International Journal of Biological Macromolecules 38 (2006) 115–119

Impact of urea on the microstructure of commercial canola

protein–carrageenan network: A research note
F.O. Uruakpa ∗ , S.D. Arntfield
Department of Food Science, University of Manitoba, Winnipeg, Man., Canada R3T 2N2
Received 30 December 2005; received in revised form 29 January 2006; accepted 30 January 2006
Available online 20 March 2006

Abstract
Biopolymer mixtures contribute to network formation in food systems. The effects of pH and urea on the structural ordering of canola protein
isolate–␬-carrageenan (CPI–␬-CAR) gels were assessed using scanning electron microscopy (SEM). At pH 6, loosely-crosslinked CPI networks
with large empty pores were produced whereas tightly-crosslinked structures were displayed at pH 10. The structure of CPI network was greatly
improved when CPI and ␬-CAR were mixed, indicating a synergistic behaviour between the two macromolecules. Urea affected the structural
arrangement and interactions involved in the formation CPI–␬-CAR gels. Urea-treated gels showed excessive network disruption and breakdown.
The microstructural results support the involvement of disulfide bonds and noncovalent interactions in the structural arrangement of CPI–␬-CAR
networks.
© 2006 Elsevier B.V. All rights reserved.

Keywords: Biopolymers; Structure; Canola protein; Microscopy; Carrageenan

1. Introduction glycoprotein. So far, studies have focused on isolation and

Plant protein functionality in food systems is determined
by the ability of these proteins to interact with compo-
nents in that system. This may involve protein–protein and
protein–polysaccharide interactions, and the formation of a
matrix which can entrap other food components. The entrap-
ment of water in a three-dimensional protein matrix can be
an important functional property often utilized in fabricated
foods. To effectively utilize networks of this type, it is neces-
sary to understand the type and role of interactions in network
formation. This information is important if alternate proteins
(e.g. canola proteins) are to be incorporated into existing food
products.
Canola meal, a major by-product of the vegetable oil indus-
try in Canada, has immense potential in human nutrition as
protein content of 40% has been reported [1]. Thermal gela-
tion of the 12S canola globulin derived from canola meal
has been studied by Léger and Arntfield [2] while Gill and
Tung [3] reported the heat-induced gelation of 12S rapeseed
∗ Corresponding author. Tel.: +1 204 474 9621; fax: +1 204 474 7630.
E-mail address: Ojiugou@yahoo.com (F.O. Uruakpa).
evaluation of laboratory-prepared canola proteins. Dynamic
rheological testing and response surface optimization model
were used to characterize and establish optimum gelling con-
ditions (15%, w/v; CPI; pH 6; 0.05 M NaCl; 3%, w/v ␬-CAR)
for commercial canola protein isolate–␬-carrageenan (CPI–␬-
CAR) mixtures [4]. Gels prepared at pH 6 showed high G
(97,465 Pa) and low tan δ (0.15) values. High G values demon-
strate stronger intermolecular network and increased inter-
actions between protein–protein and protein–polysaccharide
molecules, while low tan δ values indicate a more elastic
network.
Interactions responsible for gelation are also of concern. The
role of hydrogen bonding in amylase gelation was demonstrated
using intermolecular hydrogen bond breaking agents such as
urea [5]. Surel and Famelart [6] showed that acid milk gels
treated with urea had decreased firmness, and noted that a bal-
ance between disulfide bonds and other linkages was required
for gel strength and firmness.
In an effort to determine the molecular forces that contribute
to CPI–␬-CAR network formation, the environment in which the
gelation took place was selectively altered to identify the role
of individual molecular forces. The effects of pH and urea were
used to evaluate the role of covalent and noncovalent interactions

0141-8130/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijbiomac.2006.01.016
116 F.O. Uruakpa, S.D. Arntfield / International Journal of Biological Macromolecules 38 (2006) 115–119
in gelation of CPI–␬-CAR mixtures. The present study aims
to characterize the structural arrangement of CPI–␬-CAR gels
using scanning electron microscopy (SEM) and determine the
role of molecular interactions in network formation. Biopolymer
incompatibility, self-association and interbiopolymer complex-
ing have been shown to contribute to synergistic and antagonis-
tic effects of food formulation [7]. In a food system contain-
ing mixed biopolymers, the interactions between various con-
stituents need to be well balanced so that a stable system evolves.
2. Materials and methods
2.1. Source of materials
Commercial CPI was purchased from BMW Canola, Win-
nipeg and used without further purification. The ratio of 2S
to12S/7S proteins of the CPI was 3:97. Proximate analysis [8]
of the CPI sample indicated a protein content of 87% (N × 5.7),
0.7% oil, 2% ash, 5.9% moisture and 4.4% total carbohydrate
(determined by difference). Commercial grade ␬-carrageenan
(␬-CAR; C-1013) that contains predominantly ␬- and lesser
amounts of ␭-carrageenan, was purchased from Sigma Chemi-
cal Co. (St. Louis, MO). The exact ␬-carrageenan concentration
was not available, however, the manufacturer indicated that
it was a mixture of the following cations: K+ (10.4%), Ca2+
(2.3%) and Na+ (0.9%). Urea (U-15; Lot 863571) was purchased
from Sigma Chemical Co. (St. Louis, MO). All other chemicals
such as NaCl (BP358-212; Lot 028091), HCl (A144-225; Lot
296220), and NaOH (BP359-212; Lot 974661) were certified
reagent grade (Fisher Scientific Co., Nepean, ON).
2.2. Sample preparation
Dispersions of CPI–␬-CAR mixtures and CPI alone at pH 6
and 10 were prepared under optimum gelling conditions estab-
lished in a previous rheological study [4]. Each mixture was
stirred for approximately 1 h at room temperature or until a com-
plete dispersion of the mixture was achieved. For interaction
studies, dispersions of CPI–␬-CAR were treated with (15% w/v
CPI, 3% ␬-CAR, 6 M urea, pH 6) or without (15% w/v CPI,
3% ␬-CAR, 0.05 M NaCl, pH 6) urea prior to heat treatment.
Samples were adjusted to pH 6 with 1 M NaOH or 1 M HCl. To
ensure that pH was maintained, samples were allowed to equili-
brate for 30 min at room temperature and the pH was rechecked
prior to further testing.
2.3. Determination of optimum gelation conditions
The method described by Uruakpa and Arntfield [4] was
used to determine the optimum gelling conditions (pH 6, 15%
w/v CPI, 3% w/v ␬-CAR, 0.05 M NaCl) for CPI–␬-CAR mix-
tures. Dynamic rheological testing was conducted to charac-
terize different combinations of CPI–␬-CAR gels following a
full factorial model (Design-Expert® Software, Stat-Ease Inc.,
Minneapolis, MN). The factors were pH, CPI, NaCl and ␬-
CAR concentrations. Storage modulus (G ) and loss tangent
(tan δ = G /G ) were measured. Using these data, model fitting
was performed using the response surface optimization model
(Design-Expert® Software) to determine the optimum condi-
tions for maximizing G and minimizing tan δ to produce strong
and elastic gels.
2.4. Gel preparation and microstructural assessment
The protocol of Arntfield et al. [9] was used with the following
modifications. Using the optimum conditions stated above, heat-
induced CPI and CPI–␬-CAR gels were prepared by heating
samples in closed stainless steel vials from 25 to 95 ◦ C at a rate
of 2 ◦ C/min. At the end of the heating regime, samples were held
at 95 ◦ C for 5 min and cooled to room temperature in an ice bath.
Gels were removed from the vials, frozen, freeze-dried, then
fractured and mounted on a specimen stub with a double-sided
adhesive tape. The samples were sputter-coated (Edwards Sput-
ter Coater, model S150B) under vacuum with gold/palladium
(75/25). Scanning electron microscopy measurements were con-
ducted on a JEOL JSM-5900LV SEM (JEOL SEM, Japan),
operating at an acceleration voltage of 10 kV. Scanning electron
micrographs of each gel at ×3000 magnification were obtained.
All samples were examined in duplicate.
2.5. Sodium dodecyl sulphate polyacrylamide gel
electrophoresis (SDS-PAGE)
SDS-PAGE was carried out as described by Aluko and
McIntosh [10] with some modifications. Gel electrophoresis of
reduced and non-reduced canola protein isolate were run on
3% stacking and 12% separating gels using the Hoefer Sepa-
ration and Control Unit (Hoefer Scientific Instruments-SE600,
CA) according to the manufacturer’s instructions. Samples were
prepared for non-reduced SDS-PAGE by mixing 10 mg CPI,
1 mL deionized water and 0.4 mL stock solution (20 mL glyc-
erol, 12.5 mL stacking gel buffer solution (1 M Tris–HCl buffer
solution at pH 6.8), 24.1 mL deionized water, 4 g SDS and 20 mg
pyronin Y). Samples were heated in microcentrifuge tube at
90 ◦ C for 3 min, cooled at room temperature for approximately
2 h (or until sample dissolved), and centrifuged at 15,000 × g
for 3 min. Aliquots (4, 8, 12 ␮L) of the supernatant were loaded
onto the gel. Reduced samples were prepared by adding 5%
(v/v) 2-mercaptoethanol (ME) to an aliquot of the supernatant
from the 4 g SDS extraction, and aliquots (4, 8, 12 ␮L) were
loaded onto the gel. Standard proteins (SigmaMarker, M-4038,
Lot 122K9283; Sigma Chemical Co., St. Louis, MO) sample was
prepared according to manufacturer’s instructions and 12 ␮L
was loaded onto the gel. Separation of samples was done on a
16 cm × 18 cm gel for 4 h at 40 mA.
3. Results and discussions
3.1. Structural arrangements of canola protein isolate
networks
The SEM images of 15% (w/v) CPI gels treated with 0.05 M
NaCl at pH 6 and 10 are shown in Fig. 1A and B, respectively.
The influence of pH on the protein gel was reflected in the
 F.O. Uruakpa, S.D. Arntfield / International Journal of Biological Macromolecules 38 (2006) 115–119 117

Fig. 1. SEM images of canola protein isolate (CPI) gels (0.05 M NaCl, 15% w/v CPI) prepared at pH 6 (A) or pH 10 (B). Part (C) shows CPI–␬-carrageenan gel
network prepared under optimum gelling conditions (pH 6, 0.05 M NaCl, 15% w/v CPI, 3% w/v ␬-CAR) while part (D) displays CPI–␬-carrageenan gel network
treated with urea (6 M urea, 15% w/v CPI, 3% w/v ␬-CAR, pH 6). Magnification: ×3000.

differences in network structures. At pH 6 (Fig. 1A), a loosely
crosslinked protein network with large and small empty pores
that entrapped the liquid components was seen. The network
displayed a porous and spongy texture with regions of good
and poorly crosslinked structures. At pH 10, the gel system
had small pores that were evenly distributed throughout the
network, suggesting that the liquid phase was completely
entrapped in a continuous three-dimensional network (Fig. 1B).
The resultant network had properly crosslinked structures. In a
study with 12S rapeseed glycoprotein gels using SEM, Gill and
Tung [3] reported a decrease in pore size as pH was increased
from 6 to 10. According to Arntfield [11], it is possible that
the range for the attractive–repulsive electrostatic interaction
balance necessary for a properly crosslinked three-dimensional
network formation was maximized at alkaline pH value, so
that a good network was observed. A balance between charge
repulsion and the potential for interaction (mainly due to hydro-
gen and hydrophobic interactions) is critical in determining
gel properties [12]. Excessive attractive forces or insuffi-
cient charge repulsion results in aggregation rather than gel
formation.
3.2. Structure of canola protein–carrageenan network
Fig. 1C shows the structural arrangement of CPI–␬-CAR
gels prepared at optimum conditions (pH 6, 3% ␬-CAR, 15%
CPI, 0.05 M NaCl). The CPI–␬-CAR gel is characterized by
properly crosslinked and tightly-packed CPI–␬-CAR strands,
indicating a very strong and elastic network. Comparing the
SEM image of CPI alone (Fig. 1A) to that of CPI–␬-CAR
mixture (Fig. 1C) showed that the structure of the CPI net-
work was greatly improved when the two biopolymers (CPI and
␬-CAR) were mixed to form a gel (Fig. 1C). The improved net-
work structure due to CPI and ␬-CAR electrostatic complexing
(shown as properly formed network) suggests a synergistic inter-
action between CPI and ␬-CAR. These microstructural findings
were corroborated by the rheological data previously reported
[4].
3.3. SDS-PAGE: polypeptide composition of canola protein
isolate
The results of the SDS-PAGE assessment of canola protein
isolate in the presence and absence of 2-mercaptoethanol are
shown in Fig. 2. CPI samples without ME had polypeptides
with molecular weights (MW) ranging from 17,000 to 116,000.
There were four major polypeptides in the non-reduced samples
with MW of 17,000 (E), 21,000 (D), 27,000 (C) and 45,000 (B).
Similar results [13] showed that the 12S globulin from rapeseed
contains the following four polypeptide chains differing by their
MW in the SDS-electrophoresis: 18,500 ± 800, 21,100 ± 500,
26,800 ± 900 and 31,200 ± 1600. Thus the ‘C, D and E’ bands
are tentatively identified as components of the 12S globulin
fraction of the canola protein isolate (Fig. 2). Furthermore,
the protein profile of canola meals analyzed by non-reducing
SDS-PAGE showed that four major polypeptides bands (MW
118 F.O. Uruakpa, S.D. Arntfield / International Journal of Biological Macromolecules 38 (2006) 115–119
Fig. 2. SDS-PAGE compositions of canola protein isolate (CPI) under non-
reducing (−ME) and reducing (+ME) conditions. Lanes 1: 4 ␮L sample each;
Lanes 2: 8 ␮L sample each; Lanes 3: 12 ␮L sample each; Lane 4: stan-
dard proteins. Estimated relative molecular weights: A ≈ 52000; B ≈ 45000;
C ≈ 27000; D ≈ 21000; E ≈ 17000; F ≈ 12000; G ≈ 3000. The gel was stained
with Coomassie brilliant blue.
of 16,000, 18,000, 30,000 and 53,000) were prominent in all
samples [10].
A polypeptide with a molecular weight of 52,000 has been
shown to be present in the purified 12S preparation from Bras-
sica napus seeds [14]. The presence of this polypeptide (‘A’
band in Fig. 2) in the CPI sample examined further suggests
that 12S globulin constitutes part of the proteins extracted
by SDS-electrophoresis in the present study. In the CPI sam-
ple treated with ME, polypeptide bands with MW of 17,000
(band E), 52,000 (band A) and molecular weights above dis-
appeared, indicating the presence of disulfide bonds in the
native protein molecules (Fig. 2). In previous studies [10,14]
this 52,000 protein band of rapeseed 12S globulin has been
shown to disappear when SDS-PAGE was run in the pres-
ence of ME. The protein composition of the samples treated
with ME also showed two additional polypeptide bands with
estimated MW of 3000 and 12,000 (bands G and F, respec-
tively) that were not present under non-reducing conditions
(Fig. 2). The cleavage of disulfide bonds (by ME) respon-
sible for subunit association may have resulted in dissocia-
tion of the major polypeptides into smaller components. This
may account for the two additional polypeptide bands (G and
F) obtained in the presence of ME. It is possible that the
two additional bands may be associated with any or all the
four major polypeptide chains in CPI through disulfide bond-
ing. The presence of other polypeptide bands indicates that
commercial CPI is not a pure 12S globulin but a mixture of
globulins.
3.4. Effect of urea on canola protein–carrageenan network
structure
Standard denaturants such as urea and guanidine hydrochlo-
ride are hydrogen bond formers [15]. As a competitive hydrogen
bond former, urea is capable of strong interactions with water,
which alters the structure of the aqueous phase around the pro-
tein (and polysaccharide) molecule and increases the solubility
of hydrophobic amino acids. As a result, urea can disrupt both
hydrogen bonds and hydrophobic interactions involved in the
maintenance of protein structure. The CPI–␬-CAR gel treated
with urea (Fig. 1D) showed significant disorganization (with
large pores) of the cellular structures, with some portions of the
network completely disintegrated (with smeared appearance).
The 6 M urea, a hydrogen bond blocker, completely unfolds
the biopolymer molecule and prevents the formation of nonco-
valent interactions; indicating that any network formed under
this condition may be due to disulfide bonds (shown to be
present in canola proteins as depicted in Fig. 2). The SEM image
of urea-treated CPI–␬-CAR gel showed some network forma-
tion, an indication that CPI has sufficient disulfide linkages to
support a gel structure. Electrophoretic pattern (under reduc-
ing conditions) of the commercial CPI demonstrated that the
polypeptides are composed of subunits held together by disul-
fide bonds (Fig. 2). Furthermore, the addition of urea into a
gelling system is designed to completely unfold the protein and
prevent the formation of noncovalent intermolecular interac-
tions. This means that any networks formed in the presence
of urea may be entirely due to disulfide bonds. The exces-
sive disorganization of the gel network implicates noncova-
lent forces (e.g. hydrophobic interactions and hydrogen bonds)
and disulfide bonds as factors in the network formation of
CPI–␬-CAR systems. McGrane et al. [5] studied the gelation
of amylase and showed that the use of intermolecular hydro-
gen bond breaking agents such as urea reduced the gel strength
significantly.
3.5. Significant findings and its implications
The CPI–␬-CAR gels prepared using the optimum gelling
conditions (15% w/v CPI, 0.05 M NaCl, 3% w/v ␬-CAR, pH
6) reported by Uruakpa and Arntfield [4] showed optimal struc-
tural ordering with properly formed networks (Fig. 1C). This
finding is significant because the microstructural data supported
the previous rheological results for the mixed gels [4] where
very strong and elastic networks (G = 97,465 Pa, tan δ = 0.15)
were produced under these conditions. The synergistic inter-
action between CPI and ␬-CAR, demonstrated by the superior
gel strength for CPI–␬-CAR mixtures, was confirmed by the
SEM image (Fig. 1C) which displayed a network structure with
densely-packed strands. The low tan δ values (indicating elastic
networks) obtained for this system [4] is a result of the exten-
sive crosslinking seen in the microstructure of the CPI–␬-CAR
gels (Fig. 1C). The CPI and ␬-CAR combination reported here
can serve as a guide for food processors when alternate plant
protein sources (e.g. CPI) are to be incorporated into existing
multicomponent food products.
 F.O. Uruakpa, S.D. Arntfield / International Journal of Biological Macromolecules 38 (2006) 115–119
4. Conclusions
There is a perceived need to characterize functionally active
plant proteins to provide alternatives to animal products. It is
demonstrated that CPI–␬-CAR complexing produced networks
with improved microstructural properties, supporting previous
improvements that have been reported for the rheological prop-
erties. This indicates that this biopolymer mixture can serve as
a structure enhancer in mixed food systems. These results also
indicate that disulfide bonds and noncovalent interactions (e.g.
hydrogen bonds, hydrophobic interactions) play significant roles
in CPI–␬-CAR gels. Knowledge of the interactions between
biopolymers will enhance the manipulation of food quality (e.g.
sensory properties) by adjusting the interaction in a desirable
way. If similar structures can be obtained at lower protein con-
centrations, due to the effect of the polysaccharides, product
costs could also be reduced. This can be applied to using canola
protein as a food ingredient as well as a component of a mixed
food product.
Acknowledgements
The authors thank the Natural Sciences and Engineering
Research Council (NSERC) of Canada for providing the funds
for this study.
119
References
[1] F. Shahidi, in: F. Shahidi (Ed.), Canola and Rapeseed: Production, Chem-
istry, Nutrition and Processing Technology, AVI-Van Nostrand Reinhold,
New York, 1990, pp. 1–10.
[2] L.W. Léger, S.D. Arntfield, Thermal gelation of the 12S canola globulin,
1993.
[3] T.A. Gill, M.A. Tung, J. Food Sci. 43 (1978) 1481–1485.
[4] F.O. Uruakpa, S.D. Arntfield, Food Hydrocolloids 18 (2004) 419–
427.
[5] S.J. McGrane, D.E. Mainwaring, H.J. Cornell, C.J. Rix, Starch/Staerke
56 (2004) 122–131.
[6] O. Surel, M.H. Famelart, J. Dairy Res. 70 (2003) 253–256.
[7] V.B. Tolstoguzov, Food Hydrocolloids 17 (2003) 1–23.
[8] AOAC, Official Methods of Analysis, Association of Official Analytical
Chemists, 15th ed., Washington, DC, 1990.
[9] S.D. Arntfield, E.D. Murray, M.A.H. Ismond, J. Agric. Food Chem. 39
(1991) 1378–1385.
[10] R.E. Aluko, T. McIntosh, J. Sci. Food Agric. 81 (2001) 391–396.
[11] S.D. Arntfield, Microstructural and Rheological Properties of Protein
Networks from Ovalbumin and Vicilin, Ph.D. Thesis, University of Man-
itoba, Winnipeg, MB, 1989, 289 pages.
[12] S.D. Arntfield, Curr. Top. Pept. Protein Res. 1 (1994) 261–273.
[13] K.D. Schwenke, B. Raab, P. Plietz, G. Damaschun, Nahrung 27 (1983)
165–175.
[14] M. Dalgalarrondo, J.M. Robin, J.L. Azanza, Plant Sci. 43 (1986)
115–124.
[15] L.G. Phillips, D.M. Whitehead, J. Kinsella, Structure-Function Prop-
erties of Food Proteins, Academic Press, San Diego, 1994, pp. 153–
169.