Food Research International 100 (2017) 674–681

Contents lists available at ScienceDirect

Food Research International

journal homepage: www.elsevier.com/locate/foodres

Formation and characterization of supramolecular structures of β- MARK

lactoglobulin and lactoferrin proteins
Camila Santiago Saraivaa, Jane Sélia dos Reis Coimbraa,⁎,
Alvaro Vianna Novaes de Carvalho Teixeirab, Eduardo Basílio de Oliveiraa,
Reinaldo Francisco Teófíloc, Angélica Ribeiro da Costaa, Éverton de Almeida Alves Barbosad
a
Universidade Federal de Viçosa, Departamento de Tecnologia de Alimentos, Campus Universitário S/N, 36570-000 Viçosa, Minas Gerais, Brazil
b
Universidade Federal de Viçosa, Departamento de Física, Campus Universitário S/N, 36570-000 Viçosa, Minas Gerais, Brazil
c
Universidade Federal de Viçosa, Departamento de Química, Campus Universitário S/N, 36570-000 Viçosa, Minas Gerais, Brazil
d
Universidade Federal de Viçosa, Departamento de Bioquímica e Biologia Molecular, Campus Universitário S/N, 36570-000 Viçosa, Minas Gerais, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: Combination of β-lactoglobulin (β-Lg) and lactoferrin (Lf), biomacromolecules derived from bovine whey, was
Acidification used in the formation of supramolecular structures by thermal gelation technique to adjust the pH. Furthermore,
Calorimetry the influence of the molar ratio, temperature, pH, and heating time in the formation of supramolecular structures
Gelation were also studied. The characterization of the protein supramolecular structures was performed using dynamic
Heating
light scattering, zeta potential measurements, molecular spectrofluorimetry, and circular dichroism spectro-
Molecular interaction
scopy. The thermal behavior of the pure proteins was investigated by differential scanning calorimetry. The
Self-assembly
protein denaturation temperatures were of around 85 °C for the β-Lg and around 52 °C and 85 °C (a small
portion) for the Lf. The protein molar ratio of 2:1 Lf/β-Lg was used to form the structures, whose character-
ization showed that the best conditions of supramolecular structure formation occurred at pH 6.5 and at tem-
peratures of 62.5 °C. In those conditions, more stable systems with reduced hydrophobic surface and average
sizes between 30 and 100 nm were generated. The correlation between pH and temperature suggests that the
method of preparation of the supramolecular structure affects its size during storage.

1. Introduction biocatalysts (Hosseinipour, Khiabani, Hamishehkar, & Salehi, 2015),

Nanostructures can be defined as stable supramolecular structures,
formed through non-covalent interactions among molecules, with at
least one of their dimensions measuring < 100 nm (Tamjidi, Shahedi,
Varshosaz, & Nasirpour, 2013). Such systems exhibit different physico-
chemical properties and biological activities when compared to these
same characteristics in a macroscale level, thus reflecting on new
techno-functionalities (Tamjidi et al., 2013). Accordingly, the nanos-
tructure characterization is often based on differences in chemical,
physical, and mechanical properties compared to the respective non-
structured molecules. Therefore, studying the nanostructure formation
processes, which comprise the evaluation of their predominant types of
interactions, conformational changes, architectures, and relative stabi-
lities and this is of utmost importance to set their functionality, and thus
to understand their applicability in food for industrial purposes. In food,
the applications of nanostructured systems involve, for example, flavor
retention (Chen, Guo, Wang, Yin, & Yang, 2016), development of
the carrying and the release of bioactive compounds (Madureira,
Pereira, & Pintado, 2016), functional ingredients (Bengoechea,
Peinado, & McClements, 2011), pigments (Zhaveh et al., 2015), and
antimicrobial agents (Esmailzadeh, Sangpour, Shahraz,
Hejazi, & Khaksar, 2016). Self-assembly is a manufacturing mechanism
of nanostructures which is based on a controlled balance among at-
tractive and repulsive forces among molecules, involving mainly non-
covalent interactions, such as Van der Waals forces and other hydro-
phobic interactions, as well as hydrogen bonds (Myrick,
Vendraa, & Krishnan, 2014). For food applications, nanostructures can
be obtained by protein gelation in dilute aqueous medium, mainly by
the heating process associated or not with the acidification of the
medium. The heating triggers the denaturation of proteins and their
subsequent aggregation via hydrophobic interactions and disulfide
bonds (Wolz, Mersch, & Kulozik, 2016).
Indeed, during heating, proteins undergo conformational changes
which induce the exposure of hydrophobic amino acid residues hidden

⁎
Corresponding author.
E-mail address: jcoimbra@ufv.br (J.S. dos Reis Coimbra).

http://dx.doi.org/10.1016/j.foodres.2017.07.065
Received 7 April 2017; Received in revised form 9 July 2017; Accepted 28 July 2017
Available online 29 July 2017
0963-9969/ © 2017 Elsevier Ltd. All rights reserved.
C.S. Saraiva et al. Food Research International 100 (2017) 674–681

in the core when the protein is in its native conformation. As con- Table 1
sequence, these residues become more available to react and to form Central Composite design with original values of three variables (pH, time e temperature)
with three central points (systems 15, 16 and 17) and variation of zeta potential (mV)
new intermolecular interactions, on the surface of the molecule, fa-
versus pH of the systems, the heating time (min) and temperature (°C).
voring the formation of supramolecular species (Li,
Dalgleish, & Corredig, 2015). The three main parameters influencing System pH Time (min) Temperature (°C) Zeta Potential (mV)
such phenomena are the temperature, pH, and superficial electrical
1 3.5 20 45 99.0
charge of the proteins. The architecture acquired by the nanostructure
2 10 20 45 − 13.1
(aggregate, fiber, nanotube or spherulite) depends on the manner that 3 3.5 60 45 88.3
these factors drive the intermolecular and intramolecular forces during 4 10 60 45 − 2.0
the process (Sanguansri & Augustin, 2006). Moreover, the presence of 5 3.5 20 80 97.8
free sulfhydryl groups on the protein surface further favors the inter- 6 10 20 80 − 32
7 3.5 60 80 89.1
actions (Y. Chen, Chen, Guo, & Zhou, 2015), as it is the case of β-lac-
8 10 60 80 − 4.2
toglobulin (β-Lg). 9 2 40 62.5 38.4
Whey proteins are being increasingly used in the nanotechnology 10 12 40 62.5 − 21.7
industry due to their technological and functional properties and for 11 6.5 6 62.5 21.0
12 6.5 74 62.5 11.5
being recognized as a safe nutritional ingredient (GRAS). These proteins
13 6.5 40 33 17.7
can act as gelling agent and can form aggregate whose particle sizes are 14 6.5 40 92 37.6
easily monitored. Moreover, they have the ability to combine poly- 15 6.5 40 62.5 20.6
saccharides and bioactive compounds among other materials (Ramos 16 6.5 40 62.5 40.1
et al., 2014). 17 6.5 40 62.5 35.1
Native Lf 5.5–6.8 – 25 15.6
Bovine whey proteins can be destined to a thermal processing in
Native β-Lg 5.5–6.8 – 25 − 10.2
order to bring about changes in their physical and functional properties.
The β-lactoglobulin, the major protein in the whey, has in its structure
two disulfide bonds and a reactive sulfhydryl group (Li et al., 2015). denaturation temperature of both pure proteins in aqueous solution. In
In this context, in the present study we intended to obtain supra- order to evaluate the behavior of supramolecular structures under dif-
molecular structures of bovine whey proteins, β-Lg and Lf, via an easily ferent combinations of surface electrical charges and conformational
feasible and simple thermal gelation procedure. The characterization of structures, different pH/temperature binomials were considered in the
the supramolecular structures was performed by applying molecular study.
fluorescence spectroscopy and circular dichroism. The stability and
particle size distribution were examined by means of dynamic light 2.2. Circular dichroism analyses
scattering analyses. The electrical charges on the surfaces of the su-
pramolecular structures were assessed through zeta potential mea- Circular dichroism spectra (CD) were recorded on a Jasco J-810
surements. spectropolarimeter (Jasco Corp., Japan), equipped with a Peltier type
temperature controller (PFD 425S, Jasco, Japan) coupled to a ther-
2. Materials and methods mostatic bath (AWC 100, Julabo, Germany). Spectra were measured at
25 °C, using a quartz cuvette (Hellma Analytics, Germany) at wave-
The proteins β-lactoglobulin (β-Lg, 93.6% purity) and lactoferrin lengths ranging from 190 nm to 260 nm, with 10 accumulations in each
(Lf, 96% purity) were a kind gift from Davisco Foods International Inc. experiment. The solutions containing supramolecular structures as well
(Eden Prairie, MN, USA) and from FrieslandCampina DMV as individual proteins solution were all analyzed at a concentration of
International (Veghel, Holland), respectively. Both proteins were used 2 g·L− 1. The latter, without heat treatment and acidification, were
as furnished without further purification. taken as controls. Two repetitions were performed. The CD data were
normalized to mean residual molar ellipticity, according to Eq. (1):
2.1. Production of β-Lg/Lf supramolecular structures
θ × 100 × MM
[θ] =
n×c×l (1)
The synthesis of β-Lg/Lf supramolecular structures followed a pro-
tocol based on that one firstly proposed by Hu, Yu, & Yao (2007), in In Eq. (1), θ is the CD signal (degrees), MM is the protein molar mass
which the nanostructures are formed by thermal gelation, or self-as- (kDa), n is the number of amino acid residues, c is the protein con-
sociation, and prior adjustment of the pH of the protein solutions. In- centration (g·L− 1) and l is the optical path of the cuvette. In order to
itially, aqueous solutions of β-Lg and Lf were separately prepared, at a find MM and n of the supramolecular structures, the weighing of the
concentration of 2 g·L− 1 each. Volumes of these solutions were mixed two proteins was made. The deconvolution of mean residue ellipticity
so that the β-Lg: Lf molar ratio in the resulting mixture was 1:2; 1:1, or data was made, according to the procedure followed by Amorim et al.
2:1. The mixture was stirred on a magnetic stirrer (TE-0851, Tecnal, (2016). In parallel to this analysis, the CD signals of aqueous solutions
Brazil) for 30 min without pH adjustment, at 25 °C. After that, the so- of the pure proteins were evaluated in a temperature ranging from (10
lutions that presented some turbidity were tested in conditions of to 100) °C, at (208, 215, and 222) nm.
pH 6.5, temperature of 62.5 °C, and heated for 40 min. The β-Lg:Lf
molar ratio for subsequent studies was chosen based on the formation of 2.3. Fluorescence spectroscopy analyses
a turbid solution with no precipitate after such treatment.
Systems containing the chosen β-Lg:Lf molar ratio were formed Fluorescence analyses were performed using a K2 spectro-
from the mixture of the two protein solutions and subjected to gentle fluorometer (ISS, USA). The wavelength of the tryptophan (Trp) and
agitation for 12 h on a magnetic stirrer (TE-0851, Tecnal, Brazil) at 4 °C tyrosine (Tyr) excitation source is 280 nm and the emission spectra
in a BOD (SP-500, SPLabor). Then, the pH of the solutions were ad- were obtained at 290–450 nm. At 280 nm, Trp and Tyr residues are
justed using HCl (0.1 mol·L− 1) or NaOH (0.1 mol·L− 1) with subsequent excited and the contribution of phenylalanine (Phe) can be neglected
heating using a thermostatic bath (Bath Thermostatted, TE-184, Tecnal, (Ghisaidoobe & Chung, 2014; Teale & Weber, 1957). The selected
Brazil). Different values of pH, temperatures and heating times were emission range provides a fluorescence signal coming from the resultant
tested, following an experimental design (Table 1). The pH and tem- (summed) contribution of Trp and Tyr. The analysis of the solution
perature values were defined according to the isoelectric points and the containing the supramolecular structures with concentration 2 g·L− 1,

675
C.S. Saraiva et al. Food Research International 100 (2017) 674–681

Column A Column B

Fig. 1. Spectra of denaturation (blue curve) and refolding (red curve) of proteins. Column A: β-Lg 208, 215 and 222 nm, respectively. Column B: Lf 208, 215 and 222 nm, respectively.
(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

was performed at 25 °C in quartz cuvette with 1 cm optical path. Pure surface electrical charge of pure proteins and their supramolecular
protein solutions of β-Lg and Lf at a concentration of 2 g·L− 1, without structures. Measurements of the systems and pure proteins were made,
heat treatment and acidification, were taken as controls. Two repeti- all with a concentration of 2 g·L− 1 at 25 °C. Two repetitions were
tions were carried out. carried out.

2.4. Zeta potential analyses 2.5. Dynamic light scattering analyses

Zeta measurements were performed on a Zeta Sizer Nano (ZS90 Particle size measurements were made using the Zetasizer Nano S
Multi-Purpose Titrator MPT Malvern 2, UK) equipment to assess the net equipment (Malvern Instruments Ltd., UK). The instrument is equipped

676
C.S. Saraiva et al.
Fig. 2. CD spectra of the native proteins β-Lg (black) and Lf (red) and supramolecular
structures in different systems. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)
with a laser He/Ne 4 mW that emits light with wavelength of 632.8 nm,
a measuring cell, and an avalanche photodiode as detector. The
equipment correlator covers a time range from 500 ns to 62.9 s. The
concentration of the samples was 2 g·L− 1, which were analyzed at
(25.0 ± 0.1) °C in a rectangular polystyrene cuvette. The scattered
intensity was measured under a 173° detection angle with respect to the
source. Intensity autocorrelation functions were analyzed by the
CONTIN algorithm (Integrated in the equipment software) to determine
677
Food Research International 100 (2017) 674–681
the intensity size distribution. Each measurement represents the
average obtained from at least 10 measures accumulated during 60 s.
Two repetitions were carried out.
2.6. Data analysis
The experimental design was the main compound with three vari-
ables and three central points (Table 1) similarly to Teófilo & Ferreira
(2006), in order to evaluate the combined influence of heating time (6
to 74 min), pH (2 to 12), and temperature (33 to 92 °C) on the size of
the produced supramolecular structures. The values of the variables
(pH, time and temperature) were chosen based on the isoelectric points
and the denaturation temperatures of the β-Lg proteins (pI: 5.2 and dT:
85 °C) and Lf (pI: 8 and dT: 50 °C). The denaturation temperatures of
both proteins were found in a previous differential scanning calori-
metry experiment (Fig. 2A). Through this analysis, the optimal condi-
tions to produce the protein supramolecular structure were determined.
Data from supramolecular structures stability analyzes were processed
using STATISTICA version 7.0 software and a worksheet for central
composite design calculations with three variables.
3. Results and discussion
3.1. Production of β-Lg/Lf supramolecular structures
The systems with Lf:β-Lg molar ratio of 1:2 and 1:1, after treatment
(pH and heating), showed an increase of turbidity, with consequent
formation of undesirable precipitates. Conversely, the system con-
taining Lf:β-Lg molar ratio of 2:1 presented the best turbidity, i.e.,
opaque system in the absence of precipitates. Therefore, this molar ratio
between the two proteins was chosen to perform the subsequent steps of
this study.
3.2. Circular dichroism
In order to assess the thermal structural stability of the pure pro-
teins, they were analyzed by circular dichroism (CD), in which their
signal at 208, 215, and 222 nm during the consecutive increase in
temperature from 10 to 100 °C was monitored. Thus, a study of the pure
proteins behavior in aqueous solution during the denaturation and re-
folding process (reverse path) (Fig. 1) was developed. For β-Lg (Column
A) the temperatures of 60 °C and 80 °C did not provide complete de-
naturation, since the curve still changes its profile at higher tempera-
tures. Nevertheless, this protein is well-known for presenting capacity
of renaturation (Hattori, Ametani, Katakura, Shimizu, & Kaminogawa,
1993). A similar behavior was observed for Lf, but contrarily to β-Lg
this protein is not likely to renature when heat up to temperatures as
high as 100 °C (Paulsson, Svensson, Kishore, & Naidu, 1993). The var-
iation of the temperature profile analyzed at the three wavelengths
(208, 215 and 222 nm) allows to conclude the presence of secondary
structures (α-helix or β-sheet) in proteins. These changes were detected
only at 215 nm for β-Lg, which confirms the predominance of β-sheet
structure. For the Lf, the signal was detected in all three wavelengths
examined, as Lf has a secondary structure composed of both α-helix and
β-sheet arrangements. As the structure of β-Lg is rich in β-sheet a more
intense minimum could be seen at 215 nm. The Lf structure is com-
posed of α-helix by the existence of two negative parallels (208 and
222 nm), but the second negative parallel was located at 218 nm, which
also indicates the presence of β-sheet (Fig. 2).
The formation of protein nanoparticles is often accompanied by
their structural change (Lundqvist, Sethson, & Jonsson, 2004; Deng,
Liang, Monteiro, Toth, & Minchin, 2011; Cukalevski, Lundqvist, Osla-
kovic, Dahlbäck, Linse, & Cedervall, 2011; De Paoli Lacerda, Park,
Meuse, Pristinski, Becker, Karim, & Douglas, 2010). All supramolecular
structures (Fig. 2) possessed differentiated degree of structural changes
of both β-Lg and Lf. However, the CD results suggested a predominance
C.S. Saraiva et al. Food Research International 100 (2017) 674–681

of secondary arrangements of Lf structure, since they contained a larger

amount of this protein. Also, CD spectra for supramolecular structures
produced at pH 6.5 showed a more pronounced change in secondary
structures when compared to those produced at pH 3.5 and pH 10.
Diniz et al. (2014) studied conformational changes of supramolecular
structures formed by α-lactalbumin and GMP systems and observed
that, at pH 3.5, the temperature increase did not cause substantial
changes in the secondary structure of these supramolecular species.
The CD spectra of supramolecular structures produced at pH 3.5 and
pH 10 showed a subtle difference between them. Supramolecular
structures produced at pH 3.5 were more likely to present random
secondary arrangements (Fig. 2a–b), as indicated by the decrease of the
molar ellipticity (θ) of the signal associated with the displacement of
the most negative minimum to the region around 200 nm. Sreedhara
et al. (2010) demonstrated that the structural status of the Lf molecule
under acid conditions is different from its native state. The Lf molecule
adopts a more randomized conformation at pH 2.0 to 3.0.
A common feature in the three graphs was noticed: higher tem-
peratures led to greater unfolding of proteins and consequently caused
a larger loss of secondary structures, as expected. This is in accordance
with Day et al. (2014), who studied the changes in the secondary
structure of lysozyme caused by heating. When evaluating the CD
spectra of Lf/β-Lg supramolecular structures produced at 62.5 °C
(which is intermediate to the denaturation temperature of β-Lg and Lf
(Fig. 2c), a loss of secondary structure can be observed. At pH 2 and
pH 12, a random conformation, characterized by strong minimum at
205 nm, was predominant. The spectrum of a random coil protein
(“random coil”) has a negative band with large amplitude around
200 nm (Corrêa & Ramos, 2009).
The deconvolution of the residual molar ellipticity data of the sys-
tems (Table 2) was performed using the CONTIN/LL - Reference 7
analysis method, as described by Sreerama & Woody (2000).
In general, the treatments (pH, time and temperature) applied to
protein systems contributed to a reduction in the amount of α-helix and
an increase in the disordered structures. This indicates a loss of the
secondary structure of the proteins, confirming the results expressed in
Fig. 2.

3.3. Fluorescence spectroscopy

The tertiary structure of the proteins was greatly altered with low
pH values (pH values < 6.5 were studied in this work), thus favoring

Table 2
Mean values of deconvolution of the residual molar ellipticity data of different systems.

System Alpha-helix Beta-sheet Turns (%) Disordered structure (%)

(%) (%)

1 14.8 28.4 17.0 39.8

2 16.9 32.9 20.5 29.7
3 15.7 28.4 19.2 36.6
4 13.8 33.1 20.4 32.7
5 6.3 37.5 19.9 36.3
6 5.9 38.7 19.4 36.0
7 13.4 31.7 19.7 35.1 Fig. 3. Emission spectra of chromophores of the native proteins β-Lg (black) and Lf (red),
8 9.9 39.3 18.4 32.4 and of the supramolecular structures in different systems. (For interpretation of the re-
9 8.7 40.5 19.8 30.9 ferences to color in this figure legend, the reader is referred to the web version of this
10 8.8 29.5 17.2 44.5 article.)
11 7.6 37.6 20.3 34.6
12 4.5 40.1 20.1 35.3
13 16.0 32.3 20.6 31.0 the exposure of the tryptophan (Trp) residues and tyrosine (Tyr). These
14 5.2 39.4 19.3 36.1 residues were hidden inside the native conformation of the protein
15 7.4 38.4 20.6 33.6 molecules, which resulted in an increase in fluorescence emission in-
16 5.8 39.1 20.0 35.1 tensity (Fig. 3a–b).
17 5.2 39.7 19.9 35.1
White 23.3 27.6 20.4 28.7
At pH 10, the reduction of hydrophobic surface of the supramole-
cular structures occurs, since the emission intensity is reduced.
*White corresponds to the β-Lg + Lf system in solution, with no heat treatment and no Probably proteins end up hiding the Trp residues and Tyr in their
pH adjustment.

678
C.S. Saraiva et al. Food Research International 100 (2017) 674–681

interior (Fig. 3a–b) to reorganize their structures. Table 3

Supramolecular structures produced at pH 6.5, had emission in- Dynamic light scattering of the native proteins (β-lg and Lf) and of the protein supra-
molecular structure in different systems.
tensity inferior to the observed for other systems, since at this pH value
the β-Lg (pI: 5.2) and Lf (pI: 8) have opposite net electrical charges, so System Mean size diameter (nm) % Volume distribution
attractive interaction between them is favored. This suggests that,
during protein aggregation, a greater amount of the Trp residues and 9 Peak 1 164.90 94.8
Peak 2 5076.00 5.20
Tyr get more occluded within the formed structure. High temperatures
Peak 3 0.00 0.00
cause greater changes in the tertiary structure of the protein, leading to 10 Peak 1 146.60 0.60
increased Trp exposure and Tyr, and consequently, to an increase in Peak 2 12.70 99.40
emissions (Fig. 3c). Peak 3 4901.00 0.00
In comparison to pH and temperature, the heating time did not have 13 Peak 1 1177.00 0.20
Peak 2 15.00 99.80
such a strong influence on the conformational changes of the side
Peak 3 4804.00 0.00
chains of the proteins' molecules. Indeed, the fluorescence emission 14 Peak 1 85.00 100.00
spectra exhibited very similar profiles behavior for systems submitted Peak 2 0.00 0.00
to heating for different times. This implies the greater relevance of the Peak 3 0.00 0.00
15 Peak 1 103.30 99.20
definition of pH and temperature for the change of secondary and
Peak 2 5005.00 0.80
tertiary structure of proteins (β-Lg and Lf), which is desirable in the Peak 3 0.00 0.00
process of formation of supramolecular structures. β-Lg Peak 1 2.98 99.50
According to Zhang, Wright, & Zhong (2013), the β-Lg has two Peak 2 12.10 0.40
tryptophan residues (Trp-19 and Trp-61) that have intrinsic fluores- Peak 3 30.50 0.10
Lf Peak 1 13.59 98.80
cence. So the effect of pH on the protein structure can be revealed by
Peak 2 305.20 0.80
fluorescence spectra. These authors studied the effect of adding bixin, a Peak 3 1159.00 0.40
carotenoid used as a dye, in a β-Lg solution with different pH-values,
and concluded that when the β-Lg is pure in solution, the basic con-
ditions strongly affect the structure of this protein. When the bixin is thermal denaturation can be caused because the heating may promote
added, there is a decrease in fluorescence emission due to conforma- chemical degradation of certain charged groups. This may explain the
tional changes in β-Lg structure in order to leave occluded Trp residues. fact that the system 12 (pH 6.5; 74 min; 62.5 °C) has shown lower zeta
potential than the other systems with the same pH. Additionally, this
3.4. Zeta potential system showed significant conformational changes in the secondary
structure (Fig. 1A).
The zeta potential analysis allows the evaluation of the superficial
electrical charge of dispersed particles and macromolecules. 3.5. Dynamic light scattering
Measurements of zeta potential of the supramolecular structured sys-
tems were carried out obtaining charge spectra on the surface of such In dynamic light scattering experiments, the diffusivity of the mo-
systems (Table 1). lecules suspended in a viscous medium is reflected as a fluctuation of
For systems with pH values below the isoelectric point (IP) of β-Lg the intensity of the scattered light. The analysis of the self-correlation of
(5.2), the zeta potential presented positive values, whereas above the IP the scattered intensity gives information about the diffusivity or, in the
of Lf (8–9), the zeta potential charge was predominantly negative. last instance, the size of the many species suspended in a solution.
Lam & Nickerson (2014) reported that, an increase in the pH of β-Lg The analysis of the correlograms of pure proteins in aqueous solu-
dispersions leads to a decrease in the zeta potential value from ap- tion with a concentration of 2 mg·mL− 1 to 25 °C, gives a monomodal
proximately +22.9 mV (at pH 3.0) to around − 26.4 mV (at pH 5.0), distribution with mean diameter to a β-Lg of approximately 3 nm and to
reaching − 42.1 mV (at pH 7.0). According to these authors, depending the Lf around 14 nm (Table 3).
on how the aggregation occurs and which solvent involves the protein At basic pH (10 and 12), the average particle diameter at elevated
molecules, the superficial charge can change. This may explain the temperature was about 28 nm, while at lower temperatures, it was
decrease in zeta potential value of the system 9 with respect to the around 15 nm. At high pH (10 and 12), both β-Lg and Lf acquire ne-
systems 1, 3, 5, and 7, when actually what one expected was that the gative charges, increasing the electrostatic repulsion between the two
system presented a greater zeta potential due to its low pH. molecules. This may explain the small size of particles that were found
Conditions of pH and temperature imposed during system 9 pro- at pH 12. Similar results were found by Bourbon et al. (2015), when
duction, resulted in differences in the protein conformation, as it can be investigating the hydrodynamic diameter of nanohydrogels formed by
observed in the circular dichroism spectrum for this system (Fig. 2c). Lf and GMP (glycomacropeptide). At acidic pH (2.0–6.5), there was a
Besides, this resulted in inducing the association-dissociation of these tendency to form particles with much larger sizes, reaching a diameter
proteins subunits, which probably caused a reduction in the value of of 231 nm. In average, pH values < 5.2 molecules of both β-Lg and Lf
zeta potential. present positive net charges and, therefore, at a first attempt, one could
Under conditions of intermediate pH of the IP of β-Lg (5.2) and Lf expect no protein aggregation in such media, due to electrostatic re-
(8–9), the supramolecular structured systems showed positive zeta pelling. Nonetheless, combined pH changes and heat exposure provided
potential values. This result was expected for the used Lf/β-Lg molar important conformational modifications of the structure of the proteins'
ratio, which was 2:1, so the electrical charge of the protein present in architecture, as revealed by CD and fluorescence spectroscopy analyses
higher proportion was predominant. (Figs. 2 and 3, respectively). Two consequences of this is the exposure
In the analysis of the zeta potential of pure proteins, the expected of several hydrophobic amino acid side chains and the release of
electrical charges for both proteins between pH 5.5 and 6.8 were con- counter-ions and water molecules from the protein surface (Adamson,
firmed. The β-Lg had negative charge (ζ = − 10.2 mV) while Lf had 1982). Such hydrophobic effects, along with a rising in configurational
positive charge (ζ = + 15.6 mV). When the pure proteins (without entropy in the system, may be in some cases stronger than electrostatic
heating) are compared with the heated proteins in systems with pH 6.5 attraction, which may explain the formation of structures of larger sizes
(11–17), zeta potential values became larger in modulus. even at low pH values, when the signals of the protein charges are the
According to Peinado, Lesmes, Andrés & McClements (2010), the same. Blake, Amin, Qi, Majumdar & Lewis (2015) also found similar
change in the electrical characteristics of the Lf protein before the results, in relation to the increase in pH value before and after the

679
C.S. Saraiva et al.
Fig. 4. Average diameter of systems over the storage time.
temperature of aggregation, while studying the behavior of β-lacto-
globulin during the process of self-assembly.
At pH 6.5, particles with average diameters slightly lower than
100 nm were formed, since under such condition the proteins presented
opposite electrical charges, favoring electrostatic attractive interactions
among the molecules and thus leading to the formation of conjugates
with larger diameters (Nigen, Gaillard, Croguennec,
Madec, & Bouhallab, 2010). For supramolecular structures composed of
α-lactalbumin and lysozyme, Monteiro et al. (2016) noticed that, at
pH 7.0, there was the formation of structures at microscale.
As shown in Table 2, higher temperatures favored the increase in
size of supramolecular protein structures for acidic pH (2) and near
neutral (6.5). The importance of the temperature in the formation of
larger structures can be seen since a bi-modal distribution is observed at
33 °C with the smaller size population with diameter close to the native
Fig. 5. Supramolecular structures behavior charts during the storage time depending on their method of preparation.
680
Food Research International 100 (2017) 674–681
Lf (~ 15 nm) and a larger size population with diameter around
110 nm. In larger temperatures (62.5 °C and 92 °C) only the larger
population is observed. This suggests that part of Lf does not form su-
pramolecular structures with the β-Lg at small temperatures. This fol-
lows the trend reported by Diniz et al. (2014), when evaluating the
influence of temperature on such supramolecular structures of α-lac-
talbumin and GMP.
3.6. Stability assessment
In order to evaluate the stability of the systems, dynamic light
scattering analyses were performed along a storage time of 45 days.
From the most stable spectra, analysis of variance (ANOVA) of the
generated data (Table. 1B) shows that only for the systems 2, 4, 6, 8, 13,
and 15, the average diameter did not significantly differ throughout the
storage time (p-value > α). Fig. 4 illustrates such systems that were
statistically more stable over time.
In addition, the supramolecular structures method of preparation
affects their sizes during storage. The occurred change is related to the
pH and to the preparation temperature. Until the time of 20 days
(Fig. 5), the systems followed a trend in their behavior with respect to
preparation conditions. After that time, the analysis of the behavior was
not possible, since the heating time variable became significant.
4. Conclusion
Working with the Lf/β-Lg molar ratio of 2:1 the optimal conditions
for obtaining supramolecular structures occurred at pH 6.5 and tem-
peratures of 62.5 °C. Under these conditions, stable supramolecular
protein structures with small hydrophobic surface and average sizes
C.S. Saraiva et al.
ranged from 30 to 100 nm were formed. The formation of such supra-
molecular structures was accompanied by some unfolding of the sec-
ondary structure of both β-Lg and Lf. The exposure of such proteins to
extreme pH (2 and 12) was associated to the predominance of “random
coil”, and elevated temperatures such as 80 °C led to a major loss of the
secondary structure. Supramolecular structures produced at pH 2 and
3.5 presented a negative electric charge, at pH 10 and 12, positive
electrical charges, and at pH 6.5, presented electric charge of the
highest protein content, in this case, the Lf. Therefore, the zeta potential
was positive at that pH value.
Overall, this work showed that it is possible to obtain stable su-
pramolecular structures from two whey proteins, β-Lg and Lf, by using
thermal gelling technique with adjustment of the pH.
This result allows the production of protein supramolecular struc-
tures in a simple way and at a low cost and it enables the use of whey as
well.
Future prospects involve studies on the incorporation and controlled
release of bioactive compound and the toxicology of β-Lg and Lf su-
pramolecular structures.
Acknowledgments
The authors would like to thank the support of Professor Dr.
Leandro Licursi de Oliveira from the Departamento de Biologia Geral/
Universidade Federal de Viçosa, CNPq (Conselho Nacional de
Desenvolvimento Científico e Tecnológico), CAPES (Coordenação de
Aperfeiçoamento de Pessoal de Nível Superior), FAPEMIG (Fundação de
Amparo à Pesquisa do Estado de Minas Gerais), CNPEM-LNBio (Centro
Nacional de Pesquisa em Energia e Materiais, Laboratório Nacional de
Biociências, São Paulo), the Laboratório de Espectroscopia UV–Vis
(Universidade Federal de Minas Gerais, Minas Gerais), and Prof. Edwin
Elard Garcia Rojas from the Laboratório de Engenharia e Tecnologia
Agroindustrial (Escola de Engenharia Industrial e Metalúrgica/
Universidade Federal Fluminense, Rio de Janeiro).
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.foodres.2017.07.065.
References
Adamson, G. K. (1982). Physical chemistry of surface (4 ed.). New York: John Wiley & Sons.
Amorim, M. L., Ferreira, G. M. D., Soares, L. S., Soares, W. A. S., Ramos, A. M., Coimbra, J.
S. R., ... Oliveira, E. B. (2016). Physicochemical aspects of chitosan dispersibility in
acidic aqueous media: Effects of the food acid counter-anion. Food Biophysics, 11,
388–399. http://dx.doi.org/10.1007/s11483-016-9453-4.
Bengoechea, C., Peinado, I., & McClements, D. J. (2011). Formation of protein nano-
particles by controlled heat treatment of lactoferrin: Factors affecting particle char-
acteristics. Food Hydrocolloids, 25(5), 1354–1360. http://dx.doi.org/10.1016/j.
foodhyd.2010.12.014.
Blake, S., Amin, S., Qi, W., Majumdar, M., & Lewis, E. N. (2015). Colloidal
Stability & Conformational Changes in β-Lactoglobulin: Unfolding to Self-Assembly.
International Journal of Molecular Sciences, 16(8), 17719–17733. http://dx.doi.org/
10.3390/ijms160817719.
Bourbon, A. I., Pinheiro, A. C., Carneiro-da-Cunha, M. G., Pereira, R. N., Cerqueira, M. A.,
& Vicente, A. A. (2015). Development and characterization of lactoferrin-GMP na-
nohydrogels: evaluation of pH, ionic strength and temperature effect. Food
Hydrocolloids, 48, 292–300. http://dx.doi.org/10.1016/j.foodhyd.2015.02.026.
Chen, X.-W., Guo, J., Wang, J.-M., Yin, S.-W., & Yang, X.-Q. (2016). Controlled volatile
release of structured emulsions based on phytosterols crystallization. Food
Hydrocolloids, 56, 170–179. http://dx.doi.org/10.1016/j.foodhyd.2015.11.035.
Chen, Y., Chen, X., Guo, T. L., & Zhou, P. (2015). Improving the thermostability of β-
lactoglobulin via glycation: The effect of sugar structures. Food Research International,
69, 106–113. http://dx.doi.org/10.1016/j.foodres.2014.12.017.
Corrêa, D. H. A., & Ramos, C. H. I. (2009). The use of circular dichroism spectroscopy to
study protein folding, form and function. African Journal of Biochemistry Research,
3(5), 164–173.
Day, L., Zhai, J., Xu, M., Jones, N. C., Hoffmann, S., & Wooster, T. J. (2014).
Conformational changes of globular proteins adsorbed at oil-in-water emulsion in-
terfaces examined by synchrotron radiation circular dichroism. Food Hydrocolloids,
34, 78–87. http://dx.doi.org/10.1016/j.foodhyd.2012.12.015.
Deng, Z. J., Liang, M., Monteiro, M., Toth, I., & Minchin, R. F. (2011). Nanoparticle-
induced unfolding of fibrinogen promotes Mac-1 receptor activation and
681
Food Research International 100 (2017) 674–681
inflammation. Nature Nanotechnology, 6(1), 39–44. http://dx.doi.org/10.1038/
nnano.2010.250.
Diniz, R. S., dos Reis Coimbra, J. S., de Carvalho Teixeira, Á. V. N., da Costa, A. R., Santos,
I. J. B., Bressan, G. C., Rodrigues, A. M. C., & da Silva, L. H. M. (2014). Production,
characterization and foamability of α-lactalbumin/glycomacropeptide supramole-
cular structures. Food Research International, 64, 157–165. http://dx.doi.org/10.
1016/j.foodres.2014.05.079.
Esmailzadeh, H., Sangpour, P., Shahraz, F., Hejazi, J., & Khaksar, R. (2016). Effect of
nanocomposite packaging containing ZnO on growth of Bacillus subtilis and
Enterobacter aerogenes. Materials Science and Engineering: C, 58, 1058–1063. http://dx.
doi.org/10.1016/j.msec.2015.09.078.
Ghisaidoobe, A. B., & Chung, S. J. (2014). Intrinsic tryptophan fluorescence in the de-
tection and analysis of proteins: a focus on Förster resonance energy transfer tech-
niques. International Journal of Molecular Sciences, 15(12), 22518–22538. http://dx.
doi.org/10.3390/ijms151222518.
Hattori, M., Ametani, A., Katakura, Y., Shimizu, M., & Kaminogawa, S. (1993).
Unfolding/refolding studies on bovine β-lactoglobulin with monoclonal antibodies as
probe. The Journal of Biological Chemistry, 268(30), 22414–22419.
Hosseinipour, S. L., Khiabani, M. S., Hamishehkar, H., & Salehi, R. (2015). Enhanced
stability and catalytic activity of immobilized α-amylase on modified Fe3O4 nano-
particles for potential application in food industries. [journal article]. Journal of
Nanoparticle Research, 17(9), 1–13. http://dx.doi.org/10.1007/s11051-015-3174-3.
Hu, J., Yu, S., & Yao, P. (2007). Stable amphoteric nanogels made of ovalbumin and
ovotransferrin via self-assembly. Langmuir, 23(11), 6358–6364. http://dx.doi.org/10.
1021/la063419x.
Lam, R. S., & Nickerson, M. T. (2014). The Effect of pH and Heat Pre-Treatments on the
Physicochemical and Emulsifying Properties of β-lactoglobulin. Food Biophysics, 9(1),
20–28.
Li, Y., Dalgleish, D., & Corredig, M. (2015). Influence of heating treatment and membrane
concentration on the formation of soluble aggregates. Food Research International,
76(Part 3), 309–316. http://dx.doi.org/10.1016/j.foodres.2015.06.016.
Lundqvist, M., Sethson, I., & Jonsson, B.-H. (2004). Protein adsorption onto silica nano-
particles: Conformational changes depend on the particles' curvature and the protein
stability. Langmuir, 20(24), 10639–10647.
Madureira, A. R., Pereira, A., & Pintado, M. (2016). Chitosan nanoparticles loaded with
2,5-dihydroxybenzoic acid and protocatechuic acid: Properties and digestion. Journal
of Food Engineering, 174, 8–14. http://dx.doi.org/10.1016/j.jfoodeng.2015.11.007.
Monteiro, A. A., Monteiro, M. R., Pereira, R. N., Diniz, R., Costa, A. R., Malcata, F. X.,
Teixeira, J. A., Teixeira, A. V., Oliveira, E. B., Coimbra, J. S., Vicente, A. A., & Ramos,
O. L. (2016). Design of bio-based supramolecular structures through self-assembly of
α-lactalbumin and lysozyme. Food Hydrocolloids, 58, 60–74. http://dx.doi.org/10.
1016/j.foodhyd.2016.02.009.
Myrick, J., Vendraa, V., & Krishnan, S. (2014). Self-assembled polysaccharide nanostructures
for controlled-release applications (publication no). http://dx.doi.org/10.1515/ntrev-
2012-0050).
Nigen, M., Gaillard, C., Croguennec, T., Madec, M.-N., & Bouhallab, S. (2010). Dynamic
and supramolecular organisation of α-lactalbumin/lysozyme microspheres: A mi-
croscopic study. Biophysical Chemistry, 146(1), 30–35. http://dx.doi.org/10.1016/j.
bpc.2009.10.001.
Paulsson, M. A., Svensson, U., Kishore, A. R., & Naidu, A. S. (1993). Thermal behavior of
bovine lactoferrln in water and its relation to bacterial interaction and antibacterial
activity. Journal of Dairy Science, 76, 3711–3720.
Peinado, I., Lesmes, U., Andrés, A., & McClements, J. D. (2010). Fabrication and mor-
phological characterization of biopolymer particles formed by electrostatic com-
plexation of heat treated lactoferrin and anionic polysaccharides. Langmuir, 26(12),
9827–9834. http://dx.doi.org/10.1021/la1001013.
Ramos, O. L., Pereira, R. N., Rodrigues, R., Teixeira, J. A., Vicente, A. A., & Malcata, F. X.
(2014). Physical effects upon whey protein aggregation for nano-coating production.
Food Research International, 66, 344–355. http://dx.doi.org/10.1016/j.foodres.2014.
09.036.
Sanguansri, P., & Augustin, M. A. (2006). Nanoscale materials development — a food
industry perspective. Trends in Food Science & Technology, 17(10), 547–556. http://
dx.doi.org/10.1016/j.tifs.2006.04.010.
Sreedhara, A., Flengsrud, R., Prakash, V., Krowarsch, D., Langsrud, T., Kaul, P., Devold, T.
G., & Vegarud, G. E. (2010). A comparison of effects of pH on the thermal stability
and conformation of caprine and bovine lactoferrin. International Dairy Journal,
20(7), 487–494. http://dx.doi.org/10.1016/j.idairyj.2010.02.003.
Sreerama, N., & Woody, R. W. (2000). Estimation of protein secondary structure from
circular dichroism spectra: comparison of CONTIN, SELCON, and CDSSTR methods
with an expanded reference set. Analytical Biochemistry, 287(2), 252–260. http://dx.
doi.org/10.1006/abio.2000.4880.
Tamjidi, F., Shahedi, M., Varshosaz, J., & Nasirpour, A. (2013). Nanostructured lipid
carriers (NLC): A potential delivery system for bioactive food molecules. Innovative
Food Science & Emerging Technologies, 19, 29–43. http://dx.doi.org/10.1016/j.ifset.
2013.03.002.
Teale, F. W. J., & Weber, G. (1957). Ultraviolet fluorescence of the aromatic amino acids.
Biochemical Journal, 65, 476–482.
Teófilo, R. F., & Ferreira, M. (2006). Chemometrics II: spreadsheets for experimental
design calculations, a tutorial. Química Nova, 29(2), 338–350. http://dx.doi.org/10.
1590/S0100-40422006000200026.
Wolz, M., Mersch, E., & Kulozik, U. (2016). Thermal aggregation of whey proteins under
shear stress. Food Hydrocolloids, 56, 396–404. http://dx.doi.org/10.1016/j.foodhyd.
2015.12.036.
Zhang, Y., Wright, E., & Zhong, Q. (2013). Effects of pH on the molecular binding be-
tween β-lactoglobulin and bixin. Journal of Agricultural and Food Chemistry, 61(4),
947–954. http://dx.doi.org/10.1021/jf303844w.
Zhaveh, S., Mohsenifar, A., Beiki, M., Khalili, S. T., Abdollahi, A., Rahmani-Cherati, T., &
Tabatabaei, M. (2015). Encapsulation of Cuminum cyminum essential oils in chit-
osan-caffeic acid nanogel with enhanced antimicrobial activity against
Aspergillusflavus. Industrial Crops and Products, 69, 251–256. http://dx.doi.org/10.
1016/j.indcrop.2015.02.028.