- Email: [email protected]
FTIR characterization of protein–polysaccharide interactions in extruded blends
Accepted Manuscript Title: FTIR characterization of protein-polysaccharide interactions in extruded blends Author: Pedro Guerrero Joe P. Kerry Koro de la Caba PII: DOI: Reference:
S0144-8617(14)00471-8 http://dx.doi.org/doi:10.1016/j.carbpol.2014.05.005 CARP 8868
To appear in: Received date: Revised date: Accepted date:
10-3-2014 30-4-2014 5-5-2014
Please cite this article as: Guerrero, P., Kerry, J. P., & de la Caba, K.,FTIR characterization of protein-polysaccharide interactions in extruded blends, Carbohydrate Polymers (2014), http://dx.doi.org/10.1016/j.carbpol.2014.05.005 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Highlights •
Extruder parameters were improved by polysaccharide addition, facilitating extrusion
3
•
FTIR analysis showed changes in the secondary structure of protein
4
•
Conformational changes in blends explained physical properties of extrudates
Ac ce p
te
d
M
an
us
cr
ip t
2
1
Page 1 of 29
FTIR characterization of protein-polysaccharide interactions in extruded
a
BIOMAT Research Group,
cr
Pedro Guerreroa, Joe P. Kerryb, Koro de la Cabaa*
ip t
blends
us
University of the Basque Country (UPV/EHU),
an
Department of Chemical and Environmental Engineering,
b
M
Polytechnic School, Donostia-San Sebastián, Spain Food Packaging Group,
te
d
School of Food & Nutritional Sciences, University College Cork (UCC), Cork, Ireland
Ac ce p
*Corresponding author
Koro de la Caba
Escuela Universitaria Politécnica
Departamento de Ingeniería Química y del Medio Ambiente Plaza Europa 1
20018 Donostia-San Sebastián Spain E-mail: [email protected]
Telephone number: 00 34 943 017188 5
Abstract 2
Page 2 of 29
Soy protein-based blends were processed by double screw extrusion and the effects of
7
different types and contents of polysaccharides were analyzed. Although extrusion has not
8
been widely used for this type of blends, in this study it was observed that the increase in
9
polysaccharide content in blends caused a decrease in specific mechanical energy (SME),
10
facilitating extrusion process and showing the potential of this process, which is more cost
11
effective at industrial scale. In order to explain this behavior, infrared spectroscopy analysis
12
was carried out, mainly in the amide I and amide II region. Moreover, curve fitting analysis
13
showed the conformational changes produced in the blends due to the addition of
14
polysaccharides, which affected protein denaturation. These changes also affected properties
15
such as moisture content (MC) and total solubility matter (TSM). However, conformational
16
changes did not show significant effects with respect to piece density (PD) or in the
17
expansion ratio (ER) of the pellets. The quantitative analysis of the changes in the amide I
18
and II region provided novel information about the modifications produced in protein-based
19
blends modified with polysaccharides. In this context, infrared spectroscopy provided a
20
convenient and powerful means to monitor interactions between all ingredients used in the
21
blend formulation, which is of great importance in order to explain changes in the functional
22
properties of biodegradable materials used for industrial applications in food and
23
pharmaceutical industries.
24
Keywords: proteins, secondary structure, polysaccharides, interactions, FTIR.
Ac ce p
te
d
M
an
us
cr
ip t
6
25
3
Page 3 of 29
25 26
1. Introduction Proteins and polysaccharides are biomacromolecules employed as functional ingredients.
27
Mixtures of both types of biomacromolecules are often used in many technological
29
applications, including food, pharmaceutical, and cosmetics industries (Andrade-Mahecha,
30
Tapia-Blacido, & Menegalli, 2012; Archana et al., 2014; Guerrero, Etxabide, Leceta,
31
Peñalba, & de la Caba, 2014; Salarbashi et al., 2014). Protein-polysaccharide conjugates are
32
useful because such ingredient conjugates have improved functional properties, such as heat
33
and pH stability, solubility, emulsification, and gelation properties, offering lower
34
astringency and allergenicity compared to unmodified proteins (Doublier, Garnier, Renard, &
35
Sanchez, 2000; Klein, Aserin, Ben Ishai, & Garti, 2010; Rodríguez & Pilosof, 2011). These
36
properties, together with the increasing demands to control biochemical interactions at the
37
tissue/material interface, have prompted scientists to investigate these systems. Nevertheless,
38
the knowledge underpinning protein-polysaccharide interactions is still rather limited (Alves,
39
Costa, & Coelhoso, 2010; Wang et al., 2014). In the specific context of protein-based blends,
40
several factors such as protein concentration, pH, ionic strength, heating temperature, and
41
plasticizers affect the formation of the blend. It has been suggested that the blend network is
42
formed via hydrogen bonds and hydrophobic and electrostatic interactions. However, the
43
structural basis of the blend-forming process is, at present, not fully understood. No attempt
44
has been made to determine the relationship between the extent of formation of the blend
45
network and the structure of the protein in the blend, and nothing is known about the
46
molecular basis that leads to the formation of such blends.
Ac ce p
te
d
M
an
us
cr
ip t
28
The formation of blends from soy proteins and other globular proteins has been described
47 48
as a heat denaturation process. Heating alters the three-dimensional structure of proteins,
49
exposing functional peptide groups such as CO and NH and side chain polar and hydrophobic 4
Page 4 of 29
groups that are engaged in intramolecular hydrogen bonding and electrostatic interactions in
51
the native state; therefore, these functional groups become available for further
52
intermolecular interactions (Wang & Damodaran, 1991). The unfolded protein chains
53
approach each other and become linked through intermolecular interactions, leading to the
54
formation of a network that acts as the matrix for entrapping blend components such as
55
plasticizing agents. Since the formation of a blend occurs upon application of denaturating
56
conditions, it is quite possible that the denatured protein may undergo partial refolding, thus
57
regaining some secondary structure during the blending process (Gilbert et al., 2005). It is
58
conceivable that the extent of such refolding affects the number of functional groups
59
available for intermolecular interactions and thus, the formation and stability of the blend
60
network. Blends based on globular proteins are attractive materials because of their potential
61
applications as edible or biodegradable coatings and wrappings. This type of matrix is also
62
relevant for pharmaceutical and medical applications since it may be employed for the
63
encapsulation and release in vivo of bioactive substances at specific sites and for the growth
64
of cell cultures. A clear understanding of the structure and formation of blends is required to
65
control their properties. Besides these applications, globular protein-based blends also arouse
66
interest in the more fundamental aspects of chemical interactions involving their use since
67
they may be helpful in providing insights into the interactions that govern protein self-
68
assembly and protein conformation. In this context, many efforts have been devoted to the
69
optimization of the techno-functional properties of soy proteins, which have received
70
increasing attention since soy protein is a co-product of the soy oil industry and which can be
71
valorised (Félix, Martín-Alfonso, Romero, & Guerrero, 2014; Leceta, Etxabide, Cabezudo,
72
de la Caba, & Guerrero, 2014).
Ac ce p
te
d
M
an
us
cr
ip t
50
Regarding the processing of protein-based blends, the employment of extrusion is one of
73 74
the most versatile and well-established processes used to produce macromolecular and
5
Page 5 of 29
polymeric materials (Galicia-García et al., 2011). Moreover, extrusion is relatively
76
inexpensive (as the technology can be scaled to production volume and simplified for specific
77
commercial usage) and amenable to industrial scale-up. Extrusion conditions, such as high
78
barrel temperatures, high pressures and the use of low moisture during processing, bring
79
about major conformational changes in the ingredient components in extrusion blends during
80
processing. In addition, it is well-known that, among all of the different process variables that
81
need to be considered during extrusion, the rate and amount that moisture is introduced into
82
the extruder and the processing temperature employed at all critical points along the process
83
have important effects on the quality of the final extruded products. It is generally accepted
84
that proteins are denatured during the extrusion process, so these “reactive” unfolded protein
85
chains can interact with each other, and with other added components, allowing the formation
86
of a polymeric network. The role of covalent bonding on final product characteristics has
87
been well documented since disulfide bonds were found to develop during this process
88
(Fukushima & Van Buren, 1970a); however, physical interactions including hydrophobic
89
effects and hydrogen bonding are clearly also involved (Fukushima & Van Buren, 1970a;
90
Quinn, Monahan, O’Sullivan, & Langares, 2003), although the contribution of each type of
91
interaction remains unclear. Most of the studies on blends have focused on a selected number
92
of macroscopic characteristics attributable to the resulting films produced, such as stress and
93
strain at break point, permeability to gases and vapours, and moisture absorption, among
94
others. While the macroscopic properties of films made with globular proteins have been
95
extensively investigated, only a limited amount of information exists with respect to their
96
structure. Films made of soybean glycinin and wheat gliadin have been examined at a
97
molecular level using Fourier transform infrared (FTIR) spectroscopy (Mangavel, Barbot,
98
Popineau, & Gueguen, 2001; Subirade, Kelly, Gueguen, & Pezolet, 1998). These studies
99
revealed the presence of intermolecular β-sheets following heating and dehydration steps
Ac ce p
te
d
M
an
us
cr
ip t
75
6
Page 6 of 29
which underline the importance of hydrogen bonding in the formation of the protein network.
101
β-sheets have been proposed to act as junction zones in the cohesion of film, and in the
102
relationship between protein conformation and mechanical properties has been proposed
103
(Mangavel et al., 2001). Such reports underline the role that the β-sheet motif may play in the
104
structure and properties of globular protein films but more data are needed to evaluate the
105
importance of this type of structure.
cr
ip t
100
Since few data are available about the molecular events occurring during blend formation
107
employing globular proteins, the aim of this work was to describe the protein conformational
108
changes occurring during the extrusion process. The influence of heat treatment and the role
109
played by different types and contents of polysaccharides on final structure were also
110
considered. In order to elucidate the molecular basis pertaining to blend formation,
111
conformational changes in the biomacromolecules employed in this study were investigated
112
under extrusion conditions. In this investigation, infrared spectroscopy was employed to
113
obtain detailed information on the molecular and conformational modifications occurring
114
during the blending formation process since it has been well established that infrared
115
spectroscopy is a powerful method for the investigation of secondary protein structures in
116
complex systems, such as biological or food systems (Guerrero, Garrido, Leceta, & de la
117
Caba, 2013; Goormaghtigh, Cabiaux, & Ruysschaert, 1990; Mangavel et al., 2001).
118
2. Materials and methods
119
2.1. Materials
Ac ce p
te
d
M
an
us
106
120
Soy protein isolate (SPI), PROFAM 974, with 90% protein on a dry basis was supplied by
121
Lactotecnia S.L. (Barcelona, Spain). The SPI employed in this study was comprised of 5% of
122
moisture, 4% of fat and 5% of ash. It had an acid character and an isoelectric point of 4.6.
123
The polysaccharides used in this study were obtained from Aldrich (Madrid, Spain), and
7
Page 7 of 29
included gum arabic (ARA) from acacia tree (G9752), dextran (DEX) from Leuconostoc
125
mesenteroides (D5251), and carboxymethylcellulose (CMC) sodium salt (C4888). Glycerol
126
(GLY) was used as the plasticizer in this study and this was obtained from Panreac
127
(Barcelona, Spain).
128
2.2. Extruder and experimental configurations
ip t
124
A twin-screw extruder (MPF 19/25 APV Baker) was used in this study. The APV Baker
130
MPF19 extruder was designed to generate the high pressures needed to extrude a wide range
131
of products, in particular, edible/biodegradable blends. The MPF 19/25 possessed 19 mm
132
diameter barrel, a length to barrel diameter ratio of 25:1, and a barrel comprising four
133
controlled temperature zones, resulting in a die temperature of 70 to 100ºC.
134
2.3. Samples preparation
M
an
us
cr
129
SPI, glycerol and polysaccharide contents required to obtain the desired blend composition
135
were mixed. All components were mixed in a variable speed Stephan mixer, model UMC 5
137
(Stephan, UK), for 5 min at 1500 rpm. SPI was replaced (3, 6 and 9 wt %) by polysaccharide
138
to obtain SPI-polysaccharide blends, and samples were designated as ARA3, ARA6 and
139
ARA9 in the case of gum arabic, CMC3, CMC6 and CMC9 in the case of
140
carboxymethylcellulose, and DEX3, DEX6 and DEX9 for dextran. Glycerol was selected at
141
30 % (w/w), based on results obtained from previous work (Guerrero, Nur Hanani, Kerry, &
142
de la Caba, 2011). The sample without polysaccharides was designated as the control.
143
Mixtures were added into the feed hopper and mixed with water in the barrel of the twin-
144
screw extruder. All trials were carried out using a water speed of 2.94 g/min (0.15 kg/h).
145
Water was pumped directly into the extruder barrel zone 1 using a peristaltic pump (504U
146
MK, Watson Marlow Ltd.).
147
2.4. Specific mechanical energy (SME)
Ac ce p
te
d
136
8
Page 8 of 29
The specific mechanical energy was calculated using the expression (Hu, Hsieh, & Huff,
148 149
1993): Screw speed × Power (kW) × Torque (%) Max. screw speed × Throughput (kg/h) × 100
SME (kJ/kg)=
151
The extruder was operated at a constant speed of 250 rpm. The feed rate of extruder was
ip t
150
adjusted to 1 kg/h and had a 2 kW/h motor power in practice. Dies were scaled up by
153
maintaining a constant throughput per unit of orifice area. The extruder was equipped with a
154
single die of 3 mm diameter giving a throughput per unit area of 0.141 kg/h mm2.
155
2.5. Piece density (PD)
an
us
cr
152
Piece density is the density of each extrudate and was obtained by dividing the mass of
156
extrudate piece (Wpiece) by its volume (Vpiece), which was computed from its dimensions
158
(diameter, d; and length, l). The piece dimensions of ten, 2 cm-long, pieces were measured by
159
using a hand-held digimatic micrometer (QuantuMike Mitutoyo), and piece density was
160
calculated as:
161
PD =
162
2.6. Expansion ratio (ER)
π×d 2 ×l
te
4×Wpiece
Ac ce p
Vpiece
=
The expansion ratio is the ratio of the extrudate cross-sectional area to the die orifice cross-
163 164
Wpiece
d
M
157
sectional area, and was calculated as: ER =
165
d2 2 d die
166
where ddie is the die diameter and d is the extrudate diameter (average of 10 samples), which
167
was measured by using a hand-held digimatic micrometer (QuantuMike Mitutoyo).
168
2.7. Moisture content (MC) and total soluble matter (TSM)
9
Page 9 of 29
Three specimens of each sample were analyzed to determine moisture content and total
170
soluble matter values according to a method previously used in several studies of proteins
171
(Gontard & Guilbert, 1992; Kunte, Gennadios, Cuppett, Hanna, & Weller, 1997). Samples
172
were weighed (mw) and subsequently dried in an air-circulating oven at 105 ºC for 24 h. After
173
this time, samples were reweighed (m0) to determine MC values. Samples were dried as an
174
extrudate and MC values were calculated as:
cr
ip t
169
m w -m 0 ×100 mw
MC ( % ) =
176
Afterwards, samples were immersed in 30 mL of distilled water in the presence of sodium
us
175
azide (0.02%) in order to prevent microbial growth. The beakers were stored in an
178
environmental chamber at 25 ºC for 24 h with occasional gentle stirring. After this time,
179
specimens were dried in an air-circulating oven at 105 ºC for 24 h and weighed (mf). TSM
180
was expressed as the percentage of sample dry matter dissolved after 24 h immersion in
181
distilled water and determined by the following equation:
Ac ce p
183
m 0 -m f ×100 m0
te
TSM ( % ) =
182
d
M
an
177
2.8. Fourier transform infrared spectroscopy (FTIR) The IR spectra for blends were obtained using a Varian 600-IR Series FTIR equipped with
184 185
horizontal attenuated total reflectance (ATR) crystal (ZnSe). The spectra were collected in
186
absorbance mode on sample powders obtained by grinding pellets in an agate mortar and
187
placed directly onto the ATR crystal. Each spectrum is the result of the average of 32 scans at
188
4 cm−1 resolution. Measurements were recorded between 4000 and 700 cm−1. All spectra
189
were smoothed using the Savitzky-Golay function. Second-derivative spectra of the amide
190
region were used at peak position guides for the curve fitting procedure, using OriginPro 8.6
191
software.
10
Page 10 of 29
192
2.9. Statistical analysis Data were subjected to one-way analysis of variant (ANOVA) by means of a SPSS
193
computer program (SPSS Statistic 20.0). Post hoc multiple comparisons were determined by
195
the Tukey’s test with the level of significance set at P < 0.001.
196
3. Results and discussion
cr
ip t
194
Extrusion is generally defined as a process used to mix, homogenize, and shape material by
197
forcing it through a specifically designed opening. Extrusion process variables are critical
199
factors to be considered as they impact significantly upon ingredient blends, especially those
200
employing proteins, which ultimately impact on final film properties. Thermal extrusion
201
exposes proteinaceous ingredients to high temperature, high pressure, and mechanical shear,
202
which converts an ingredient like soy protein into a continuous plastic “melt”, resulting in
203
protein denaturation (Harper, 1989). Within the process, the main fractions of SPI (7S and
204
11S globulins) undergo a complex pattern of association-dissociation reactions (Cheftel, Cuq,
205
& Lorient, 1985). The effect of extrusion is to disassemble proteins and then reassemble them
206
using disulfide bonds, hydrogen bonds, and non-covalent interactions forming fibrous
207
extrudates. Therefore, extruded blends of various ingredients provide an excellent model
208
system to study the effects of extrusion parameters on protein structure and to monitor
209
protein-carbohydrate interactions, where carbohydrate ingredients have been added.
210
3.1. Effect of polysaccharides on extruder parameters and physical properties
Ac ce p
te
d
M
an
us
198
Firstly, the effect of polysaccharide content on SME values was analyzed and results are
211 212
shown in Table 1. The SME value indicates the extent of molecular breakdown that
213
mechanical force undergoes during the extrusion process, so this value is an important
214
parameter since it can influence final product characteristics (Godavarti & Karwe, 1997).
215
Since SME is the amount of mechanical energy dissipated as heat inside the material, 11
Page 11 of 29
expressed per unit mass of the material (Harper, 1989), increasing polysaccharide content in
217
this study resulted in higher heat transfer from the extruder barrel to the feed material and
218
consequently, a lower viscosity, lower shear, and lower friction during extrusion. The higher
219
the polysaccharide content (6-9 %), the lower the torque and SME (P < 0.001) observed.
220
Because of the reduction of the force required to push the mass through the die, a
221
consequential reduction in friction occurred between the raw processing materials, screw
222
shaft, and extruder barrel (Chen, Wei, & Ojokoh, 2010).
PD (g/cm3)
ER
MC (%)
TSM (%)
Control
972±20.13a
1.63±0.14a
1.12±0.14a
17.36±0.05a
31.49±0.06a
ARA3
972±29.16a
1.62±0.12a
1.13±0.07a
18.88±0.07a
31.42±0.03a
ARA6
648±15.21b
1.60±0.26a
1.14±0.16a
14.65±0.26bc
36.51±0.56b
ARA9
612±10.91c
1.59±0.34a
1.15±0.16a
13.79±0.21b
38.65±0.23b
CMC3
972±31.51a
1.63±0.19a
1.12±0.12a
18.36±0.51a
30.55±0.21a
CMC6
648±20.11b
1.65±0.12a
1.05±0.04b
15.49±0.29b
30.27±0.09a
CMC9
648±22.31b
1.66±0.06a
1.03±0.09b
15.13±0.49b
30.38±0.61a
DEX3
1044±45.51e 1.64±0.17a
1.10±0.16ab
18.70±0.51a
31.01±0.07a
DEX6
900±32.53d
1.63±0.16a
1.12±0.23a
14.53±0.32bc
32.35±0.47ab
DEX9
900±35.91d
1.61±0.24a
1.13±0.32a
12.75±0.41c
33.87±0.53b
Ac ce p
te
d
an
SME (kJ/kg)
M
us
cr
ip t
216
223
Table 1. Effect of gum arabic (ARA), carboxymethylcellulose (CMC), and dextran (DEX)
224
content on specific mechanical energy (SME), piece density (PD), expansion ratio (ER),
225
moisture content (MC) and total soluble matter (TSM).
226
a-e
227
different through the Tukey´s multiple range test.
Two means followed by the same letter in the same column are not significantly (P > 0.001)
228
During the extrusion process, protein denaturation takes place, as has been described
229
previously (Guerrero, Beatty, Kerry, & de la Caba, 2012), which leads to the decrease of 12
Page 12 of 29
electrostatic repulsions and allows intermolecular interactions and network formation to
231
occur (Song, Tang, Wang, & Wang, 2011). The nature of these interactions and the structure
232
of the network formed may affect the macroscopic properties of the blends. The reduction in
233
conformational freedom would explain the reason for the decrease (P < 0.001) in moisture
234
absorption shown in lower MC values, when polysaccharides were added (Table 1).
235
Conversely, TSM values increased (P < 0.001) when the polysaccharide content in blends
236
increased, probably due to the reduction of protein-protein interactions in the presence of
237
polysaccharides and due to the formation of a less tight polymeric network. These facts were
238
confirmed by FTIR analysis, as shown below.
239
3.2. Effect of polysaccharides on protein structure
an
us
cr
ip t
230
Firstly, the band corresponding to amide I depends on the secondary structure of the protein
M
240
backbone and it is the most commonly used band for the quantitative analysis of secondary
242
structures. Secondly, the band associated with amide II is caused mainly by C-N stretching
243
and N-H in plane bending. Therefore, alterations of these bands might clarify to what extent
244
polysaccharide addition to ingredient blends might assist in the stabilization of added protein,
245
since the progressive exposure of the proteins to polysaccharides may govern the structural
246
changes that ultimately take place during the extrusion process. In order to analyze this
247
influence, the FTIR spectra for the blends in this target IR region are shown in Figure 1.
Ac ce p
te
d
241
248 249 250 251 252 253 254
13
Page 13 of 29
0.4
ARA ARA9 ARA6 ARA3 Control
255 0.3 Absorbance
256 257 258
0.2
0.1
0.0
260
1750
1700
1650
1600
1550
1500
ip t
259 1450
-1
Wavenumber (cm )
0.4
Absorbance
263 264 265
us
0.3
0.2
0.1
0.0 1750
267
1700
1650
1600
1550 -1
Wavenumber (cm )
0.4
0.2
te
Absobance
271
d
0.3
270
0.1
Ac ce p
272
0.0 1750
273
1450
DEX DEX9 DEX6 DEX3 Control
268 269
1500
M
266
cr
262
CMC CMC9 CMC6 CMC3 Control
an
261
1700
1650
1600
1550
1500
1450
-1
Wavenumber (cm )
274
Figure 1. Effect of polysaccharide content on FTIR spectra in the amide I and II region for
275
soy protein blends with gum arabic (ARA), carboxymethylcellulose (CMC), and dextran
276
(DEX).
When polysaccharides were added, the frequencies of the bands assigned to the amide I and
277 278
amide II regions remained constant at 1628 cm-1 and 1541 cm-1, respectively. However, the
279
band assigned to carboxylate in the polysaccharides at 1587 cm-1 for CMC, 1600 cm-1 for
280
ARA, and 1643 cm-1 for DEX was not detectable in the blends with SPI, indicating
281
interactions between the protein and the polysaccharides and highlights the lower capacity of 14
Page 14 of 29
the carbohydrate molecules to form intermolecular hydrogen bonds between themselves in
283
presence of the protein (Carpenter & Crowe, 1989). Additionally, the dominant spectral
284
change observed was the decrease in the intensity observed for the amide I and amide II
285
bands, thereby indicating conformational changes occurring through interactions between
286
protein and the polysaccharides. These changes were quantitatively measured and results are
287
shown in Table 2.
cr
ip t
282
1628 cm-1
1541 cm-1
2930 cm-1
1628/2930 cm-1
1541/2930 cm-1
Control
14.624
7.649
0.168
87.048
45.530
ARA3
13.151
6.736
0.151
87.093
44.609
ARA6
12.456
5.525
0.149
83.597
37.081
ARA9
12.023
5.087
0.156
77.071
32.609
CMC3
13.666
6.896
0.157
87.045
43.924
CMC6
11.735
5.636
0.139
84.424
40.547
CMC9
10.681
4.888
0.136
78.537
35.941
DEX3
13.404
6.853
0.154
87.039
44.500
DEX6
12.579
6.143
0.152
82.757
40.414
5.830
0.161
74.155
36.211
an
M
d
te
11.939
Ac ce p
DEX9
us
Amide area
288
Table 2. FTIR absorption area values for amide I (1628 cm-1), amide II (1541 cm-1), and
289
reference (2930 cm-1) bands as they pertain to soy protein blends with different gum arabic
290
(ARA), carboxymethylcellulose (CMC), and dextran (DEX) contents. The relationship of amide I and amide II to total carbohydrate was estimated from the ratio
291 292
of the absorbances at 1628 cm-1 and 1541 cm-1 to 2930 cm-1. The well-defined band at 2930
293
cm-1 due to C-H content represents a good reference for total carbohydrate content, because
294
the number of C-H groups remains constant, irrespective of changes in the ratio of
295
polysaccharides and protein contents (Rochas, Lahaye, & Yaphe, 1986). As can be seen in
296
Table 2, the area of the normalized bands, corresponding to amide I and amide II, decreased
15
Page 15 of 29
when the content of the polysaccharide increased. This change was lower for those blends
298
containing ARA and higher for the blends containing DEX. The results indicate that
299
carbohydrates caused changes in the infrared spectrum of soy protein, decreasing amide I and
300
II band area and eliminating the carboxylate band, which would be indicative of the degree of
301
hydrogen bonding occurring between proteins and carbohydrates which are necessary to
302
provide stability to the extruded blends.
303
3.2.1. FTIR in amide I region
us
cr
ip t
297
In addition to the analysis carried out above and since the band corresponding to amide I
304
depends on the secondary structure of the protein backbone, it can also be used for the
306
quantitative analysis of different secondary structures. Specifically, protein unfolding is
307
characterized by changes in the intensity of the dominant band in the native protein at 1650
308
cm-1 (assigned to α-helix/unordered structures) and in the regions around 1615 to 1630 cm-1
309
and 1680 to 1700 cm-1 (assigned to β-sheets) (Hu, Lu, Sun, Cebe, Wang, Zhang, & Kaplan,
310
2010; Lefevre & Subirade, 2000), as shown in Figure 2.
te
d
M
an
305
0.4 Control
Ac ce p
311
0.4 ARA9
0.3
Absorbance (a. u.)
313 314 315
0.2
0.1
0.0
316
Absorbance (a. u.)
0.3
312
1720
1700
1680
1660
1640
1620
1600
0.2
0.1
0.0
1580
1720
1700
-1
1600
1580
1620
1600
1580
0.3
Absorbance (a. u.)
Absorbance (a. u.)
320
1620
DEX9
0.3
319
1640
0.4
CMC9
318
1660
Wavenumber (cm )
0.4
317
1680
-1
Wavenumber (cm )
0.2
0.1
0.2
0.1
321 0.0
0.0 1720
1700
1680
1660
1640
1620 -1
Wavenumber (cm )
1600
1580
1720
1700
1680
1660
1640 -1
Wavenumber (cm )
16
Page 16 of 29
322
Figure 2. Original spectrum (black) and curve fitting spectra of amide I for soy protein
323
(control) and soy protein blends containing 9% gum arabic (ARA), carboxymethylcellulose
324
(CMC) or dextran (DEX).
ip t
Therefore, it appears wiser to consider the individual components determined by curve
325
fitting than a mixture of many subcomponents with slightly different frequencies. This
327
quantitative analysis is shown in Table 3.
1680 cm-1
36.31
53.68
10.01
ARA3
35.77
55.02
9.21
ARA6
35.55
55.34
9.11
ARA9
35.31
55.28
9.40
CMC3
37.48
53.05
9.47
CMC6
38.67
52.10
CMC9
41.41
49.45
9.14
DEX3
35.07
54.88
10.05
DEX6
32.88
56.66
10.46
58.59
10.27
31.14
M
d
Ac ce p
DEX9
an
Control
us
1650 cm-1
te
Area Amide I 1624 cm-1
cr
326
9.23
328
Table 3. Resulting percentage of the curve fitting of amide I for soy protein (control) and soy
329
protein blends with different contents of gum arabic (ARA), carboxymethylcellulose (CMC)
330
or dextran (DEX).
331
The area of the bands at 1624 cm-1, 1650 cm-1, and 1680 cm-1 increased, were maintained
332
or decreased depending on the polysaccharide-type employed, thus differences in the spectra
333
of the blends indicated that the use of different polysaccharides resulted in the production of
334
different molecular conformations. It is worth noting that in all of the blends studied, the
335
broad band produced at 1650 cm-1 reflected the presence of residual regular structures. At this
336
position, α-helices and unordered structures are most likely to predominate in blends.
17
Page 17 of 29
Regarding the bands observed at 1624 cm-1 and 1680 cm-1, upon increasing CMC, these two
338
bands increased due to the formation of intermolecular antiparallel β-sheets (Lefevre,
339
Subirade, & Pezolet, 2005). Extended antiparallel β-sheets are commonly found in
340
aggregated proteins, especially in heat-denatured proteins (Susi, Timasheff, & Stevens,
341
1967). These results lead to the view that aggregation occurred and that the aggregates are
342
partly composed of β-sheets. This observation represents a progressive conversion of residual
343
regular structures and unordered segments into intermolecular β-sheets. The β-sheet structure
344
has a higher degree of intermolecular hydrogen bonding and explains the important role of
345
water molecules during the unfolding-folding process (Quinn et al., 2003). Conversely, when
346
DEX was added to blends, the bands at 1624 cm-1 and 1680 cm-1 decreased, and the band at
347
1650 cm-1 increased, thus DEX did not represent a progressive conversion of residual regular
348
structures and unordered segments into intermolecular β-sheets. Finally, in the case of ARA,
349
the area of the three bands remained practically constant.
d
M
an
us
cr
ip t
337
The area of aggregation bands at 1680 cm-1 and 1624 cm-1 for the control indicated that
351
46% of the amide I band is due to intermolecular β-sheets. This behavior was maintained
352
with the addition of ARA; however, the addition of DEX decreased intermolecular β-sheets
353
from 46% to 41%, and intermolecular β-sheets increased from 46% to 51% when CMC was
354
added. Therefore, it can be estimated that between 41 to 51% of the amino acids present in
355
blends are engaged in β-sheets, whereas 59 to 49% of amino acids are present in α-helix,
356
random coil segment and turn structures. However, the polysaccharides used in this study did
357
not lead to dramatic spectral differences in the blends. This might explain why there was no
358
significant difference observed in some product characteristics such as density and expansion
359
ratio (Table 1). Protein can be resistant to conformational changes during mixing and may
360
retain its native conformation, but in the extrusion process, the protein may unfold and
361
consequently, not regain its native conformation, resulting in conformational changes and
Ac ce p
te
350
18
Page 18 of 29
362
denaturation. Regaining protein structure is dependent on the polysaccharide content and type
363
used.
364
3.2.2. FTIR in amide II region
ip t
In relation to the band corresponding to amide II shown in Figure 3, the band at 1517 cm-1
365
was attributed to the vibrational modes of Tyrosine (Tyr) side chains (Hu, Kaplan, & Cebe,
367
2008). 0.30
0.30
ARA9
Control
us
368
cr
366
369 Absorbance (a. u.)
0.20
0.10 1600
373
1580
1560
1540
1520
1500
-1
Wavenumber (cm )
0.30
374
377
1480
0.10 1600
379
1540
1520
1500
1480
1500
1480
-1
Wavenumber (cm )
d te
0.25
0.20
0.15
0.10 1600
1560
DEX9
0.25
378
1580
0.30
Ac ce p
376
0.15
CMC9
Absorbance (a. u.)
375
0.20
M
0.15
372
1580
1560
1540
1520 -1
Wavenumber (cm )
1500
Absorbance (a. u.)
371
Absorbance (a. u.)
370
an
0.25
0.25
1480
0.20
0.15
0.10 1600
1580
1560
1540
1520 -1
Wavenumber (cm )
380
Figure 3. Original spectrum (black) and curve fitting spectra of amide II for soy protein
381
(control) and soy protein blends containing 9% gum arabic (ARA), carboxymethylcellulose
382
(CMC) or dextran (DEX).
383
The Tyr amino acids in SPI contribute at the interfaces of the β-sheet regions and act as
384
turns, so this band provided a specific local monitor for protein conformational changes due
385
to reduction of the conformational freedom in the protein structure (Murayama & Tomida,
386
2004; Reinstadler, Fabian, & Naumann, 1999), as shown quantitatively in Table 4. 19
Page 19 of 29
1517 cm-1
1544 cm-1
Control
26.49
73.51
ARA3
27.99
72.01
ARA6
27.85
72.15
ARA9
26.12
73.88
CMC3
24.72
75.28
CMC6
23.58
76.42
CMC9
22.15
77.85
DEX3
25.44
74.56
DEX6
23.78
76.22
DEX9
20.56
79.44
an
us
cr
ip t
Area Amide II
Table 4. Resulting percentage of the curve fitting of amide II for soy protein (control) and
388
soy protein blends with different contents of gum arabic (ARA), carboxymethylcellulose
389
(CMC) or dextran (DEX).
390
4. Conclusions
te
d
M
387
Soy protein-based blends were processed by twin-screw extrusion. During trials, a decrease
Ac ce p
391 392
in extrusion torque and specific mechanic energy was achieved through the addition of
393
polysaccharides to blends, indicating that a polysaccharide content of 6-9% in blends was
394
enough to decrease friction and facilitate macromolecules pass through the extruder die.
395
These improvements were the result of interactions between the extruder-induced
396
denaturation of soy protein and polysaccharides, which was analyzed by infrared
397
spectroscopy. The analysis of FTIR spectra in the band regions for amide I and II provided
398
novel information about conformational changes in the protein following extrusion. Although
399
the band frequency was maintained in a constant manner with respect to polysaccharide type
400
and content, further analysis of each band, which can be deconvoluted into bands
401
corresponding to the different secondary structures of the protein such as α-helix and β20
Page 20 of 29
sheets, showed that changes could be related to the decrease in SME values and to the
403
changes of some physical properties like MC and TSM. This study clearly highlights the
404
usefulness of employing FTIR as an analytical tool to assess conformational changes in the
405
structure of proteins and interactions that have taken place between biopolymers used in
406
blends following extrusion.
ip t
402
In summary, two features are the most relevant findings achieved in this work. Firstly, the
cr
407
utility of polysaccharide addition to facilitate the extrusion process, which is not the most
409
widely used method to produce this type of blends nowadays, although it is the most cost
410
effective process at industrial scale. Secondly, the value of using FTIR to analyze
411
conformational changes in globular proteins, providing novel information to understand the
412
relationship between protein structure and final properties of soy protein-based products.
413
Acknowledgments
M
an
us
408
The authors thank the University of the Basque Country (research group GIU12/06) as well
d
414
as the Basque Government (projects S-PE12UN002 and S-PE13UN057) for their financial
416
support.
417
References
418
Archana, G., Sabina, K., Babuskin, S., Radhakrishnan, K., Fayidh, M.A., Azhagu Saravana
419
Babu, P., Sivarajan, M., & Sukumar, M. (2013). Preparation and characterization of mucilage
420
polysaccharide for biomedical applications. Carbohydrate Polymers, 98, 89-94.
421
Alves, V.D., Costa, N., & Coelhoso, I.M. (2010). Barrier properties of biodegradable
422
composite films based on kappa-carrageenan/pectin blends and mica flakes. Carbohydrate
423
Polymers, 79, 269-276.
Ac ce p
te
415
21
Page 21 of 29
Andrade-Mahecha, M.M., Tapia-Blácido, D.R., & Menegalli, F.C. (2012). Development and
425
optimization of biodegradable films based on achira flour. Carbohydrate Polymers, 88, 449-
426
458.
427
Carpenter, J.F., & Crowe, J.H. (1989). An infrared spectroscopic study of the interactions of
428
carbohydrates with dried proteins. Biochemistry, 28, 3916-3922.
429
Cheftel, J.C., Cuq, J.L., & Lorient, D. (1985). In O.R. Fennema (Ed.), Food Chemistry (pp.
430
245-369). New York: Marcel Dekker.
431
Chen, F.L., Wei, Y.M., & Ojokoh, A.O. (2010). System parameters and product properties
432
response of soybean protein extruded at wide moisture range. Journal of Food Engineering,
433
96, 208-213.
434
Doublier, J.L., Garnier, C., Renard, D., & Sanchez, C. (2000). Protein-polysaccharide
435
interactions. Current Opinion in Colloid & Interface Science, 5, 202-214.
436
Félix, M., Martín-Alfonso, J.E., Romero, A., & Guerrero, A. (2014). Development of
437
albumen/soy biobased plastic materials processed by injection molding. Journal of Food
438
Engineering, 125, 7-16.
439
Fukushima, D., & Van Buren, J. (1970a). Effect of physical and chemical processing factors
440
on the redispersibility of dried soy milk proteins. Cereal Chemistry, 47, 571-578.
441
Fukushima, D., & Van Buren, J. (1970b). Mechanisms of protein insolubilization during the
442
drying of soy milk role of sisulfide and hydrophobic bonds. Cereal Chemistry, 47, 687-696.
443
Galicia-García, T., Martínez-Bustos, F., Jiménez-Arevalo, O., Martínez, A.B., Ibarra-
444
Gómez, R., Gaytán-Martínez, M., & Mendoza-Duarte, M. (2011). Thermal and
Ac ce p
te
d
M
an
us
cr
ip t
424
22
Page 22 of 29
microstructural characterization of biodegradable films prepared by extrusion-calendering
446
process. Carbohydrate Polymers, 83, 354-361.
447
Gilbert, V., Rouabhia, M., Wang, H., Arnould, A.L., Remondetto, G., & Subirade, M. (2005).
448
Characterization and evaluation of whey protein-based biofilms as substrates for in vitro cell
449
cultures. Biomaterials, 26, 7471-7480.
450
Godavarti, S., & Karwe, M.V. (1997). Determination of specific mechanical energy
451
distribution on a twin-screw extruder. Journal of Agricultural Engineering Research, 67,
452
277-287.
453
Gontard, N. & Guilbert, S. (1992). Edible wheat gluten films: influence of the main process
454
variables on film properties using response surface methodology. Journal of Food Science,
455
57, 190-195.
456
Goormaghtigh, E., Cabiaux, V., & Ruysschaert, J.M. (1990). Secondary structure and dosage
457
of soluble and membrane proteins by attenuated total reflection Fourier-transform infrared
458
spectroscopy on hydrated films. European Journal of Biochemistry, 193, 409-420.
459
Guerrero, P, Beatty, E., Kerry, J.P., & de la Caba, K. (2012). Extrusion of soy protein with
460
gelatin and sugars at low moisture content. Journal of Food Engineering, 110, 53-59.
461
Guerrero, P., Etxabide, A., Leceta, I., Peñalba, M., & de la Caba, K. (2014). Extraction of
462
agar from Gelidium sesquipedale (Rodhopyta) and surface characterization of agar based
463
films. Carbohydrate Polymers, 99, 491-498.
464
Guerrero, P., Garrido, T., Leceta, I., & de la Caba, K. (2013). Films based on proteins and
465
polysaccharides: Preparation and physical–chemical characterization. European Polymer
466
Journal, 49, 3713-3721.
Ac ce p
te
d
M
an
us
cr
ip t
445
23
Page 23 of 29
Guerrero, P., Nur Hanani, Z.A., Kerry, J.P., & de la Caba, K. (2011). Characterization of soy
468
protein-based films prepared with acids and oils by compression. Journal of Food
469
Engineering, 107, 41-49.
470
Harper, J.M. (1989). Food extruders and their applications. In C. Mercier, P. Linko, & J.P.
471
Harper (Eds.), Extrusion Cooking (pp. 1-15). Minnesota: American Assoc. Cereal Chemistry.
472
Hu, L., Hsieh, F., & Huff, H.E. (1993). Corn meal extrusion with emulsifier and soy bean
473
fiber. LWT-Food Science and Technology, 26, 544-551.
474
Hu, X., Kaplan, D., & Cebe, P. (2008). Dynamic protein-water relationships during β-sheet
475
formation. Macromolecules, 41, 3939-3948.
476
Hu, X., Lu, Q., Sun, L., Cebe, P., Wang, X., Zhang, X., & Kaplan, D. (2010). Biomaterials
477
from ultrasonicatio-induced silk fibroin-hyaluronic acid hydrogels. Biomacromolecules, 11,
478
3178-3188.
479
Klein, M., Aserin, A., Ben Ishai, P., & Garti, N. (2010). Interactions between whey protein
480
isolate and gum Arabic. Colloids and Surfaces B: Biointerfaces, 79, 377-383.
481
Kunte, L., Gennadios, A., Cuppett, S., Hanna, M., & Weller, C. (1997). Cast films from soy
482
protein isolates and fractions. Cereal Chemistry, 72, 115-118.
483
Leceta, I., Etxabide, A., Cabezudo, S., de la Caba, K., & Guerrero, P. (2014). Bio-based films
484
prepared with by-products and wastes: environmental assessment. Journal of Cleaner
485
Production, 64, 218-227.
486
Lefevre, T., & Subirade, M. (2000). Molecular differences in the formation and structure of
487
fine-stranded and particulate β-lactoglobulin gels. Biopolymers, 54, 578-586.
Ac ce p
te
d
M
an
us
cr
ip t
467
24
Page 24 of 29
Lefevre, T., Subirade, M., & Pezolet, M. (2005). Molecular description of the formation and
489
structure of plasticized globular protein films. Biomacromolecules, 6, 3209-3219.
490
Mangavel, C., Barbot, J., Popineau, Y., & Gueguen, J. (2001). Evolution of wheat gliadins
491
conformation during film formation: a Fourier transform infrared study. Journal of
492
Agricultural and Food Chemistry, 49, 867-872.
493
Murayama, K., & Tomida, M. (2004). Heat-induced secondary structure and conformation
494
change of bovine serum albumin investigated by Fourier transform infrared spectroscopy.
495
Biochemistry, 43, 11526-11532
496
Quinn, G., Monahan, F.J., O’Sullivan, M., & Langares, A. (2003). Role of covalent and
497
noncovalent interactions in the formation of films from unheated whey protein solutions
498
following pH adjustment. Journal of Food Science, 68, 2284-2288.
499
Reinstadler, D., Fabian, H., & Naumann, D. (1999). New structural insights into the refolding
500
of ribonuclease T1 as seen by time-resolved Fourier-transform infrared spectroscopy.
501
Proteins: Structure, Function, and Bioinformatics, 34, 303-316.
502
Rochas, C., Lahaye, M., & Yaphe, W. (1986). Sulfate content of carrageenan and agar
503
determined by infrared spectroscopy. Botanica Marina, 29, 335-340.
504
Rodríguez, J.M., & Pilosof, A.M.R. (2011). Protein-polysaccharide interactions at fluid
505
interfaces. Food Hydrocolloids, 25, 1925-1937.
506
Salarbashi, D., Tajik, S., Ghasemlou, M., Shojaee-Aliabadi, S., Shahidi Noghabi, M., &
507
Khaksar, R. (2013). Characterization of soluble soybean polysaccharide film incorporated
508
essential oil intended for food packaging. Carbohydrate Polymers, 98, 1127-1136.
Ac ce p
te
d
M
an
us
cr
ip t
488
25
Page 25 of 29
Song, F., Tang, D.L., Wang, X.L., & Wang, Y.Z. (2011). Biodegradable soy protein isolate-
510
based materials: a review. Biomacromolecules, 12, 3369-3380.
511
Subirade, M., Kelly, I., Gueguen, J., & Pezolet, M. (1998). Molecular basis of film formation
512
from a soybean protein: comparison between the conformation of glycinin in aqueous
513
solution and in films. International Journal of Biological Macromolecules, 23, 241-249.
514
Susi, H., Timasheff, S.N., & Stevens, L. (1967). Infrared spectra and protein conformations
515
in aqueous solutions: I. The amide I band in H2O and D2O solutions. Journal of Biological
516
Chemistry, 242, 5460-5466.
517
Wang, L., Xiao, M., Dai, S., Song, J., Ni, X., Fang, Y., Corke, H., & Jiang, F. (2014).
518
Interactions between carboxymethyl konjac glucomannan and soy protein isolate in blended
519
films. Carbohydrate Polymers, 101, 136-145.
520
Wang, C.H., & Damodaran, S. J. (1991). Thermal gelation of globular proteins. Journal of
521
Agricultural and Food Chemistry, 39, 433-438.
te
d
M
an
us
cr
ip t
509
Ac ce p
522
26
Page 26 of 29
522
Figure captions
524
Figure 1. Effect of polysaccharide content on FTIR spectra in the amide I and II region for
525
soy protein blends with gum arabic (ARA), carboxymethylcellulose (CMC), and dextran
526
(DEX).
527
Figure 2. Original spectrum (black) and curve fitting spectra of amide I for soy protein
528
(control) and soy protein blends containing 9% gum arabic (ARA), carboxymethylcellulose
529
(CMC) or dextran (DEX).
530
Figure 3. Original spectrum (black) and curve fitting spectra of amide II for soy protein
531
(control) and soy protein blends containing 9% gum arabic (ARA), carboxymethylcellulose
532
(CMC) or dextran (DEX).
533
Table captions
534
Table 1. Effect of gum arabic (ARA), carboxymethylcellulose (CMC), and dextran (DEX)
535
content on specific mechanical energy (SME), piece density (PD), expansion ratio (ER),
536
moisture content (MC) and total soluble matter (TSM).
537
a-e
538
different through the Tukey´s multiple range test.
539
Table 2. FTIR absorption area values for amide I (1628 cm-1), amide II (1541 cm-1), and
540
reference (2930 cm-1) bands as they pertain to soy protein blends with different gum arabic
541
(ARA), carboxymethylcellulose (CMC), and dextran (DEX) contents.
542
Table 3. Resulting percentage of the curve fitting of amide I for soy protein (control) and soy
543
protein blends with different contents of gum arabic (ARA), carboxymethylcellulose (CMC)
544
or dextran (DEX).
Ac ce p
te
d
M
an
us
cr
ip t
523
Two means followed by the same letter in the same column are not significantly (P > 0.001)
27
Page 27 of 29
Table 4. Resulting percentage of the curve fitting of amide II for soy protein (control) and
546
soy protein blends with different contents of gum arabic (ARA), carboxymethylcellulose
547
(CMC) or dextran (DEX).
Ac ce p
te
d
M
an
us
cr
ip t
545
28
Page 28 of 29
Page 29 of 29
ed
pt
ce
Ac M
an
us
cr
S0144-8617(14)00471-8 http://dx.doi.org/doi:10.1016/j.carbpol.2014.05.005 CARP 8868
To appear in: Received date: Revised date: Accepted date:
10-3-2014 30-4-2014 5-5-2014
Please cite this article as: Guerrero, P., Kerry, J. P., & de la Caba, K.,FTIR characterization of protein-polysaccharide interactions in extruded blends, Carbohydrate Polymers (2014), http://dx.doi.org/10.1016/j.carbpol.2014.05.005 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Highlights •
Extruder parameters were improved by polysaccharide addition, facilitating extrusion
3
•
FTIR analysis showed changes in the secondary structure of protein
4
•
Conformational changes in blends explained physical properties of extrudates
Ac ce p
te
d
M
an
us
cr
ip t
2
1
Page 1 of 29
FTIR characterization of protein-polysaccharide interactions in extruded
a
BIOMAT Research Group,
cr
Pedro Guerreroa, Joe P. Kerryb, Koro de la Cabaa*
ip t
blends
us
University of the Basque Country (UPV/EHU),
an
Department of Chemical and Environmental Engineering,
b
M
Polytechnic School, Donostia-San Sebastián, Spain Food Packaging Group,
te
d
School of Food & Nutritional Sciences, University College Cork (UCC), Cork, Ireland
Ac ce p
*Corresponding author
Koro de la Caba
Escuela Universitaria Politécnica
Departamento de Ingeniería Química y del Medio Ambiente Plaza Europa 1
20018 Donostia-San Sebastián Spain E-mail: [email protected]
Telephone number: 00 34 943 017188 5
Abstract 2
Page 2 of 29
Soy protein-based blends were processed by double screw extrusion and the effects of
7
different types and contents of polysaccharides were analyzed. Although extrusion has not
8
been widely used for this type of blends, in this study it was observed that the increase in
9
polysaccharide content in blends caused a decrease in specific mechanical energy (SME),
10
facilitating extrusion process and showing the potential of this process, which is more cost
11
effective at industrial scale. In order to explain this behavior, infrared spectroscopy analysis
12
was carried out, mainly in the amide I and amide II region. Moreover, curve fitting analysis
13
showed the conformational changes produced in the blends due to the addition of
14
polysaccharides, which affected protein denaturation. These changes also affected properties
15
such as moisture content (MC) and total solubility matter (TSM). However, conformational
16
changes did not show significant effects with respect to piece density (PD) or in the
17
expansion ratio (ER) of the pellets. The quantitative analysis of the changes in the amide I
18
and II region provided novel information about the modifications produced in protein-based
19
blends modified with polysaccharides. In this context, infrared spectroscopy provided a
20
convenient and powerful means to monitor interactions between all ingredients used in the
21
blend formulation, which is of great importance in order to explain changes in the functional
22
properties of biodegradable materials used for industrial applications in food and
23
pharmaceutical industries.
24
Keywords: proteins, secondary structure, polysaccharides, interactions, FTIR.
Ac ce p
te
d
M
an
us
cr
ip t
6
25
3
Page 3 of 29
25 26
1. Introduction Proteins and polysaccharides are biomacromolecules employed as functional ingredients.
27
Mixtures of both types of biomacromolecules are often used in many technological
29
applications, including food, pharmaceutical, and cosmetics industries (Andrade-Mahecha,
30
Tapia-Blacido, & Menegalli, 2012; Archana et al., 2014; Guerrero, Etxabide, Leceta,
31
Peñalba, & de la Caba, 2014; Salarbashi et al., 2014). Protein-polysaccharide conjugates are
32
useful because such ingredient conjugates have improved functional properties, such as heat
33
and pH stability, solubility, emulsification, and gelation properties, offering lower
34
astringency and allergenicity compared to unmodified proteins (Doublier, Garnier, Renard, &
35
Sanchez, 2000; Klein, Aserin, Ben Ishai, & Garti, 2010; Rodríguez & Pilosof, 2011). These
36
properties, together with the increasing demands to control biochemical interactions at the
37
tissue/material interface, have prompted scientists to investigate these systems. Nevertheless,
38
the knowledge underpinning protein-polysaccharide interactions is still rather limited (Alves,
39
Costa, & Coelhoso, 2010; Wang et al., 2014). In the specific context of protein-based blends,
40
several factors such as protein concentration, pH, ionic strength, heating temperature, and
41
plasticizers affect the formation of the blend. It has been suggested that the blend network is
42
formed via hydrogen bonds and hydrophobic and electrostatic interactions. However, the
43
structural basis of the blend-forming process is, at present, not fully understood. No attempt
44
has been made to determine the relationship between the extent of formation of the blend
45
network and the structure of the protein in the blend, and nothing is known about the
46
molecular basis that leads to the formation of such blends.
Ac ce p
te
d
M
an
us
cr
ip t
28
The formation of blends from soy proteins and other globular proteins has been described
47 48
as a heat denaturation process. Heating alters the three-dimensional structure of proteins,
49
exposing functional peptide groups such as CO and NH and side chain polar and hydrophobic 4
Page 4 of 29
groups that are engaged in intramolecular hydrogen bonding and electrostatic interactions in
51
the native state; therefore, these functional groups become available for further
52
intermolecular interactions (Wang & Damodaran, 1991). The unfolded protein chains
53
approach each other and become linked through intermolecular interactions, leading to the
54
formation of a network that acts as the matrix for entrapping blend components such as
55
plasticizing agents. Since the formation of a blend occurs upon application of denaturating
56
conditions, it is quite possible that the denatured protein may undergo partial refolding, thus
57
regaining some secondary structure during the blending process (Gilbert et al., 2005). It is
58
conceivable that the extent of such refolding affects the number of functional groups
59
available for intermolecular interactions and thus, the formation and stability of the blend
60
network. Blends based on globular proteins are attractive materials because of their potential
61
applications as edible or biodegradable coatings and wrappings. This type of matrix is also
62
relevant for pharmaceutical and medical applications since it may be employed for the
63
encapsulation and release in vivo of bioactive substances at specific sites and for the growth
64
of cell cultures. A clear understanding of the structure and formation of blends is required to
65
control their properties. Besides these applications, globular protein-based blends also arouse
66
interest in the more fundamental aspects of chemical interactions involving their use since
67
they may be helpful in providing insights into the interactions that govern protein self-
68
assembly and protein conformation. In this context, many efforts have been devoted to the
69
optimization of the techno-functional properties of soy proteins, which have received
70
increasing attention since soy protein is a co-product of the soy oil industry and which can be
71
valorised (Félix, Martín-Alfonso, Romero, & Guerrero, 2014; Leceta, Etxabide, Cabezudo,
72
de la Caba, & Guerrero, 2014).
Ac ce p
te
d
M
an
us
cr
ip t
50
Regarding the processing of protein-based blends, the employment of extrusion is one of
73 74
the most versatile and well-established processes used to produce macromolecular and
5
Page 5 of 29
polymeric materials (Galicia-García et al., 2011). Moreover, extrusion is relatively
76
inexpensive (as the technology can be scaled to production volume and simplified for specific
77
commercial usage) and amenable to industrial scale-up. Extrusion conditions, such as high
78
barrel temperatures, high pressures and the use of low moisture during processing, bring
79
about major conformational changes in the ingredient components in extrusion blends during
80
processing. In addition, it is well-known that, among all of the different process variables that
81
need to be considered during extrusion, the rate and amount that moisture is introduced into
82
the extruder and the processing temperature employed at all critical points along the process
83
have important effects on the quality of the final extruded products. It is generally accepted
84
that proteins are denatured during the extrusion process, so these “reactive” unfolded protein
85
chains can interact with each other, and with other added components, allowing the formation
86
of a polymeric network. The role of covalent bonding on final product characteristics has
87
been well documented since disulfide bonds were found to develop during this process
88
(Fukushima & Van Buren, 1970a); however, physical interactions including hydrophobic
89
effects and hydrogen bonding are clearly also involved (Fukushima & Van Buren, 1970a;
90
Quinn, Monahan, O’Sullivan, & Langares, 2003), although the contribution of each type of
91
interaction remains unclear. Most of the studies on blends have focused on a selected number
92
of macroscopic characteristics attributable to the resulting films produced, such as stress and
93
strain at break point, permeability to gases and vapours, and moisture absorption, among
94
others. While the macroscopic properties of films made with globular proteins have been
95
extensively investigated, only a limited amount of information exists with respect to their
96
structure. Films made of soybean glycinin and wheat gliadin have been examined at a
97
molecular level using Fourier transform infrared (FTIR) spectroscopy (Mangavel, Barbot,
98
Popineau, & Gueguen, 2001; Subirade, Kelly, Gueguen, & Pezolet, 1998). These studies
99
revealed the presence of intermolecular β-sheets following heating and dehydration steps
Ac ce p
te
d
M
an
us
cr
ip t
75
6
Page 6 of 29
which underline the importance of hydrogen bonding in the formation of the protein network.
101
β-sheets have been proposed to act as junction zones in the cohesion of film, and in the
102
relationship between protein conformation and mechanical properties has been proposed
103
(Mangavel et al., 2001). Such reports underline the role that the β-sheet motif may play in the
104
structure and properties of globular protein films but more data are needed to evaluate the
105
importance of this type of structure.
cr
ip t
100
Since few data are available about the molecular events occurring during blend formation
107
employing globular proteins, the aim of this work was to describe the protein conformational
108
changes occurring during the extrusion process. The influence of heat treatment and the role
109
played by different types and contents of polysaccharides on final structure were also
110
considered. In order to elucidate the molecular basis pertaining to blend formation,
111
conformational changes in the biomacromolecules employed in this study were investigated
112
under extrusion conditions. In this investigation, infrared spectroscopy was employed to
113
obtain detailed information on the molecular and conformational modifications occurring
114
during the blending formation process since it has been well established that infrared
115
spectroscopy is a powerful method for the investigation of secondary protein structures in
116
complex systems, such as biological or food systems (Guerrero, Garrido, Leceta, & de la
117
Caba, 2013; Goormaghtigh, Cabiaux, & Ruysschaert, 1990; Mangavel et al., 2001).
118
2. Materials and methods
119
2.1. Materials
Ac ce p
te
d
M
an
us
106
120
Soy protein isolate (SPI), PROFAM 974, with 90% protein on a dry basis was supplied by
121
Lactotecnia S.L. (Barcelona, Spain). The SPI employed in this study was comprised of 5% of
122
moisture, 4% of fat and 5% of ash. It had an acid character and an isoelectric point of 4.6.
123
The polysaccharides used in this study were obtained from Aldrich (Madrid, Spain), and
7
Page 7 of 29
included gum arabic (ARA) from acacia tree (G9752), dextran (DEX) from Leuconostoc
125
mesenteroides (D5251), and carboxymethylcellulose (CMC) sodium salt (C4888). Glycerol
126
(GLY) was used as the plasticizer in this study and this was obtained from Panreac
127
(Barcelona, Spain).
128
2.2. Extruder and experimental configurations
ip t
124
A twin-screw extruder (MPF 19/25 APV Baker) was used in this study. The APV Baker
130
MPF19 extruder was designed to generate the high pressures needed to extrude a wide range
131
of products, in particular, edible/biodegradable blends. The MPF 19/25 possessed 19 mm
132
diameter barrel, a length to barrel diameter ratio of 25:1, and a barrel comprising four
133
controlled temperature zones, resulting in a die temperature of 70 to 100ºC.
134
2.3. Samples preparation
M
an
us
cr
129
SPI, glycerol and polysaccharide contents required to obtain the desired blend composition
135
were mixed. All components were mixed in a variable speed Stephan mixer, model UMC 5
137
(Stephan, UK), for 5 min at 1500 rpm. SPI was replaced (3, 6 and 9 wt %) by polysaccharide
138
to obtain SPI-polysaccharide blends, and samples were designated as ARA3, ARA6 and
139
ARA9 in the case of gum arabic, CMC3, CMC6 and CMC9 in the case of
140
carboxymethylcellulose, and DEX3, DEX6 and DEX9 for dextran. Glycerol was selected at
141
30 % (w/w), based on results obtained from previous work (Guerrero, Nur Hanani, Kerry, &
142
de la Caba, 2011). The sample without polysaccharides was designated as the control.
143
Mixtures were added into the feed hopper and mixed with water in the barrel of the twin-
144
screw extruder. All trials were carried out using a water speed of 2.94 g/min (0.15 kg/h).
145
Water was pumped directly into the extruder barrel zone 1 using a peristaltic pump (504U
146
MK, Watson Marlow Ltd.).
147
2.4. Specific mechanical energy (SME)
Ac ce p
te
d
136
8
Page 8 of 29
The specific mechanical energy was calculated using the expression (Hu, Hsieh, & Huff,
148 149
1993): Screw speed × Power (kW) × Torque (%) Max. screw speed × Throughput (kg/h) × 100
SME (kJ/kg)=
151
The extruder was operated at a constant speed of 250 rpm. The feed rate of extruder was
ip t
150
adjusted to 1 kg/h and had a 2 kW/h motor power in practice. Dies were scaled up by
153
maintaining a constant throughput per unit of orifice area. The extruder was equipped with a
154
single die of 3 mm diameter giving a throughput per unit area of 0.141 kg/h mm2.
155
2.5. Piece density (PD)
an
us
cr
152
Piece density is the density of each extrudate and was obtained by dividing the mass of
156
extrudate piece (Wpiece) by its volume (Vpiece), which was computed from its dimensions
158
(diameter, d; and length, l). The piece dimensions of ten, 2 cm-long, pieces were measured by
159
using a hand-held digimatic micrometer (QuantuMike Mitutoyo), and piece density was
160
calculated as:
161
PD =
162
2.6. Expansion ratio (ER)
π×d 2 ×l
te
4×Wpiece
Ac ce p
Vpiece
=
The expansion ratio is the ratio of the extrudate cross-sectional area to the die orifice cross-
163 164
Wpiece
d
M
157
sectional area, and was calculated as: ER =
165
d2 2 d die
166
where ddie is the die diameter and d is the extrudate diameter (average of 10 samples), which
167
was measured by using a hand-held digimatic micrometer (QuantuMike Mitutoyo).
168
2.7. Moisture content (MC) and total soluble matter (TSM)
9
Page 9 of 29
Three specimens of each sample were analyzed to determine moisture content and total
170
soluble matter values according to a method previously used in several studies of proteins
171
(Gontard & Guilbert, 1992; Kunte, Gennadios, Cuppett, Hanna, & Weller, 1997). Samples
172
were weighed (mw) and subsequently dried in an air-circulating oven at 105 ºC for 24 h. After
173
this time, samples were reweighed (m0) to determine MC values. Samples were dried as an
174
extrudate and MC values were calculated as:
cr
ip t
169
m w -m 0 ×100 mw
MC ( % ) =
176
Afterwards, samples were immersed in 30 mL of distilled water in the presence of sodium
us
175
azide (0.02%) in order to prevent microbial growth. The beakers were stored in an
178
environmental chamber at 25 ºC for 24 h with occasional gentle stirring. After this time,
179
specimens were dried in an air-circulating oven at 105 ºC for 24 h and weighed (mf). TSM
180
was expressed as the percentage of sample dry matter dissolved after 24 h immersion in
181
distilled water and determined by the following equation:
Ac ce p
183
m 0 -m f ×100 m0
te
TSM ( % ) =
182
d
M
an
177
2.8. Fourier transform infrared spectroscopy (FTIR) The IR spectra for blends were obtained using a Varian 600-IR Series FTIR equipped with
184 185
horizontal attenuated total reflectance (ATR) crystal (ZnSe). The spectra were collected in
186
absorbance mode on sample powders obtained by grinding pellets in an agate mortar and
187
placed directly onto the ATR crystal. Each spectrum is the result of the average of 32 scans at
188
4 cm−1 resolution. Measurements were recorded between 4000 and 700 cm−1. All spectra
189
were smoothed using the Savitzky-Golay function. Second-derivative spectra of the amide
190
region were used at peak position guides for the curve fitting procedure, using OriginPro 8.6
191
software.
10
Page 10 of 29
192
2.9. Statistical analysis Data were subjected to one-way analysis of variant (ANOVA) by means of a SPSS
193
computer program (SPSS Statistic 20.0). Post hoc multiple comparisons were determined by
195
the Tukey’s test with the level of significance set at P < 0.001.
196
3. Results and discussion
cr
ip t
194
Extrusion is generally defined as a process used to mix, homogenize, and shape material by
197
forcing it through a specifically designed opening. Extrusion process variables are critical
199
factors to be considered as they impact significantly upon ingredient blends, especially those
200
employing proteins, which ultimately impact on final film properties. Thermal extrusion
201
exposes proteinaceous ingredients to high temperature, high pressure, and mechanical shear,
202
which converts an ingredient like soy protein into a continuous plastic “melt”, resulting in
203
protein denaturation (Harper, 1989). Within the process, the main fractions of SPI (7S and
204
11S globulins) undergo a complex pattern of association-dissociation reactions (Cheftel, Cuq,
205
& Lorient, 1985). The effect of extrusion is to disassemble proteins and then reassemble them
206
using disulfide bonds, hydrogen bonds, and non-covalent interactions forming fibrous
207
extrudates. Therefore, extruded blends of various ingredients provide an excellent model
208
system to study the effects of extrusion parameters on protein structure and to monitor
209
protein-carbohydrate interactions, where carbohydrate ingredients have been added.
210
3.1. Effect of polysaccharides on extruder parameters and physical properties
Ac ce p
te
d
M
an
us
198
Firstly, the effect of polysaccharide content on SME values was analyzed and results are
211 212
shown in Table 1. The SME value indicates the extent of molecular breakdown that
213
mechanical force undergoes during the extrusion process, so this value is an important
214
parameter since it can influence final product characteristics (Godavarti & Karwe, 1997).
215
Since SME is the amount of mechanical energy dissipated as heat inside the material, 11
Page 11 of 29
expressed per unit mass of the material (Harper, 1989), increasing polysaccharide content in
217
this study resulted in higher heat transfer from the extruder barrel to the feed material and
218
consequently, a lower viscosity, lower shear, and lower friction during extrusion. The higher
219
the polysaccharide content (6-9 %), the lower the torque and SME (P < 0.001) observed.
220
Because of the reduction of the force required to push the mass through the die, a
221
consequential reduction in friction occurred between the raw processing materials, screw
222
shaft, and extruder barrel (Chen, Wei, & Ojokoh, 2010).
PD (g/cm3)
ER
MC (%)
TSM (%)
Control
972±20.13a
1.63±0.14a
1.12±0.14a
17.36±0.05a
31.49±0.06a
ARA3
972±29.16a
1.62±0.12a
1.13±0.07a
18.88±0.07a
31.42±0.03a
ARA6
648±15.21b
1.60±0.26a
1.14±0.16a
14.65±0.26bc
36.51±0.56b
ARA9
612±10.91c
1.59±0.34a
1.15±0.16a
13.79±0.21b
38.65±0.23b
CMC3
972±31.51a
1.63±0.19a
1.12±0.12a
18.36±0.51a
30.55±0.21a
CMC6
648±20.11b
1.65±0.12a
1.05±0.04b
15.49±0.29b
30.27±0.09a
CMC9
648±22.31b
1.66±0.06a
1.03±0.09b
15.13±0.49b
30.38±0.61a
DEX3
1044±45.51e 1.64±0.17a
1.10±0.16ab
18.70±0.51a
31.01±0.07a
DEX6
900±32.53d
1.63±0.16a
1.12±0.23a
14.53±0.32bc
32.35±0.47ab
DEX9
900±35.91d
1.61±0.24a
1.13±0.32a
12.75±0.41c
33.87±0.53b
Ac ce p
te
d
an
SME (kJ/kg)
M
us
cr
ip t
216
223
Table 1. Effect of gum arabic (ARA), carboxymethylcellulose (CMC), and dextran (DEX)
224
content on specific mechanical energy (SME), piece density (PD), expansion ratio (ER),
225
moisture content (MC) and total soluble matter (TSM).
226
a-e
227
different through the Tukey´s multiple range test.
Two means followed by the same letter in the same column are not significantly (P > 0.001)
228
During the extrusion process, protein denaturation takes place, as has been described
229
previously (Guerrero, Beatty, Kerry, & de la Caba, 2012), which leads to the decrease of 12
Page 12 of 29
electrostatic repulsions and allows intermolecular interactions and network formation to
231
occur (Song, Tang, Wang, & Wang, 2011). The nature of these interactions and the structure
232
of the network formed may affect the macroscopic properties of the blends. The reduction in
233
conformational freedom would explain the reason for the decrease (P < 0.001) in moisture
234
absorption shown in lower MC values, when polysaccharides were added (Table 1).
235
Conversely, TSM values increased (P < 0.001) when the polysaccharide content in blends
236
increased, probably due to the reduction of protein-protein interactions in the presence of
237
polysaccharides and due to the formation of a less tight polymeric network. These facts were
238
confirmed by FTIR analysis, as shown below.
239
3.2. Effect of polysaccharides on protein structure
an
us
cr
ip t
230
Firstly, the band corresponding to amide I depends on the secondary structure of the protein
M
240
backbone and it is the most commonly used band for the quantitative analysis of secondary
242
structures. Secondly, the band associated with amide II is caused mainly by C-N stretching
243
and N-H in plane bending. Therefore, alterations of these bands might clarify to what extent
244
polysaccharide addition to ingredient blends might assist in the stabilization of added protein,
245
since the progressive exposure of the proteins to polysaccharides may govern the structural
246
changes that ultimately take place during the extrusion process. In order to analyze this
247
influence, the FTIR spectra for the blends in this target IR region are shown in Figure 1.
Ac ce p
te
d
241
248 249 250 251 252 253 254
13
Page 13 of 29
0.4
ARA ARA9 ARA6 ARA3 Control
255 0.3 Absorbance
256 257 258
0.2
0.1
0.0
260
1750
1700
1650
1600
1550
1500
ip t
259 1450
-1
Wavenumber (cm )
0.4
Absorbance
263 264 265
us
0.3
0.2
0.1
0.0 1750
267
1700
1650
1600
1550 -1
Wavenumber (cm )
0.4
0.2
te
Absobance
271
d
0.3
270
0.1
Ac ce p
272
0.0 1750
273
1450
DEX DEX9 DEX6 DEX3 Control
268 269
1500
M
266
cr
262
CMC CMC9 CMC6 CMC3 Control
an
261
1700
1650
1600
1550
1500
1450
-1
Wavenumber (cm )
274
Figure 1. Effect of polysaccharide content on FTIR spectra in the amide I and II region for
275
soy protein blends with gum arabic (ARA), carboxymethylcellulose (CMC), and dextran
276
(DEX).
When polysaccharides were added, the frequencies of the bands assigned to the amide I and
277 278
amide II regions remained constant at 1628 cm-1 and 1541 cm-1, respectively. However, the
279
band assigned to carboxylate in the polysaccharides at 1587 cm-1 for CMC, 1600 cm-1 for
280
ARA, and 1643 cm-1 for DEX was not detectable in the blends with SPI, indicating
281
interactions between the protein and the polysaccharides and highlights the lower capacity of 14
Page 14 of 29
the carbohydrate molecules to form intermolecular hydrogen bonds between themselves in
283
presence of the protein (Carpenter & Crowe, 1989). Additionally, the dominant spectral
284
change observed was the decrease in the intensity observed for the amide I and amide II
285
bands, thereby indicating conformational changes occurring through interactions between
286
protein and the polysaccharides. These changes were quantitatively measured and results are
287
shown in Table 2.
cr
ip t
282
1628 cm-1
1541 cm-1
2930 cm-1
1628/2930 cm-1
1541/2930 cm-1
Control
14.624
7.649
0.168
87.048
45.530
ARA3
13.151
6.736
0.151
87.093
44.609
ARA6
12.456
5.525
0.149
83.597
37.081
ARA9
12.023
5.087
0.156
77.071
32.609
CMC3
13.666
6.896
0.157
87.045
43.924
CMC6
11.735
5.636
0.139
84.424
40.547
CMC9
10.681
4.888
0.136
78.537
35.941
DEX3
13.404
6.853
0.154
87.039
44.500
DEX6
12.579
6.143
0.152
82.757
40.414
5.830
0.161
74.155
36.211
an
M
d
te
11.939
Ac ce p
DEX9
us
Amide area
288
Table 2. FTIR absorption area values for amide I (1628 cm-1), amide II (1541 cm-1), and
289
reference (2930 cm-1) bands as they pertain to soy protein blends with different gum arabic
290
(ARA), carboxymethylcellulose (CMC), and dextran (DEX) contents. The relationship of amide I and amide II to total carbohydrate was estimated from the ratio
291 292
of the absorbances at 1628 cm-1 and 1541 cm-1 to 2930 cm-1. The well-defined band at 2930
293
cm-1 due to C-H content represents a good reference for total carbohydrate content, because
294
the number of C-H groups remains constant, irrespective of changes in the ratio of
295
polysaccharides and protein contents (Rochas, Lahaye, & Yaphe, 1986). As can be seen in
296
Table 2, the area of the normalized bands, corresponding to amide I and amide II, decreased
15
Page 15 of 29
when the content of the polysaccharide increased. This change was lower for those blends
298
containing ARA and higher for the blends containing DEX. The results indicate that
299
carbohydrates caused changes in the infrared spectrum of soy protein, decreasing amide I and
300
II band area and eliminating the carboxylate band, which would be indicative of the degree of
301
hydrogen bonding occurring between proteins and carbohydrates which are necessary to
302
provide stability to the extruded blends.
303
3.2.1. FTIR in amide I region
us
cr
ip t
297
In addition to the analysis carried out above and since the band corresponding to amide I
304
depends on the secondary structure of the protein backbone, it can also be used for the
306
quantitative analysis of different secondary structures. Specifically, protein unfolding is
307
characterized by changes in the intensity of the dominant band in the native protein at 1650
308
cm-1 (assigned to α-helix/unordered structures) and in the regions around 1615 to 1630 cm-1
309
and 1680 to 1700 cm-1 (assigned to β-sheets) (Hu, Lu, Sun, Cebe, Wang, Zhang, & Kaplan,
310
2010; Lefevre & Subirade, 2000), as shown in Figure 2.
te
d
M
an
305
0.4 Control
Ac ce p
311
0.4 ARA9
0.3
Absorbance (a. u.)
313 314 315
0.2
0.1
0.0
316
Absorbance (a. u.)
0.3
312
1720
1700
1680
1660
1640
1620
1600
0.2
0.1
0.0
1580
1720
1700
-1
1600
1580
1620
1600
1580
0.3
Absorbance (a. u.)
Absorbance (a. u.)
320
1620
DEX9
0.3
319
1640
0.4
CMC9
318
1660
Wavenumber (cm )
0.4
317
1680
-1
Wavenumber (cm )
0.2
0.1
0.2
0.1
321 0.0
0.0 1720
1700
1680
1660
1640
1620 -1
Wavenumber (cm )
1600
1580
1720
1700
1680
1660
1640 -1
Wavenumber (cm )
16
Page 16 of 29
322
Figure 2. Original spectrum (black) and curve fitting spectra of amide I for soy protein
323
(control) and soy protein blends containing 9% gum arabic (ARA), carboxymethylcellulose
324
(CMC) or dextran (DEX).
ip t
Therefore, it appears wiser to consider the individual components determined by curve
325
fitting than a mixture of many subcomponents with slightly different frequencies. This
327
quantitative analysis is shown in Table 3.
1680 cm-1
36.31
53.68
10.01
ARA3
35.77
55.02
9.21
ARA6
35.55
55.34
9.11
ARA9
35.31
55.28
9.40
CMC3
37.48
53.05
9.47
CMC6
38.67
52.10
CMC9
41.41
49.45
9.14
DEX3
35.07
54.88
10.05
DEX6
32.88
56.66
10.46
58.59
10.27
31.14
M
d
Ac ce p
DEX9
an
Control
us
1650 cm-1
te
Area Amide I 1624 cm-1
cr
326
9.23
328
Table 3. Resulting percentage of the curve fitting of amide I for soy protein (control) and soy
329
protein blends with different contents of gum arabic (ARA), carboxymethylcellulose (CMC)
330
or dextran (DEX).
331
The area of the bands at 1624 cm-1, 1650 cm-1, and 1680 cm-1 increased, were maintained
332
or decreased depending on the polysaccharide-type employed, thus differences in the spectra
333
of the blends indicated that the use of different polysaccharides resulted in the production of
334
different molecular conformations. It is worth noting that in all of the blends studied, the
335
broad band produced at 1650 cm-1 reflected the presence of residual regular structures. At this
336
position, α-helices and unordered structures are most likely to predominate in blends.
17
Page 17 of 29
Regarding the bands observed at 1624 cm-1 and 1680 cm-1, upon increasing CMC, these two
338
bands increased due to the formation of intermolecular antiparallel β-sheets (Lefevre,
339
Subirade, & Pezolet, 2005). Extended antiparallel β-sheets are commonly found in
340
aggregated proteins, especially in heat-denatured proteins (Susi, Timasheff, & Stevens,
341
1967). These results lead to the view that aggregation occurred and that the aggregates are
342
partly composed of β-sheets. This observation represents a progressive conversion of residual
343
regular structures and unordered segments into intermolecular β-sheets. The β-sheet structure
344
has a higher degree of intermolecular hydrogen bonding and explains the important role of
345
water molecules during the unfolding-folding process (Quinn et al., 2003). Conversely, when
346
DEX was added to blends, the bands at 1624 cm-1 and 1680 cm-1 decreased, and the band at
347
1650 cm-1 increased, thus DEX did not represent a progressive conversion of residual regular
348
structures and unordered segments into intermolecular β-sheets. Finally, in the case of ARA,
349
the area of the three bands remained practically constant.
d
M
an
us
cr
ip t
337
The area of aggregation bands at 1680 cm-1 and 1624 cm-1 for the control indicated that
351
46% of the amide I band is due to intermolecular β-sheets. This behavior was maintained
352
with the addition of ARA; however, the addition of DEX decreased intermolecular β-sheets
353
from 46% to 41%, and intermolecular β-sheets increased from 46% to 51% when CMC was
354
added. Therefore, it can be estimated that between 41 to 51% of the amino acids present in
355
blends are engaged in β-sheets, whereas 59 to 49% of amino acids are present in α-helix,
356
random coil segment and turn structures. However, the polysaccharides used in this study did
357
not lead to dramatic spectral differences in the blends. This might explain why there was no
358
significant difference observed in some product characteristics such as density and expansion
359
ratio (Table 1). Protein can be resistant to conformational changes during mixing and may
360
retain its native conformation, but in the extrusion process, the protein may unfold and
361
consequently, not regain its native conformation, resulting in conformational changes and
Ac ce p
te
350
18
Page 18 of 29
362
denaturation. Regaining protein structure is dependent on the polysaccharide content and type
363
used.
364
3.2.2. FTIR in amide II region
ip t
In relation to the band corresponding to amide II shown in Figure 3, the band at 1517 cm-1
365
was attributed to the vibrational modes of Tyrosine (Tyr) side chains (Hu, Kaplan, & Cebe,
367
2008). 0.30
0.30
ARA9
Control
us
368
cr
366
369 Absorbance (a. u.)
0.20
0.10 1600
373
1580
1560
1540
1520
1500
-1
Wavenumber (cm )
0.30
374
377
1480
0.10 1600
379
1540
1520
1500
1480
1500
1480
-1
Wavenumber (cm )
d te
0.25
0.20
0.15
0.10 1600
1560
DEX9
0.25
378
1580
0.30
Ac ce p
376
0.15
CMC9
Absorbance (a. u.)
375
0.20
M
0.15
372
1580
1560
1540
1520 -1
Wavenumber (cm )
1500
Absorbance (a. u.)
371
Absorbance (a. u.)
370
an
0.25
0.25
1480
0.20
0.15
0.10 1600
1580
1560
1540
1520 -1
Wavenumber (cm )
380
Figure 3. Original spectrum (black) and curve fitting spectra of amide II for soy protein
381
(control) and soy protein blends containing 9% gum arabic (ARA), carboxymethylcellulose
382
(CMC) or dextran (DEX).
383
The Tyr amino acids in SPI contribute at the interfaces of the β-sheet regions and act as
384
turns, so this band provided a specific local monitor for protein conformational changes due
385
to reduction of the conformational freedom in the protein structure (Murayama & Tomida,
386
2004; Reinstadler, Fabian, & Naumann, 1999), as shown quantitatively in Table 4. 19
Page 19 of 29
1517 cm-1
1544 cm-1
Control
26.49
73.51
ARA3
27.99
72.01
ARA6
27.85
72.15
ARA9
26.12
73.88
CMC3
24.72
75.28
CMC6
23.58
76.42
CMC9
22.15
77.85
DEX3
25.44
74.56
DEX6
23.78
76.22
DEX9
20.56
79.44
an
us
cr
ip t
Area Amide II
Table 4. Resulting percentage of the curve fitting of amide II for soy protein (control) and
388
soy protein blends with different contents of gum arabic (ARA), carboxymethylcellulose
389
(CMC) or dextran (DEX).
390
4. Conclusions
te
d
M
387
Soy protein-based blends were processed by twin-screw extrusion. During trials, a decrease
Ac ce p
391 392
in extrusion torque and specific mechanic energy was achieved through the addition of
393
polysaccharides to blends, indicating that a polysaccharide content of 6-9% in blends was
394
enough to decrease friction and facilitate macromolecules pass through the extruder die.
395
These improvements were the result of interactions between the extruder-induced
396
denaturation of soy protein and polysaccharides, which was analyzed by infrared
397
spectroscopy. The analysis of FTIR spectra in the band regions for amide I and II provided
398
novel information about conformational changes in the protein following extrusion. Although
399
the band frequency was maintained in a constant manner with respect to polysaccharide type
400
and content, further analysis of each band, which can be deconvoluted into bands
401
corresponding to the different secondary structures of the protein such as α-helix and β20
Page 20 of 29
sheets, showed that changes could be related to the decrease in SME values and to the
403
changes of some physical properties like MC and TSM. This study clearly highlights the
404
usefulness of employing FTIR as an analytical tool to assess conformational changes in the
405
structure of proteins and interactions that have taken place between biopolymers used in
406
blends following extrusion.
ip t
402
In summary, two features are the most relevant findings achieved in this work. Firstly, the
cr
407
utility of polysaccharide addition to facilitate the extrusion process, which is not the most
409
widely used method to produce this type of blends nowadays, although it is the most cost
410
effective process at industrial scale. Secondly, the value of using FTIR to analyze
411
conformational changes in globular proteins, providing novel information to understand the
412
relationship between protein structure and final properties of soy protein-based products.
413
Acknowledgments
M
an
us
408
The authors thank the University of the Basque Country (research group GIU12/06) as well
d
414
as the Basque Government (projects S-PE12UN002 and S-PE13UN057) for their financial
416
support.
417
References
418
Archana, G., Sabina, K., Babuskin, S., Radhakrishnan, K., Fayidh, M.A., Azhagu Saravana
419
Babu, P., Sivarajan, M., & Sukumar, M. (2013). Preparation and characterization of mucilage
420
polysaccharide for biomedical applications. Carbohydrate Polymers, 98, 89-94.
421
Alves, V.D., Costa, N., & Coelhoso, I.M. (2010). Barrier properties of biodegradable
422
composite films based on kappa-carrageenan/pectin blends and mica flakes. Carbohydrate
423
Polymers, 79, 269-276.
Ac ce p
te
415
21
Page 21 of 29
Andrade-Mahecha, M.M., Tapia-Blácido, D.R., & Menegalli, F.C. (2012). Development and
425
optimization of biodegradable films based on achira flour. Carbohydrate Polymers, 88, 449-
426
458.
427
Carpenter, J.F., & Crowe, J.H. (1989). An infrared spectroscopic study of the interactions of
428
carbohydrates with dried proteins. Biochemistry, 28, 3916-3922.
429
Cheftel, J.C., Cuq, J.L., & Lorient, D. (1985). In O.R. Fennema (Ed.), Food Chemistry (pp.
430
245-369). New York: Marcel Dekker.
431
Chen, F.L., Wei, Y.M., & Ojokoh, A.O. (2010). System parameters and product properties
432
response of soybean protein extruded at wide moisture range. Journal of Food Engineering,
433
96, 208-213.
434
Doublier, J.L., Garnier, C., Renard, D., & Sanchez, C. (2000). Protein-polysaccharide
435
interactions. Current Opinion in Colloid & Interface Science, 5, 202-214.
436
Félix, M., Martín-Alfonso, J.E., Romero, A., & Guerrero, A. (2014). Development of
437
albumen/soy biobased plastic materials processed by injection molding. Journal of Food
438
Engineering, 125, 7-16.
439
Fukushima, D., & Van Buren, J. (1970a). Effect of physical and chemical processing factors
440
on the redispersibility of dried soy milk proteins. Cereal Chemistry, 47, 571-578.
441
Fukushima, D., & Van Buren, J. (1970b). Mechanisms of protein insolubilization during the
442
drying of soy milk role of sisulfide and hydrophobic bonds. Cereal Chemistry, 47, 687-696.
443
Galicia-García, T., Martínez-Bustos, F., Jiménez-Arevalo, O., Martínez, A.B., Ibarra-
444
Gómez, R., Gaytán-Martínez, M., & Mendoza-Duarte, M. (2011). Thermal and
Ac ce p
te
d
M
an
us
cr
ip t
424
22
Page 22 of 29
microstructural characterization of biodegradable films prepared by extrusion-calendering
446
process. Carbohydrate Polymers, 83, 354-361.
447
Gilbert, V., Rouabhia, M., Wang, H., Arnould, A.L., Remondetto, G., & Subirade, M. (2005).
448
Characterization and evaluation of whey protein-based biofilms as substrates for in vitro cell
449
cultures. Biomaterials, 26, 7471-7480.
450
Godavarti, S., & Karwe, M.V. (1997). Determination of specific mechanical energy
451
distribution on a twin-screw extruder. Journal of Agricultural Engineering Research, 67,
452
277-287.
453
Gontard, N. & Guilbert, S. (1992). Edible wheat gluten films: influence of the main process
454
variables on film properties using response surface methodology. Journal of Food Science,
455
57, 190-195.
456
Goormaghtigh, E., Cabiaux, V., & Ruysschaert, J.M. (1990). Secondary structure and dosage
457
of soluble and membrane proteins by attenuated total reflection Fourier-transform infrared
458
spectroscopy on hydrated films. European Journal of Biochemistry, 193, 409-420.
459
Guerrero, P, Beatty, E., Kerry, J.P., & de la Caba, K. (2012). Extrusion of soy protein with
460
gelatin and sugars at low moisture content. Journal of Food Engineering, 110, 53-59.
461
Guerrero, P., Etxabide, A., Leceta, I., Peñalba, M., & de la Caba, K. (2014). Extraction of
462
agar from Gelidium sesquipedale (Rodhopyta) and surface characterization of agar based
463
films. Carbohydrate Polymers, 99, 491-498.
464
Guerrero, P., Garrido, T., Leceta, I., & de la Caba, K. (2013). Films based on proteins and
465
polysaccharides: Preparation and physical–chemical characterization. European Polymer
466
Journal, 49, 3713-3721.
Ac ce p
te
d
M
an
us
cr
ip t
445
23
Page 23 of 29
Guerrero, P., Nur Hanani, Z.A., Kerry, J.P., & de la Caba, K. (2011). Characterization of soy
468
protein-based films prepared with acids and oils by compression. Journal of Food
469
Engineering, 107, 41-49.
470
Harper, J.M. (1989). Food extruders and their applications. In C. Mercier, P. Linko, & J.P.
471
Harper (Eds.), Extrusion Cooking (pp. 1-15). Minnesota: American Assoc. Cereal Chemistry.
472
Hu, L., Hsieh, F., & Huff, H.E. (1993). Corn meal extrusion with emulsifier and soy bean
473
fiber. LWT-Food Science and Technology, 26, 544-551.
474
Hu, X., Kaplan, D., & Cebe, P. (2008). Dynamic protein-water relationships during β-sheet
475
formation. Macromolecules, 41, 3939-3948.
476
Hu, X., Lu, Q., Sun, L., Cebe, P., Wang, X., Zhang, X., & Kaplan, D. (2010). Biomaterials
477
from ultrasonicatio-induced silk fibroin-hyaluronic acid hydrogels. Biomacromolecules, 11,
478
3178-3188.
479
Klein, M., Aserin, A., Ben Ishai, P., & Garti, N. (2010). Interactions between whey protein
480
isolate and gum Arabic. Colloids and Surfaces B: Biointerfaces, 79, 377-383.
481
Kunte, L., Gennadios, A., Cuppett, S., Hanna, M., & Weller, C. (1997). Cast films from soy
482
protein isolates and fractions. Cereal Chemistry, 72, 115-118.
483
Leceta, I., Etxabide, A., Cabezudo, S., de la Caba, K., & Guerrero, P. (2014). Bio-based films
484
prepared with by-products and wastes: environmental assessment. Journal of Cleaner
485
Production, 64, 218-227.
486
Lefevre, T., & Subirade, M. (2000). Molecular differences in the formation and structure of
487
fine-stranded and particulate β-lactoglobulin gels. Biopolymers, 54, 578-586.
Ac ce p
te
d
M
an
us
cr
ip t
467
24
Page 24 of 29
Lefevre, T., Subirade, M., & Pezolet, M. (2005). Molecular description of the formation and
489
structure of plasticized globular protein films. Biomacromolecules, 6, 3209-3219.
490
Mangavel, C., Barbot, J., Popineau, Y., & Gueguen, J. (2001). Evolution of wheat gliadins
491
conformation during film formation: a Fourier transform infrared study. Journal of
492
Agricultural and Food Chemistry, 49, 867-872.
493
Murayama, K., & Tomida, M. (2004). Heat-induced secondary structure and conformation
494
change of bovine serum albumin investigated by Fourier transform infrared spectroscopy.
495
Biochemistry, 43, 11526-11532
496
Quinn, G., Monahan, F.J., O’Sullivan, M., & Langares, A. (2003). Role of covalent and
497
noncovalent interactions in the formation of films from unheated whey protein solutions
498
following pH adjustment. Journal of Food Science, 68, 2284-2288.
499
Reinstadler, D., Fabian, H., & Naumann, D. (1999). New structural insights into the refolding
500
of ribonuclease T1 as seen by time-resolved Fourier-transform infrared spectroscopy.
501
Proteins: Structure, Function, and Bioinformatics, 34, 303-316.
502
Rochas, C., Lahaye, M., & Yaphe, W. (1986). Sulfate content of carrageenan and agar
503
determined by infrared spectroscopy. Botanica Marina, 29, 335-340.
504
Rodríguez, J.M., & Pilosof, A.M.R. (2011). Protein-polysaccharide interactions at fluid
505
interfaces. Food Hydrocolloids, 25, 1925-1937.
506
Salarbashi, D., Tajik, S., Ghasemlou, M., Shojaee-Aliabadi, S., Shahidi Noghabi, M., &
507
Khaksar, R. (2013). Characterization of soluble soybean polysaccharide film incorporated
508
essential oil intended for food packaging. Carbohydrate Polymers, 98, 1127-1136.
Ac ce p
te
d
M
an
us
cr
ip t
488
25
Page 25 of 29
Song, F., Tang, D.L., Wang, X.L., & Wang, Y.Z. (2011). Biodegradable soy protein isolate-
510
based materials: a review. Biomacromolecules, 12, 3369-3380.
511
Subirade, M., Kelly, I., Gueguen, J., & Pezolet, M. (1998). Molecular basis of film formation
512
from a soybean protein: comparison between the conformation of glycinin in aqueous
513
solution and in films. International Journal of Biological Macromolecules, 23, 241-249.
514
Susi, H., Timasheff, S.N., & Stevens, L. (1967). Infrared spectra and protein conformations
515
in aqueous solutions: I. The amide I band in H2O and D2O solutions. Journal of Biological
516
Chemistry, 242, 5460-5466.
517
Wang, L., Xiao, M., Dai, S., Song, J., Ni, X., Fang, Y., Corke, H., & Jiang, F. (2014).
518
Interactions between carboxymethyl konjac glucomannan and soy protein isolate in blended
519
films. Carbohydrate Polymers, 101, 136-145.
520
Wang, C.H., & Damodaran, S. J. (1991). Thermal gelation of globular proteins. Journal of
521
Agricultural and Food Chemistry, 39, 433-438.
te
d
M
an
us
cr
ip t
509
Ac ce p
522
26
Page 26 of 29
522
Figure captions
524
Figure 1. Effect of polysaccharide content on FTIR spectra in the amide I and II region for
525
soy protein blends with gum arabic (ARA), carboxymethylcellulose (CMC), and dextran
526
(DEX).
527
Figure 2. Original spectrum (black) and curve fitting spectra of amide I for soy protein
528
(control) and soy protein blends containing 9% gum arabic (ARA), carboxymethylcellulose
529
(CMC) or dextran (DEX).
530
Figure 3. Original spectrum (black) and curve fitting spectra of amide II for soy protein
531
(control) and soy protein blends containing 9% gum arabic (ARA), carboxymethylcellulose
532
(CMC) or dextran (DEX).
533
Table captions
534
Table 1. Effect of gum arabic (ARA), carboxymethylcellulose (CMC), and dextran (DEX)
535
content on specific mechanical energy (SME), piece density (PD), expansion ratio (ER),
536
moisture content (MC) and total soluble matter (TSM).
537
a-e
538
different through the Tukey´s multiple range test.
539
Table 2. FTIR absorption area values for amide I (1628 cm-1), amide II (1541 cm-1), and
540
reference (2930 cm-1) bands as they pertain to soy protein blends with different gum arabic
541
(ARA), carboxymethylcellulose (CMC), and dextran (DEX) contents.
542
Table 3. Resulting percentage of the curve fitting of amide I for soy protein (control) and soy
543
protein blends with different contents of gum arabic (ARA), carboxymethylcellulose (CMC)
544
or dextran (DEX).
Ac ce p
te
d
M
an
us
cr
ip t
523
Two means followed by the same letter in the same column are not significantly (P > 0.001)
27
Page 27 of 29
Table 4. Resulting percentage of the curve fitting of amide II for soy protein (control) and
546
soy protein blends with different contents of gum arabic (ARA), carboxymethylcellulose
547
(CMC) or dextran (DEX).
Ac ce p
te
d
M
an
us
cr
ip t
545
28
Page 28 of 29
Page 29 of 29
ed
pt
ce
Ac M
an
us
cr