionic- strength stimulator. The sodium bisulfite in solid
form was the product of Fisher Chemicals (Fisher Scientific,
Houston, Texas, USA), with 56.7% SO2, 0.005% chloride, 0.0003%
heavy metal (Pb), 0.003% water insoluble, and 0.0005% iron.
Protein Samples Preparation
NaHSO3 was added to the SPI solution at various amounts to
produce four ionic strengths (Im): 0.000, 0.029, 0.058, and
0.115 mol L1. pH of the protein solution was kept at 9.5, and
the solution was stirred for about two hours at room conditions.
Macromol. Biosci. 2008, 8, 295–303
ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
A model CM 100 (FEI Company, Hillsboro, Oregon, USA) was
operated at 100 kV. Protein and polypeptide samples were diluted
to 1% with distilled water and sonicated for 3 min in an L&R 320
ultrasonic stirrer (L&R Manufacturing Company, Keary, N.J., USA).
Samples were absorbed for approximately 30 s at room
temperature onto Formvar/carbon-coated 200-mesh copper grids
(Electron Microscopy Sciences, Fort Washington, Penn., USA), and
stained with 2% (w/v) uranyl acetate (Ladd Research Industries,
Inc., Burlington, V.T., USA) for 60 s at room temperature before
being viewed by TEM.
Light Microscopy (LM)
Images were taken on an Axioplan 2 MOT research microscope
(Carl Zeiss, Inc., Thornwood, N.Y., USA) equipped with a Zeiss
Axiocam HR digital camera, a fully motorized stage with
mark-and-find software, plan neofluor objectives (1.25/0.035,
10/0.3, 20/0.5, 40/0.75, 40/1.3 oil), plan apochromat
objectives (63/1.4 oil, 100/1.4 oil), an achroplan objective
(4/0.1), differential contrast interference (DIC), phase contrast
(ph), dark field, bright field, and Axiovision 3.1 software with
interactive measurements and D deconvolution modules.
A small drop of a fresh protein sample at 38% solid, pH ¼ 4.8,
and various levels of Im was deposited onto a 300 100 glass slide
(Fisher Scientific), with and without spreading force. The protein
was allowed to set at room conditions for 5 min. Thickness of the
protein sample drop without spreading force was about 1.5 mm,
with a diameter of 10 mm. Thickness of the protein film with
spreading force was about 0.5 mm. In all samples, protein became
completely cured because of water evaporation. DIC images were
taken at various times and magnifications. LM images reported in
this paper were only those samples derived from the SPI solutions
treated with NaHSO3 at 38% solid content.
www.mbs-journal.de
297
X. S. Sun, D. Wang, L. Zhang, X. Mo, L. Zhu, D. Bolye
Sodium Dodecyl
Sulfate-Polyacrylamide
Gel Electrophoresis
(SDS-PAGE)
To estimate the polypeptide composition, SDS-PAGE was performed using a discontinuous
buffer system on a 12% separating gel and 4% stacking gel, as
described by Laemmli.[20] Protein
samples were mixed with SDSPAGE sample buffer solution containing 5% b-mercaptoethanol, 2%
SDS, 25% glycerol, and 0.01%
bromphenol blue. Approximately
5 mg of protein sample were
loaded per well. The gel electrophoresis was carried out at a
100 V constant voltage. The gel
was stained with 0.25% Coomassie Brilliant Blue-R250 and
destained with solution containing 10% acetic acid and 40%
methanol. Molecular weightmarker proteins were run along
with the samples. The percentage
of polypeptide component was
estimated by analyzing the gel
image with Kodak 1D Image
Analysis software, version 4.6
(Eastman
Kodak
Company,
Rochester N.Y., USA).
Figure 1. TEM images of soy protein isolate at 3% solid content: A) Control soy protein sample (SPI at
Im ¼ 0.0, and pH ¼ 7.0); SPI treated with NaHSO3 at Im ¼ 0.029 at B) pH ¼ 9.5, C) at pH ¼ 5.4, and D) at
pH ¼ 4.8; SPI treated with NaHSO3 at Im ¼ 0.058 at E) pH ¼ 9.5, at F) pH ¼ 5.4, and at G) pH ¼ 4.8; SPI
treated with NaHSO3 at Im ¼ 0.115 at H) pH ¼ 9.5, at I) pH ¼ 5.4, and at J) pH ¼ 4.8.
X-Ray Diffraction
An X-ray scattering method was used to determine crystal and
amorphous phases of the protein samples. A Philips APD 3520,
wide-angle X-ray diffractometer was used. A voltage of 35 kV,
current of 20 mA, and a curved crystal graphite monochromator
(l ¼ 0.154 nm) were employed. The protein samples were
freeze-dried and ground into powder with particle size less than
100 mesh. The powder sample was continuously scanned from 108
to 358 (2u) with a speed of 28 (2u) per min.
Results
TEM Images of Soy Protein Polypeptides in
Aqueous System
SPI Solution with 3% Solid Concentration
For the control sample (SPI) without NaHSO3 treatment,
irregular clusters were observed, formed by a mixture of
smaller spherical clusters and rod-shape clusters
(Figure 1 A). Diameter of the clusters was approximately
50 to 70 nm. At low ionic strength, Im ¼ 0.029 mol L1,
298
Macromol. Biosci. 2008, 8, 295–303
ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
treated with NaHSO3 reducing agent, small uniform
spherical clusters were observed at pH ¼ 9.5. These clusters
had an average diameter of nearly 30 nm (Figure 1B). As pH
decreased, the number of clusters decreased, but diameters
became larger. For example, at pH ¼ 5.4, the average
diameter of the cluster was about 45 nm, and while at
pH ¼ 4.8, these clusters became aggregated into huge
irregular spherical shape clusters with diameters ranging
from 200 nm to 2 000 nm (Figure 1D). In the case of
pH ¼ 5.4, ‘‘Free-walking’’ polypeptides (Figure 1C arrow)
existed, mainly the hydrophilic polypeptides. Distance
between the cluster lines was from 50 nm to 150 nm.
However, at pH ¼ 4.8, these free-walking polypeptides
disappeared, which attracted to the large hydrophobic
clusters. The smaller, dark-color spherical spots observed
on the surface of the large clusters, with an average
diameter of about 50 nm (Figure 1D, arrow), could be those
hydrophilic polypeptides.
At medium ionic strength, Im ¼ 0.058 mol L1, treated
with NaHSO3 reducing agent, no cluster was formed
at pH ¼ 9.5 (Figure 1E). As pH decreased, for example at
DOI: 10.1002/mabi.200700235
Morphology and Phase Separation of Hydrophobic Clusters of Soy Globular Protein Polymers
pH ¼ 5.4, a uniform chain network-like structure was
observed (Figure 1F). While at pH ¼ 4.8, uniform spherical
clusters were formed with an average diameter of about
25 nm. Distance between clusters was from 10 to 40 nm.
At strong ionic strength, Im ¼ 0.115 mol L1, treated
with NaHSO3 reducing agent, uniform spherical clusters
were observed at pH ¼ 9.5 (Figure 1H), and rod-shape
clusters were in between these spherical clusters (Figure 1F
arrow). The rod-shape clusters were mainly hydrophilic
polypeptides. Such morphology remained the same at
pH ¼ 5.4 (Figure 1I), except that the rod-shape clusters
were not as clear as those at pH ¼ 9.5. At pH ¼ 4.8, large,
non-uniform aggregates were observed again (Figure 1 J)
similar to those aggregates shown in Figure 1D. However,
the diameter of these aggregates was larger than those in
Figure 1 D, ranging from 500 nm to 5 mm.
SPI Solution Centrifuged at pH 5.4
The protein (3% solid content) in the supernatant, pH
remained 5.4, formed uniform spherical clusters in the
presence of a reducing agent (NaHSO3). As Im increased, the
number and diameter of the clusters increased. At
Im > 0.058 mol L1, large-diameter clusters aggregated.
All clusters had dark centers and light rings surrounded by
dark rings (Figure 2 A), which indicated the spherical
clusters collapsed after coating onto the copper grids.
Diameter of samples treated at Im ¼ 0.029 mol L1was
from 70 to 500 nm, and diameter was from 100 to 1 500 nm
at Im ¼ 0.058 mol L1. Clusters at Im ¼ 0.029 mol L1 had
a clear boundary with their environment. At Im ¼
0.058 mol L1, although the clusters presented a clear
line of the sphere (arrows) (Figure 2B), numerous polypeptides were tightly attached to the cluster, forming a
continuous boundary surrounding the cluster (arrows)
(Figure 2 C). More interesting, at Im ¼ 0.115 mol L1, these
large clusters formed large aggregates (arrows) (Figure 2D),
and many small spherical clusters (arrow I) and irregular
aggregates (arrow II) were attached to the surface of the
large aggregates (Figure 2E). Size of the large aggregates
was up to 2 500 nm.
At 3% solid, pH ¼ 4.8, network-clusters formed.
Figure 3 A (arrow) is a typical network structure observed.
The degree of the network increased as Im increased from
0.029 to 0.058 mol L1, and the network structure was
uniform. But at Im > 0.058 mol L1, aggregates of large
clusters were favored over a network of small clusters. The
complex structure was not uniform at Im ¼ 0.115 mol L1.
Some spherical clusters remained intact (Figure 3B) or
coupled with other clusters.
As solid content increased, for example at 38% solids,
pH ¼ 4.8, with NaHSO3 used as a reducing agent, chains of
clusters formed at Im < 0.115 mol L1 (Figure 3 C, arrow).
The chain was made up of numerous polypeptides clusters
with diameters ranging from 5 to 25 nm. But aggregate
formation was still favored over chain formation at
Im 0.115 mol L1 (Figure 3D). Many large spherical
clusters remained as tightly aggregated balls with
diameters ranging from 100 to 600 nm (Figure 3D, arrow
I), and these aggregated balls were connected with
chain-like
network-structures
(Figure 3D, arrow II).
LM Images and Phase Separation of
Cured Proteins
Figure 2. TEM images of soy protein samples at 3% solid content treated with NaHSO3
and centrifuged at A) pH ¼ 5.4 at Im ¼ 0.058 mol L1; B) and C) are enlarged images of
A); D) at Im ¼ 0.115 mol L1, and E) is an enlarged image of D).
Macromol. Biosci. 2008, 8, 295–303
ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
At pH ¼ 4.8, treated with NaHSO3, solid
content was increased from 38% to 90%
by curing at room condition. Phase
separation between hydrophobic and
hydrophilic polypeptides was clearly
observed (Figure 4). At Im ¼ 0.0 mol L1,
protein molecules were individually
uniformly packed next to each other,
forming a typical uniform brittle structure, and some cavities were also present, due to local protein aggregates, but
no clear phase separation occurred. At
Im 0.115 mol L1, spider-web-ordered
structure was observed (Figure 4A). The
degree of ordered spider-web structures
became stronger as Im increased. The
average diameter of the spider-web
structures also increased as Im increased.
www.mbs-journal.de
299
X. S. Sun, D. Wang, L. Zhang, X. Mo, L. Zhu, D. Bolye
during protein curing upon water removal,
and spherical clusters are formed by mainly
hydrophobic polypeptides. Therefore, the
line phase should contain mainly hydrophilic polypeptides, which are squeezed out
due to hydrophobic aggregation during
curing. To confirm this conclusion, the cured
protein gel was soaked in distilled water at
room conditions for 48 h; the protein gel
remaining on the glass surface was examined by using LM, and was then collected for
SDS-PAGE analysis. A LM image showed that
the majority of the line phase was dissolved
into water, leaving cavities (Figure 4C).
SDS-PAGE results confirmed that proteins
in the confined area contained both hydrophobic and hydrophilic amino acids
(Figure 5), but the percentage of the hydrophilic component was significantly reduced
(Table 1). a0 and a polypeptides are considered hydrophilic polypeptides; and their
contents were reduced from 13.3% and 9.6%
to 4.1% and 4.6%, respectively. The acidic
polypeptide is also considered a hydrophilic polypeptide, and its content was
reduced from 25.7% to 7.3%. These results
confirmed that the line phase was mainly
constructed by hydrophilic polypeptides.
Figure 3. TEM images of soy protein samples at 3% solid content: treated with
NaHSO3 centrifuged at pH ¼ 5.4 and then the pH was adjusted to 4.8. A) at
Im ¼ 0.058 mol L1, B) at Im ¼ 0.115 mol L1; images of soy protein samples
at 38% solid content, treated with NaHSO3 at pH ¼ 4.8 at C) Im ¼ 0.058 mol L1,
and at D) Im ¼ 0.115 mol L1.
When the protein sample was spread into a thin film,
the spider-web pattern was totally destroyed and turned
into a randomly oriented net pattern (Figure 4B).
According to theories of hydrophobic interactions,[21–24]
phase separation is caused by hydrophobic aggregation
X-Ray Diffraction
Like most globular proteins, soy proteins are
also, in most instances, in amorphous form.
X-Ray diffraction patterns of both the
control and sodium bisulfite-treated samples were almost identical, and no obvious crystal peaks
were observed. Both control and treated protein samples
were a mixture of polypeptides and were tested in powder
form. This result indicates that none of the treated
polypeptide formed crystal-like structures during curing.
Figure 4. Light microscope images of soy protein gel cured at room temperature on a glass surface, with the protein samples at 38% solid
content treated with NaHSO3 at Im ¼ 0.115 mol L1. A) A free drop of the protein sample with a little pressure; B) a free drop spread into
about a 0.5-mm film, and C) the image of the cured protein gel of image B) after soaking in distilled water for 48 h.
300
Macromol. Biosci. 2008, 8, 295–303
ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/mabi.200700235
Morphology and Phase Separation of Hydrophobic Clusters of Soy Globular Protein Polymers
Figure 5. An SDS-PAGE image of soy protein samples: A) left side
molecular marker; B) soy protein isolate control sample; C) soy
protein sample at 38% solid content treated with NaHSO3 at
Im ¼ 0.115 mol L1; D) soy protein gel cured at room temperature
on a glass surface, left on the glass, and collected after soaking in
distilled water for 48 h.
Discussion
Sodium bisulfite, as a reducing agent, breaks down the
sulfur-sulfur covalent bonds in the protein. When this
happens, protein conformation changes, and the hydrogen
Table 1. Polypeptide contents in percentages analyzed by using
SDS-PAGE and band-intensity scan methods.
Name of
polypeptide
Soy protein fraction distribution
%
Treated with
NaHSO3a)
Treated with NaHSO3
plus curing and
soakingb)
a(
13.3
4.1
a
9.6
4.6
b
8.7
14.6
Acidic
25.7
7.3
Basic
21.5
29.2
a)
Soy protein samples treated with NaHSO3 at Im ¼ 0.115 mol LS1
(Figure 5, line C); b)Soy protein sample treated with NaHSO3 at Im
0.115 mol LS1, cured gel remained on the glass after soaking for
48 h (Figure 5D).
Macromol. Biosci. 2008, 8, 295–303
ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
bonding structure would also be changed. When there are
more ions than needed for reduction, electrostatic bonding
is interrupted. All these changes induce conformation
change; in consequence, hydrophobic bonding is also
interrupted among those polypeptides, resulting in a
mixture of many free polypeptides in the solution, such as
a, a0 , b, and acidic and basic polypeptides. Each polypeptide
has its own isoelectric point (pI) (Table 2).[25,26] a, a0 , and
acidic polypeptides are considered hydrophilic proteins,
and b and basic polypeptides are considered hydrophobic
proteins.[27–30]
Hydrophobic force (HBF), related to protein structure
and composition, is a major driving force to form
clusters;[3,4] and electrostatic force (ESF) favored (ESFat[31,32]
or disfavored (ESFrepell) hydrophobic interaction
tract)
depending on protein structure, composition, pH, and ion
contents in the environment. We propose a model
(Scheme 1) that would help in understanding the
phenomena observed in this study. At a given protein
structure and composition, when HBF is smaller than
ESFrepell, a uniform mixture of polypeptides exists
(Scheme 1B). The untreated SPI presented in Figure 1 A
belongs to this scenario. At HBF > ESFrepell, clusters would
form as shown in Scheme 1 D, and the cluster size
increased as the force difference (DF) (DF ¼ HBF-ESFrepell)
increased. Meanwhile, the number of clusters decreased
(Scheme 1E). This phenomenon was also observed in this
study. The soy protein treated with NaHSO3 formed
uniform hydrophobic clusters (Figure 1B and C; Figure 2A),
and the number and size of these clusters are significantly
affected by Im and pH as well as hydrophobic polypeptides
content.
At HBF ¼ ESFrepell, a chain-like network would form as
proposed in Scheme 1C. The network structure presented
in Figure 3A is considered to fit this model. As pH
approached to 4.8, which is close to the isoelectric point of
SPI (pI ¼ 4.5), positive charge on the surface of protein
molecules increased that would make the ESFrepell equal or
close to HBF, resulting in a chain-like network. However,
when ESF becomes an attractive force, ESFattract favors
hydrophobic interaction. In this case, protein aggregates
formed as proposed in Scheme 1F, which was observed in
this study as shown in Figure 2D.
The native protein structure described in Scheme 1A
follows the theory proposed by Van Holde.[33] The cluster
formation phenomena described in Scheme 1C and D are in
agreement with the thermodynamic theory proposed by
Anfinsen.[23] The model proposed in Scheme 1 can be a
generic model for globular proteins. Proteins with both
hydrophobic and hydrophilic components, as well as
charged amino acids, have to be first unfolded and
disassociated if the protein contains more than one
polypeptide. The unfolding method should include physical, chemical, and enzymatic methods. At a given protein
www.mbs-journal.de
301
X. S. Sun, D. Wang, L. Zhang, X. Mo, L. Zhu, D. Bolye
Table 2. Isoelectric point of soy polypeptides, diameters of clusters, and SDS-PAGE estimated contents (%) of polypeptides of soy protein
isolates (SPI) treated with NaHSO3 at Im ¼ 0.029, 0.058 mol L1, and 0.115 levels; where 7S stands for conglycinin, 11S for glycinin, HB for
hydrophobic polypeptides, and HL hydrophilic polypeptides. Polypeptides contents were estimated based on the percentage of 11S and 7S.
Polypeptides
Isoelectric pointa)
0.029
0.058
0.115
0.029–0.115
a
43.6
31.2
11.0
11.6
4.9
a(
33.5
34.8
8.6
12.5
5.18
b
14.0
18.1
7.9
8.4
5.6–6.0
Acidic
8.9
15.9
16.1
31.3
4.7–5.4
Basic
0.0
0.0
56.4
36.2
8.0–8.5
11S
8.9
15.9
72.5
67.5
5.4–6.4
7S
91.1
84.1
27.5
32.5
4.8
11S/7S ratio
0.1
0.2
2.4
2.1
Total HB
14.0
18.1
64.3
44.6
Total HL
86.0
82.0
35.7
55.4
D of clusterb) [nm]
0.16
0.22
1.8
0.8
70–500
100–1500
500–2500
20–60
Data from ref.[24,25]; b)D ¼ diameter.
structure and composition, hydrophobic clusters, aggregates, or a network can be achieved by manipulating pH
and Im to change the DF between HBF and either ESFrepell or
ESFattract. Properties of the clusters and network depend on
protein structure, composition, pH, and Im. At a higher
protein concentration (i.e., 38% solid shown in Figure 3C),
unfolded polypeptides can become associated into a
continuous liquid, and in another case shown in
Scheme 1F, would form into viscous cohesive substances
(i.e., 38% solid shown in Figure 3D). Proteins at each stage
described in the proposed model have their unique
properties that would be good for a specific application.
The ratio of hydrophobic and hydrophilic (HB/HL)
polypeptides is critical in controlling cluster size and
aggregation. As HB/HL increases from 0.16 to 0.22,
corresponding to Im increases from 0.029 to 0.058 mol L1,
1, cluster size increases (Table 2). At HB/HL ¼ 0.8, uniform
and small-size clusters can be obtained regardless of Im
range used in this study, though occurring at different pH
values (Figure 1). At HB/HL ¼ 1.8, proteins became large
aggregates (Figure 2D). There should be a critical HB/HL
value (Hc) around 0.8 at which the average cluster size
reaches its lowest peak. When HB/HL > Hc, protein
becomes aggregated, and aggregation degree increases
as HB/HL value increases. While, when HB/HL < Hc,
proteins form uniform clusters, and cluster size increased
as HB/HL ratio increases up to a certain value, then
decreases as HB/HL ratio approaches to Hc.
302
SPI without centrifuge
Im [mol LS1]
HB/HL ratio
a)
SPI centrifuged at pH 5.4
Macromol. Biosci. 2008, 8, 295–303
ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Conclusion
Soy protein can be unfolded by sodium bisulfite, resulting
in five major hydrophilic (acidic, a, a0 ) and hydrophobic (b
and basic) polypeptides. The hydrophobic polypeptides
formed clusters and networks. The number and diameter
of these clusters and network structure can be manipulated by changing pH and ionic strength. At higher protein
concentration, these clusters turned into chain-like continuous liquid at pH ¼ 4.8 and Im < 0.058 mol L1, which
has strong adhesion to various surfaces, while at Im 0.115 mol L1, a viscous and cohesive substance formed,
which has potential for fabricating fibers, films, and
plastics. Upon curing, hydrophobic clusters aggregated
and phase separation between hydrophobic and hydrophilic polypeptides occurred.
Hydrophobic interactions follow the global, minimumfree-energy theory when proteins are not biologically
functional, which is the major reason causing cluster
formation. While Van der Waals forces become important
at higher protein concentration to form continuous chain
like structure.
Received: September 17, 2007; Accepted: October 12, 2007; DOI:
10.1002/mabi.200700235
Keywords: globular protein; hydrophobic clusters; ionic
strength; morphology; pH; phase separation; soy protein
DOI: 10.1002/mabi.200700235
Morphology and Phase Separation of Hydrophobic Clusters of Soy Globular Protein Polymers
[1] D. M. Curley, T. F. Kumosinski, J. J. Unruh, H. M. Farrell, Jr.,
J. Dairy Sci. 1998, 81, 3154.
[2] A. Mora-Guteirrez, H. M. Farrell, Jr., T. F. Kumosinski, J. Dairy
Sci. 1993, 76, 3690.
[3] N. Vetier, S. Desobry-Banon, M. M. O. Eleya, J. Hardy, J. Dairy
Sci. 1997, 80, 3161.
[4] E. C. Needs, M. Capellas, A. P. Bland, P. Manoj, D. Macdougal,
G. Paul, J. Dairy Res. 2000, 67, 329.
[5] J. Chen, E. Dickinson, M. Edwards, J. Texture Stud. 1999, 30,
377.
[6] D. E. Otzen, Biophy. J. 2002, 83, 2219.
[7] X. Sun, ‘‘Biobased Polymers and Composites’’, R. P. Wool,
X. S. Sun, Eds., Elsevier Academic Press, New York 2005,
Chapter 10, p. 327.
[8] X. Sun, ‘‘Biobased Polymers and Composites’’, R. P. Wool,
X. S. Sun, Eds., Elsevier Academic Press, New York 2005,
Chapter 9, p. 292.
[9] J. W. Wolf, J. Agric. Food Chem. 1970, 18, 969.
[10] P. E. Staswick, M. A. Hermodson, N. C. Nielsen, J. Biol. Chem.
1984, 259, 13431.
[11] N. C. Nielsen, ‘‘New Protein Foods 5: Seed Storage proteins’’,
A. M. Altshul, H. L. Wilcke, Eds., Academic Press, Orlando, FL
1985, p. 27.
[12] V. H. K. Thanh, Biochem. Biophys. Acta. 1977, 490, 370.
[13] J. E. Kinsella, S. Damodaran, B. German, ‘‘New Protein Foods 5:
Seed Storage Proteins’’, A. M. Altshul, H. L. Wilcke, Eds., Academic Press, Orlando, FL 1985, p. 107.
[14] V. H. Thanh, K. Shibasaki, J. Agric. Food Chem. 1979, 27, 805.
[15] W. J. Wolf, D. R. Briggs, Arch. Biochem. Biophys. 1958, 76, 377.
[16] T. Mori, T. Nakamura, S. Utsumi, J. Food Sci. 1982, 47, 26.
Macromol. Biosci. 2008, 8, 295–303
ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[17] J. E. Kinsella, J. Am. Oil. Chem. Soc. 1979, 56, 242.
[18] J. Zhang, X. Y. Liu, J. Chem. Phy. 2003, 119, 10972.
[19] H. P. Lehmann, X. Fuentes-Arderiu, L. F. Bertello, Pure & Appl.
Chem. 1996, 68, 977.
[20] U. K. Laemmli, Nature 1970, 227, 680.
[21] C. J. Tsai, S. L. Lin, H. J. Wolfson, R. Nussinov, Protein Sci. 2006,
6, 53.
[22] N. Catsimpoolas, FEBS Lett. 1969, 4, 259.
[23] C. B. Anfinsen, Science 1973, 181, 223.
[24] R. L. Baldwin, Science 2002, 295, 1657.
[25] V. H. Thanh, K. Shibasaki, Biochim. Biophys. Acta. 1977, 490,
370.
[26] V. H. Thanh, K. Shibsaki, J. Agric. Food Chem. 1976, 24, 1117.
[27] T. J. Cho, N. C. Nielsen, Nucleic Acids Res. 1989, 17, 4388.
[28] N. C. Nielsen, C. D. Dickinson, T. J. Cho, V. H. Thanh, B.
J. Scallon, R. L. Fischer, T. L. Sims, G. N. Drews, R. B. Goldberg,
Plant Cell. 1989, 1, 313.
[29] N. Maruyama, T. Katsube, Y. Wada, M. H. Oh, A. P. Barba De
La Rosa, E. Okuda, S. Nakagawa, S. Utsumi, Eur. J. Biochem.
1998, 258, 854.
[30] N. Maruyama, M. Adachi, K. Takahashi, K. Yagasaki,
M. Kohno, Y. Takenaka, E. Okuda, S. Nakagawa, B. Mikami,
S. Utsumi, Eur. J. Biochem. 2001, 268, 3595.
[31] N. K. D. Kella, Y. J. Kang, J. E. Kinsella, J. Protein Chem. 1988, 7,
535.
[32] N. K. D. Kella, W. F. Barbeau, J. E. Kinsella, Int. J. Peptide Protein
Res. 1986, 27, 421.
[33] K. E. Van Holde, ‘‘Food Proteins’’, J. R. Whitaker, S. R. Tannenbaum, Eds., AVI Publishing Company, Inc., Westport, CN 1977,
Chapter 1, p. 1ff.
www.mbs-journal.de
303