- Email: [email protected]
Foam fractionation as a tool to study the air-water interface structure-function relationship of wheat gluten hydrolysates
Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Foam fractionation as a tool to study the air-water interface structure-function relationship of wheat gluten hydrolysates Arno G.B. Wouters a,∗ , Ine Rombouts a , Nele Schoebrechts a , Ellen Fierens a , Kristof Brijs a , Christophe Blecker b , Jan A. Delcour a a Laboratory of Food Chemistry and Biochemistry and Leuven Food Science and Nutrition Research Center (LFoRCe), KU Leuven, Kasteelpark Arenberg 20, B-3001 Leuven, Belgium b Department of Food Sciences and Formulation, Gembloux Agro-Bio Tech, University of Liège, 5030 Gembloux, Belgium
a r t i c l e
i n f o
Article history: Received 11 October 2016 Received in revised form 2 December 2016 Accepted 20 December 2016 Available online 22 December 2016 Keywords: Wheat gluten Hydrolysate Foam fractionation Hydrophobicity Air-water interfacial properties
a b s t r a c t Enzymatic hydrolysis of wheat gluten protein improves its solubility and produces hydrolysates with foaming properties which may find applications in food products. First, we here investigated whether foam-liquid fractionation can concentrate wheat gluten peptides with foaming properties. Foam and liquid fractions had high and very low foam stability (FS), respectively. In addition, foam fractions were able to decrease surface tension more pronouncedly than un-fractionated samples and liquid fractions, suggesting they are able to arrange themselves more efficiently at an interface. As a second objective, foam fractionation served as a tool to study the structural properties of the peptides, causing these differences in air-water interfacial behavior. Zeta potential and surface hydrophobicity measurements did not fully explain these differences but suggested that hydrophobic interactions at the air-water interface are more important than electrostatic interactions. RP-HPLC showed a large overlap between foam and liquid fractions. However, a small fraction of very hydrophobic peptides with relatively high average molecular mass was clearly enriched in the foam fraction. These peptides were also more concentrated in un-fractionated DH 2 hydrolysates, which had high FS, than in DH 6 hydrolysates, which had low FS. These peptides most likely play a key role in stabilizing the air-water interface. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Commercial wheat gluten is the co-product of the industrial starch isolation. It mainly consists of wheat storage proteins [1]. It is predominantly used in bakery and animal feed applications, but there is also a clear industrial interest in alternative valorization routes [2,3]. One of the main obstacles for a wide application of gluten proteins in food is their low solubility in aqueous media [4]. Enzymatic hydrolysis not only strongly improves their solubility but also induces foaming properties [5]. In a food and beverage context, foams are important in e.g. meringues, beer and whipped dairy products. They consist of a gaseous phase dispersed in a liquid, usually in the form of closely packed air bubbles in an aqueous phase. While foams, which consist of many air-water (A-W) interfaces, are thermodynamically unstable, they can be stabilized by surface-active compounds [6,7]. Because of their amphiphilic nature, proteins and peptides have
∗ Corresponding author. E-mail address: [email protected] (A.G.B. Wouters). http://dx.doi.org/10.1016/j.colsurfb.2016.12.031 0927-7765/© 2016 Elsevier B.V. All rights reserved.
some affinity for A-W interfaces. They can adsorb to interfaces, thereby lowering the surface tension but also sterically preventing gas bubbles to approach and eventually merge with other gas bubbles [6,8]. After adsorption at the interface, proteins tend to interact and form a visco-elastic film which stabilizes the foam. Several authors have discussed the link between foaming and structural properties of wheat gluten hydrolysates. A recurring observation is that hydrolysates with a relatively low degree of hydrolysis (DH), which represents the percentage of peptide bonds cleaved (see below), lead to better foam stability than hydrolysates with a high DH [9–13]. Evidently, a higher DH implies a lower average molecular mass (MM). The importance of a high MM has also been illustrated by the improvement of foaming properties upon transglutaminase treatment [14,15]. Additionally, fractionation with membrane technology has shown that peptide fractions with high average MM have better foaming properties than fractions with lower MM. It is important to note that in different studies [16–19] the peptides in the high MM fractions were also the more hydrophobic ones. The relevance of hydrophobicity for foaming of peptides has also been suggested in a previous study from our group [13]. In this context, it is necessary to keep in mind
296
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
that protein hydrolysates contain a very heterogeneous mixture of peptides, of which it is not always clear to what extent certain peptides contribute to a hydrolysate’s overall functionality. Selectively enriching gluten hydrolysates in peptides which are more efficient in stabilizing A-W interfaces will help to gain insight in the mechanism whereby they stabilize foam. Ways to do so include separation with membranes, as already mentioned above, or by chromatography. However, these techniques are often laborintensive and time-consuming. In addition, they are not selective for peptide fractions with a higher affinity for A-W interfaces. In contrast, foam fractionation processes usually consist of a fairly simple setup [20–22]. They are mostly used industrially to isolate proteins from more complex waste-streams [21,23], such as e.g. from whey [22]. However, they can also be used to separate peptides or proteins with high affinity for an A-W interface from a mixture of peptides or proteins with a lower affinity for such interface. This concept has been used to selectively enrich either ␣-amylase or lysozyme from a mixture of both by foam fractionation at different pH values [24], and to enrich specific bio-active peptides from complex protein hydrolysates [25,26]. Thus far, the potential of foam fractionation to alter the functional properties of protein hydrolysates has not been investigated. In this work, a fairly simple foam fractionation was carried out to investigate its potential for concentrating peptides with improved foaming and interfacial properties. The peptide composition, structural properties and A-W interfacial properties of these fractions were assessed to gain insight in the mechanism of foam stabilization by wheat gluten peptides. 2. Materials and methods 2.1. Materials Commercial wheat gluten was from Tereos Syral (Aalst, Belgium). It contained 82.4% protein (N x 5.7) on dry matter basis when determined using an adaptation of the AOAC Official Method [27] to an EA1108 Elemental Analyzer (Carlo Erba/Thermo Scientific, Waltham, MA, USA). Trypsin (EC 3.4.21.4) from porcine pancreas and pepsin (EC 3.4.23.1) from porcine gastric mucosa were from Sigma-Aldrich (Bornem, Belgium), as were all other chemicals, solvents and reagents. 2.2. Enzymatic hydrolysis A 6.0% (wprotein /v) wheat gluten aqueous dispersion in 0.03% (w/v) NaCl was incubated with trypsin or pepsin at pH-stat conditions. For each enzyme, hydrolysis was performed until DH 2 and DH 6. Tryptic hydrolysis was at 50 ◦ C, pH 8.0 and using an enzyme to substrate ratio of 1:480 (DH 2) or 1:20 (DH 6) on protein mass basis. Peptic hydrolysis was performed at 37 ◦ C, pH 3.5 and using an enzyme to substrate ratio of 1:1200 (DH 2) or 1:300 (DH 6) on protein mass basis. When the desired DH (see Section 2.3) was reached, the pH was adjusted to 6.0 with 0.5 M NaOH and proteolysis was stopped by heating the protein suspension for 15 min at 95 ◦ C. The hydrolysates were then centrifuged (10 min, 12,100g) at room temperature and supernatants were filtered and freeze-dried. Tryptic DH 2 or DH 6 hydrolysates are further referred to as T2 and T6, respectively, and those of peptic DH 2 or DH 6 hydrolysates as P2 and P6, respectively. 2.3. Determination of degree of hydrolysis DH is the percentage of the number of peptide bonds hydrolysed (h) to the total number of peptide bonds per unit weight present in wheat gluten protein (htot ). DH was then calculated from
the amount of base (trypsin) or acid (pepsin) used to keep the pH constant during hydrolysis, using the formula: DH (%) =
h X.Mx .100 = htot ˛.Mp .htot
(1)
X is the consumption (ml) of acid or base needed to keep the pH constant during hydrolysis and Mx the molarity of the acid or base. The term ␣ is a measure for the degree of dissociation of ␣-NH3 + (neutral or alkaline conditions) or ␣-COOH (acidic conditions) groups. Under the conditions used, in tryptic hydrolysis ␣ is 0.89 [28], whereas in peptic hydrolysis it is 0.29 [29]. Mp is the mass of protein used, h are the hydrolysis equivalents [milli-equivalents (meq/g) protein] and htot is the theoretical number of peptide bonds per unit protein. For gluten protein, htot is 8.3 meq/g protein [28,30]. 2.4. Foam fractionation Aliquots (50 ml) of protein solutions [0.15% (wprot /v)] of T2, T6, P2 and P6 were temperature equilibrated in graduated glass cylinders (internal diameter 60.0 mm) in a water bath at 20 ◦ C. A standardized stirring test based on Caessens et al. [31] was performed. The protein solutions were stirred for 70 s using a rotating propeller (outer diameter 45.0 mm, thickness 0.4 mm) at 2000 rpm. After stirring, the propeller was immediately removed and the glass cylinder was sealed with a plastic paraffin film. After 15 min, the foam and liquid phases were freeze-dried separately. This yielded eight samples, with an extra letter in their code (F for foam or L for liquid fraction). 2.5. Analysis of foaming properties Foaming properties of protein solutions [0.05% (wprot /v)] of T2, T6, P2 and P6 and of their respective foam or liquid fractions were determined with the standardized stirring procedure described above (Section 2.4). The foaming capacity (FC) was defined as the foam volume 120 s after the start of stirring. The foam volume was also measured 4, 10, 15, 30, 45 and 60 min after the start of stirring. Foam volumes were calculated based on foam height and cylinder diameter, and expressed in ml. The decrease of foam volume over time was an indication for the foam stability (FS) of a given sample. 2.6. Analysis of zeta potential Protein solutions of T2, P2, T6 and P6 [0.15% (wprotein /v)] in deionized water and of their respective foam and liquid fractions were placed in a disposable capillary zeta cell (Malvern Instruments, Malvern, United Kingdom) to determine zeta potential in a Zetasizer Nano ZS (Malvern) based on laser Doppler microelectrophoresis. 2.7. Analysis of protein surface hydrophobicity The protein surface hydrophobicity of solutions of T2, T6, P2 and P6 and of their respective foam and liquid fractions was determined with 1-anilino-8-naphtalene sulfonic acid (ANS) as fluorescent probe. Samples containing between 0.18 and 0.90 mg protein/ml deionized water were prepared. Aliquots (200.0 l) of these samples were transferred to a 96-well plate, and 10.0 l 8 mM ANS in deionized water was added. The fluorescence intensity of the protein samples was measured with a Synergy Multi-Mode Microplate Reader (BioTek, Winooski, VT, USA), using 390 and 480 nm as excitation and emission wavelengths, respectively. The relative fluorescence intensity was then calculated as the difference in intensity of the protein-ANS mixture and the control sample (ANS in water), divided by the intensity of the control sample. The slope of the plot of relative fluorescence intensity as a function of
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
protein concentration for each sample was a measure for surface hydrophobicity.
2.8. Reversed phase high performance liquid chromatography Protein solutions (1.0 mg/ml in deionized water) of T2, T6, P2 and P6 and of their respective foam and liquid fractions were centrifuged (10 min, 10,000g), filtered (Millex-GP, 0.45 m, polyethersulfone; Millipore), loaded on a Jupiter C18 column (Phenomenex, Torrance, CA, USA) and analyzed at 60 ◦ C using a LC-20AT modular HPLC system (Shimadzu, Kyoto, Japan) with protein elution monitored at 214 nm. The eluents were milli-Q water (solvent A) and acetonitrile (solvent B), both containing 0.1% (v/v) trifluoroacetic acid. Peptides were eluted at 0.3 ml/min with the following gradient conditions: linear from 0 to 56% (v/v) eluent B in 35 min, linear from 56 to 90% (v/v) eluent B in 2 min, isocratic with 90% eluent B for 2 min, linear from 90 to 0% (v/v) eluent B in 2 min. Hydrophobicity distributions of all liquid and foam fractions were compared to those of their respective un-fractionated samples. While all liquid fractions were enriched in peptides eluting before 30 min, all foam fractions were enriched in peptides eluting after 30 min. Therefore, a distinction was made between more hydrophilic (eluting in the first 30 min) and more hydrophobic (eluting thereafter) peptides in the obtained profiles. The relative proportions of hydrophilic and hydrophobic peptides were deduced from the respective areas in the profile and expressed as percentages of the total area. A further separation was then made to divide the hydrophobic (HPHO) peptides in a HPHO-1 (eluting between 30 and 33 min) and a HPHO-2 fraction (eluting between 33 and 42 min).
2.9. Size exclusion high performance liquid chromatography HPHO-1 and HPHO-2 fractions of T2F, T6F, P2F and P6F were collected during RP-HPLC, as described in Section 2.8, after injecting 100 g in 100 l. The solvent (a mixture of milli-Q water and acetonitrile) was removed from these samples in a Rotational Vacuum Concentrator (Martin Christ, Osterode am Harz, Germany), after which the pellet was dissolved in 120 l 0.05 M sodium phosphate buffer (pH 6.8) containing 2.0% sodium dodecyl sulfate. An aliquot (90 l) of each sample was loaded on a BioSep SEC-S2000 column (separation range from 0.2 to 75 kDa, Phenomenex, Torrance, CA, USA) using a Shimadzu LC-2010 integrated system monitoring protein elution at 214 nm. The mobile phase was the above mentioned sodium phosphate buffer as mentioned above. Elution was at 0.5 ml/min and 30 ◦ C. As the protein content of the pellets varied, comparison of the peptide composition for the different samples was difficult. Therefore, all chromatograms were normalized to the same total area. The column was calibrated with MM markers [chicken egg ovalbumin (45 kDa), lysozyme (14.3 kDa), and aprotinin (6.5 kDa)].
2.10. Maximum bubble pressure method In the maximum bubble pressure method (MBP), solutions [0.05% (wprotein /v)] of T2, T6, P2 and P6 and of their respective foam and liquid fractions were used to determine concentrationdependent kinetics of diffusion to and adsorption at the A-W interface in a timescale up to 10 ms. In a BP 100 bubble pressure tensiometer (Krüss, Hamburg, Germany), air bubbles are generated at 20 ◦ C at a constant rate (depending on the desired bubble formation time) through a capillary (diameter 0.200 mm) in the liquid phase. When the bubble radius equals the capillary radius (rcap ), the pressure in the bubble is maximal (Pmax ) and measured. This
297
pressure can be used in the following equation (with P0 the initial hydrostatic pressure) to determine surface tension (␥): =
(Pmax − P0 ) .rcap 2
Surface tension was determined as a function of bubble formation times in a 10 ms to 10 s time frame. 2.11. (Oscillating) pendant drop measurements Solutions [0.05% (wprotein /v)] of T2, T6, P2 and P6 and of their respective foam and liquid fractions were used in a Theta optical tensiometer (Biolin Scientific Attension, Stockholm, Sweden) to create a hanging drop with a fixed volume (8 l). The surface tension at the A-W interface can be calculated by means of drop shape analysis and a Young-Laplace fit. For every drop, the decrease in surface tension was measured over 10 min to assess protein adsorption and rearrangement at the A-W interface. To this end, images were taken at a rate of 0.14 frames per second. After this, a sinusoidal oscillation (50 cycles) was performed at a frequency of 1 Hz with an amplitude set at 1.00 in the OneAttension software (Biolin Scientific Attension), which corresponded to a volume of about 1.0 l. During oscillation, images were recorded at a rate of 7 frames per second. The dilatational elastic modulus E’ of the protein interfacial film could be calculated from the drop shape analysis data during oscillation. 2.12. Statistical analysis MBP measurements were carried out at least in duplicate. All other experiments were performed at least fourfold. All data were analyzed using statistical software JMP Pro 11 (SAS Institute, Cary, NC, USA) with a Tukey multiple comparison test at a significance level ␣ = 0.05. 3. Results and discussion 3.1. Foaming properties Fig. 1 shows the foaming properties of T2, T6, P2 and P6 and of their respective foam and liquid fractions. The first data point (2 min after the start of whipping) in the foam volume vs. time curves indicated that all un-fractionated samples had comparable FC (Fig. 1). However, there were clear differences in FS between the different un-fractionated samples, assessed by the course of foam volume over time, as well as in the final foam volume after 60 min. Both DH 2 samples had better FS than their DH 6 counterparts. This is in line with observations by others that low DH hydrolysates have better FS than high DH hydrolysates [9,32–34,13]. Fractionation only caused minor alteration of the FC. T2F and T2 had comparable FC, whereas T6F, P2F and P6F had slightly higher FC than the un-fractionated samples. All liquid fractions had a FC comparable to that of their respective un-fractionated samples, except for T6L which showed slightly better FC than T6. In contrast, fractionation largely impacted FS. All foam fractions had FS as good or better than their respective un-fractionated samples. In addition, T6F and P6F displayed significantly higher (P < 0.05) foam volume after 60 min than did the un-fractionated samples, while this was not the case for the DH 2 hydrolysates. Liquid fractions had much poorer FS than both the un-fractionated samples and foam fractions. Thus, a relatively simple foam fractionation yields foam and liquid fractions with distinct foaming properties. The rate of drainage of a protein foam is related to the viscosity of the continuous phase [6]. However, viscosities of solutions of the un-fractionated samples did not differ significantly from that of pure water, even when tested in concentrations three times as high as the ones used in this work
298
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
Fig. 1. Foam volume versus time plots of un-fractionated samples of T2, T6, P2 and P6 and their liquid and foam fractions at a concentration of 0.05% wprot /v in deionized water. The first data point of every curve is the foam volume 2 min after the start of whipping and therefore corresponds to the foaming capacity (FC). The data correspond to the mean of four independent measurements ± standard deviations. The codes T and P refer to gluten trypsin and pepsin hydrolysates, the codes 2 and 6 refer to degrees of hydrolysis of 2 and 6.
(data not shown). Therefore, none of the differences in foaming properties observed here were caused by differences in viscosity of the continuous phase. In what follows, structural properties of the peptides in the foam and liquid fractions of all samples will be related to their potential to stabilize interfaces.
3.2. Zeta potential and surface hydrophobicity The capacity of proteins to stabilize foams is generally dictated by their ability to interact and form a cohesive film at the A-W interface [6]. The nature of these interactions can vary, but electrostatic and hydrophobic interactions are assumed to be most common [34]. Fig. 2A shows the zeta potential of T2, T6, P2 and P6 and of their respective foam and liquid fractions. All samples had a negative zeta potential, which indicated an overall negative charge of the peptide mixture in deionized water. T2, P2 and P6 had comparable zeta potential values, whereas that of T6 was much more negative. This difference does not seem to directly relate to differences in foaming properties, given e.g. that T6 and P6 had very similar FS. While the zeta potential varied for foam and liquid fractions, no general trend was observed. Zeta potential readings of T2L and T2 were similar, whereas that of T2F was more negative. In contrast, the zeta potentials of T6F and T6L were closer to zero than that of T6. For P2 and P6, the liquid fractions had more negative zeta potential values than, in that order, the foam fractions and the un-fractionated samples.
Fig. 2B shows the surface hydrophobicity of all un-fractionated and fractionated samples. P2 had the highest surface hydrophobicity, followed by P6, T2 and T6. All foam fractions had higher surface hydrophobicity than the corresponding un-fractionated samples and liquid fractions as a result of the high affinity of hydrophobic peptides for A-W interfaces. Thus, as generally accepted for other protein-stabilized foams, surface hydrophobicity seems important for foaming of gluten hydrolysates. However, not all observations in Section 3.1 can be explained based on surface hydrophobicity readings. After all, DH 2 hydrolysates had better FS than DH 6 hydrolysates, but e.g. P6 had higher surface hydrophobicity than T2. Of course, while the measured surface hydrophobicity results from the entire peptide mixture, certain peptides with a greater tendency to adsorb at A-W interfaces, may dominate the foam stabilizing effect. 3.3. Chromatography Fig. 3A shows RP-HPLC chromatograms of the foam and liquid fractions of T2, T6, P2 and P6. The chromatograms of the un-fractionated samples (not on the graph) were intermediate between those of their foam and liquid fractions. Even though large differences in foaming properties were noted, there was a large overlap between the chromatograms of foam and liquid fractions for each sample. Thus, foam fractionation did not drastically change the overall peptide composition. However, small changes evidently induced significant differences in foaming properties. An
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
299
Fig. 2. Zeta potential (A) and surface hydrophobicity (B) of the T2, T6, P2 and P6 foam (F) and liquid (L) fractions compared to the values for the un-fractionated (U) samples. Codes T2, T6, P2 and P6 as in Fig. 1.
Fig. 3. A. RP-HPLC chromatograms of the liquid and foam fractions of T2, T6, P2 and P6 (from bottom to top). Peptides eluting before 30 min considered as more hydrophilic, whereas those eluting after 30 min as more hydrophobic. B. Relative amounts of hydrophobic peptides in the un-fractionated and the liquid and foam fractions of T2, T6, P2 and P6. Numbers between brackets indicate, within each sample, relative changes in the percentages of hydrophobic peptides from those in each un-fractionated sample (of which the% of hydrophobic peptides is regarded as 1). Codes T2, T6, P2 and P6 as in Fig. 1.
enrichment in hydrophobic peptides (those eluting after 30 min) in the foam fractions was noted. Because the foam fractions had much better FS than the liquid fractions, the hydrophobic peptides are likely to be very efficient at stabilizing A-W interfaces. Fig. 3B shows the percentages of hydrophobic peptides present in each sample. Firstly, the share of hydrophobic peptides was higher in DH 2 hydrolysates (P2: 33%, T2: 25%) than in DH 6 hydrolysates
(P6: 21%, T6: 13%). Secondly, all foam fractions were enriched in these hydrophobic peptides, while all liquid fractions contained less hydrophobic peptides than the un-fractionated samples. Foam fractions of DH 6 hydrolysates contained 1.7 times the amount of hydrophobic peptides of their un-fractionated samples while for foam fractions of DH 2 hydrolysates this was only 1.4 times. This more pronounced enrichment could be related to T6F and P6F hav-
300
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
ing an even higher residual foam volume after 60 min than their un-fractionated samples, which was not the case for T2F and P2F (see Section 3.1). However, these observations still did not explain the overall better FS of DH 2 hydrolysates than of DH 6 hydrolysates, given that T2F and P6F both contained the same level of hydrophobic peptides (36%). In Fig. 4A, the RP-HPLC chromatograms of the liquid fractions were subtracted from those of the corresponding foam fractions. Negative peaks (areas below the gray reference line) correspond to peptides more abundantly present in the liquid fractions, while positive peaks correspond to peptides more abundantly present in the foam fractions. The foams of the DH 6 hydrolysates were especially enriched in peptides eluting between 30 and 33 min (HPHO-1 peptides), whereas the foams of DH 2 hydrolysates were especially enriched in even more hydrophobic peptides eluting between 33 and 42 min (HPHO-2 peptides). Indeed, comparing the distribution of hydrophobic peptides over the HPHO-1 and HPHO-2 fractions (Fig. 4B) allowed concluding that DH 2 hydrolysates had a higher share of HPHO-2 peptides while DH 6 hydrolysates had a higher share of HPHO-1 peptides. So, even if these hydrolysates contained a variety of peptides with varying structure, our results suggest that only a small portion of hydrophobic peptides, more in particular the HPHO-2 peptides, are of utmost importance for foam stabilization. Not only their hydrophobicity but also their size is important for foam stabilization by peptides. Fig. 5 shows the apparent MM distributions of the HPHO-1 and HPHO-2 fractions of T2F, T6F, P2F and P6F when analyzed with SE-HPLC. To enable comparison of peptide compositions, all chromatograms were normalized to the same total area. The HPHO-2 fractions of all samples contained some larger peptides (apparent MM exceeding 45 kDa) which were not or to a much lesser extent present in the HPHO-1 fraction. While hydrophobic interactions obviously contribute to A-W interfacial stabilization, it is very well possible that the higher average MM of these peptides also facilitates interfacial stabilization. However, although certain specific peptides are essential to the foam stabilizing ability of a protein hydrolysate, the mechanism of this process is not yet clear. Evaluation of the rate of diffusion and adsorption to the A-W interface, as well as the E’ of the different samples and their foam and liquid fractions may yield valuable insights. 3.4. Air-water interfacial properties Fig. 6 allows evaluating the surface tension decrease at the earliest stages of protein adsorption (MBP method) of T2, T6, P2 and P6, as well as of their foam and liquid fractions. The initial surface tension of all samples equaled 72 mN/m, indicating that no peptides were adsorbed at the A-W interface. The drop formation time at which surface tension started to decrease, here referred to as the lag time, is indicative for the rate of diffusion of peptides to the interface. Lag time values of all samples ranged between 37 and 53 ms, but were not significantly (P > 0.05) different. However, from that point onwards, differences in the slope of the decrease of surface tension with time, which is indicative for the rate of adsorption to the A-W interface, were observed. While each un-fractionated sample and its respective liquid fraction had a comparable rate of surface tension decrease, the adsorption rate for foam fractions was significantly (P < 0.05) higher. Thus, foam fractions showed substantially more affinity for the interface. This was in line with the fact that the overall surface hydrophobicity of all foam fractions was higher than that of their corresponding un-fractionated samples and liquid fractions (see Section 3.2). However, a higher adsorption rate did not in all cases imply an increased FC (see Section 3.1). For example, T6L had a FC comparable to that of T6F and higher than that of T6, while the adsorption rate of T6L was lower than those of T6 and T6F. It is possible that differences in FC were not
Table 1 Values for equilibrium surface tension of T2, T6, P2 and P6, as well as of their respec√ tive foam and liquid fractions, were estimated at t → ∞ in surface tension vs 1/ t plots with the same data as in Fig. 7. The data correspond to the mean of six independent measurements ± standard deviations. Capital letters indicate significant (␣ = 0.05) differences between the different fractions for a given samples, while lowercase letter indicate significant (␣ = 0.05) differences between different samples in a given phase. T2 Foam Un-fractionated Liquid
T6
47.9 ± 0.5 50.3 ± 0.5 B,bc 52.3 ± 0.8 A,c C,c
P2
49.1 ± 0.3 50.6 ± 0.5 B,ab 55.5 ± 0.8 A,a C,b
P6
47.1 ± 0.6 49.7 ± 0.1 B,c 52.8 ± 0.8 A,c C,d
50.3 ± 0.3 C,a 51.3 ± 0.7 B,a 54.1 ± 0.3 A,b
observed due to the limitations of the foam test in terms of the maximal amount of foam formed during whipping. All things considered, foam fractions were enriched in peptides that were able to decrease surface tension more pronouncedly. Measurements with a pendant drop device show a more pronounced decrease of surface tension over time for the foam fractions of all samples (Fig. 7), not only during the early stages of adsorption (as was also the case for the MBP method described earlier), but also after 10 min. At this point, equilibrium surface ten√ sion had not been reached. To this end, surface tension vs 1/ t plots (plots not shown here) were made of the same data as in Fig. 7, and equilibrium surface tension values were estimated at t → ∞ (Table 1). For any sample, estimated equilibrium surface tension value of its foam fraction was significantly (P < 0.05) higher than that of its un-fractionated sample and of its liquid fraction. This suggests that the peptides enriched in the foam fraction not only had a higher affinity for the A-W interface, they also seemed able to pack more closely or arrange themselves more efficiently at this interface. After subsequent sinusoidal volume oscillations of the equilibrated drops, E’ could be calculated for all samples. E’ is a measure of the elasticity of the protein film and a high E’ is often associated with high foam stability [35]. E’ values of both (un-fractionated) DH 2 hydrolysates were significantly (P < 0.05) higher than E’ of both (un-fractionated) DH 6 hydrolysates. This is in accordance with the observation that FS of DH 2 hydrolysates was in general better than that of DH 6 hydrolysates (see Section 3.1). The relationship between protein film elasticity and FS of enzymatic gluten hydrolysates has also been described elsewhere [13]. All foam fractions and their respective un-fractionated samples had similar E’ values, except for T2F which had significantly higher (P < 0.05) E’ than T2. All liquid fractions but P6L had significantly (P < 0.05) lower E’ than their respective un-fractionated samples. As discussed above, these liquid fractions all had very low FS (See section 3.1). It seems that, while fractionation led to some differences in E’ between the different fractions, these could not explain all observations regarding FS (see Section 3.1). For example, E’ of T2L was equal to that of T6F, but the latter had much better FS than the former. That foam fractionation only had a limited impact on overall peptide composition of wheat gluten hydrolysates (see Section 3.2) seems an important observation in this context. Although E’ and FS are related to some extent, E’ seems to depend on overall peptide composition while slight changes in peptide composition, such as those induced by the foam fractionation process, can strongly affect FS. 4. Conclusions While the choice of enzyme and hydrolysis conditions can to a certain extent determine which peptides are released, the resultant protein hydrolysate always consists of a heterogeneous mixture of peptides that are efficient at stabilizing interfaces and peptides that are not. This work has shown that a relatively simple foam fractionation process can have a drastic impact on the foaming
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
301
Fig. 4. A. Profiles resulting from subtracting the RP-HPLC chromatograms of the liquid fractions from those of the foam fractions for T2, T6, P2 and P6. The hydrophobic (HPHO) peptides are further divided in HPHO-1 and HPHO-2 peptides. B. Relative amounts of HPHO-1 (30–33 min) and HPHO-2 (33–42 min) peptides in the hydrophobic (30–42 min) part of the profiles for all samples. Codes T2, T6, P2 and P6 as in Fig. 1.
Fig. 5. Comparison of the average apparent molecular mass (MM) distribution of the HPHO-1 and HPHO-2 fractions of T2, T6, P2 and P6, after these fractions were collected from RP-HPLC (as described in 2.9). Dots above the curves indicate apparent MM marker proteins of 6.5, 14.3 and 45 kDa. In all cases, the HPHO-2 fraction contained a number of peptides with a higher apparent MM (>45 kDa), which was much less the case for the HPHO-1 fraction. Codes T2, T6, P2 and P6 as in Fig. 1.
302
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
Fig. 6. Surface tension in the early stages of adsorption at the air-water (A-W) interface (from 10 ms to 10 s after bubble formation in a protein solution) of T2, T6, P2 and P6 liquid and foam fractions, compared to their respective un-fractionated samples. Codes T2, T6, P2 and P6 as in Fig. 1.
Fig. 7. Surface tension at the air-water interface after forming a hanging drop with solutions of T2, T6, P2 and P6 and their respective foam and liquid fractions. After 10 min, the drops were sinusoidally oscillated to determine surface dilatational moduli (E’). Codes T2, T6, P2 and P6 as in Fig. 1.
properties of such a complex enzymatic protein hydrolysate. Freeze dried foam fractions had FS as good or better than those of the respective un-fractionated samples, and much better than those of
the liquid fractions, which all had very low FS. When comparing surface-active properties, the foam fractions seemed to have the higher affinity for the A-W interface and caused a larger decrease in
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
surface tension when approaching equilibrium. This suggests that peptides in the foam fraction arrange themselves in a more efficient way at the interface than those in the liquid fraction. However, although there were some notable differences, most foam and liquid fractions did not have remarkably high or low E’, respectively, compared to those of their respective un-fractionated samples. This suggests that, although protein film elasticity is important for gluten hydrolysate foaming properties, other mechanisms are also in play. These will be addressed in follow-up work. Even more important is that foam fractionation is an excellent tool for studying the structural properties of the peptides which cause the differences in air-water interfacial behavior described above. When comparing the peptide composition of each sample before and after fractionation, only small differences were observed, despite large differences in FS. However, a small fraction of hydrophobic peptides with a relatively high apparent MM was enriched in foam fractions. These peptides were also present in higher levels in un-fractionated DH 2 hydrolysates than in DH 6 hydrolysates, which is in accordance with the better FS of the first. These peptides likely play a key role in efficiently stabilizing foam. In contrast to other studies, in which usually the overall properties of the entire hydrolysate were studied, our evidence suggests that it is rather the presence of certain specific peptides that determines the foam stabilizing ability of the entire hydrolysate. Acknowledgements I. Rombouts thanks the Research Foundation − Flanders (FWO, Brussels, Belgium) for financial support. K. Brijs acknowledges the Industrial Research Fund (KU Leuven, Leuven, Belgium) for a position as Industrial Research Manager. J.A. Delcour holds the W.K. Kellogg Chair in Cereal Science and Nutrition at KU Leuven. This work is part of the Methusalem program “Food for the Future” at KU Leuven. References [1] A. Van Der Borght, H. Goesaert, W.S. Veraverbeke, J.A. Delcour, Fractionation of wheat and wheat flour into starch and gluten: overview of the main processes and the factors involved, J. Cereal Sci. 41 (2005) 221–237. [2] L. Day, M.A. Augustin, I.L. Batey, C.W. Wrigley, Wheat gluten uses and industry needs, Trends Food Sci. Technol. 17 (2006) 82–90. [3] W.S. Veraverbeke, J.A. Delcour, Wheat protein composition and properties of wheat glutenin in relation to breadmaking functionality, Crit. Rev. Food Sci. Nutr. 42 (2002) 179–208. [4] J.A. Delcour, I.J. Joye, B. Pareyt, E. Wilderjans, K. Brijs, B. Lagrain, Wheat gluten functionality as a quality determinant in cereal-based food products, Annu. Rev. Food Sci. Technol. 3 (2012) 469–492. [5] J. Adler-Nissen, Enzymatic hydrolysis of proteins for increased solubility, J. Agric. Food Chem. 24 (1976) 1090–1093. [6] S. Damodaran, Protein stabilization of emulsions and foams, J. Food Sci. 70 (2005) R54–R66. [7] B.S. Murray, Stabilization of bubbles and foams, Curr. Opin. Colloid Interface Sci. 12 (2007) 232–241. [8] T.N. Hunter, R.J. Pugh, G.V. Franks, G.J. Jameson, The role of particles in stabilising foams and emulsions, Adv. Colloid Interface Sci. 137 (2008) 57–81. [9] S.R. Drago, R.J. González, Foaming properties of enzymatically hydrolysed wheat gluten, Innov. Food Sci. Emerg. Technol. 1 (2000) 269–273. [10] X.Z. Kong, H.M. Zhou, H.F. Qian, Enzymatic preparation and functional properties of wheat gluten hydrolysates, Food Chem. 101 (2007) 615–620.
303
[11] E. Linares, C. Larre, M. Lemeste, Y. Popineau, Emulsifying and foaming properties of gluten hydrolysates with an increasing degree of hydrolysis: role of soluble and insoluble fractions, Cereal Chem. 77 (2000) 414–420. [12] B. Mimouni, J. Raymond, A.M. Merle-Desnoyers, J.L. Azanza, A. Ducastaing, Combined acid deamidation and enzymic hydrolysis for improvement of the functional properties of wheat gluten, J. Cereal Sci. 20 (1994) 153–165. [13] A.G.B. Wouters, I. Rombouts, M. Legein, E. Fierens, K. Brijs, C. Blecker, J.A. Delcour, Air–water interfacial properties of enzymatic wheat gluten hydrolyzates determine their foaming behavior, Food Hydrocolloid 55 (2016) 155–162. [14] K.K. Agyare, K. Addo, Y.L. Xiong, Emulsifying and foaming properties of transglutaminase-treated wheat gluten hydrolysate as influenced by ph, temperature and salt, Food Hydrocolloid 23 (2009) 72–81. [15] E.E. Babiker, N. Fujisawa, N. Matsudomi, A. Kato, Improvement in the functional properties of gluten by protease digestion or acid hydrolysis followed by microbial transglutaminase treatment, J. Agric. Food Chem. 44 (1996) 3746–3750. [16] S. Berot, Y. Popineau, J.P. Compoint, C. Blassel, B. Chaufer, Ultrafiltration to fractionate wheat polypeptides, J. Chromatogr. B 753 (2001) 29–35. [17] Y. Popineau, B. Huchet, C. Larre, S. Berot, Foaming and emulsifying properties of fractions of gluten peptides obtained by limited enzymatic hydrolysis and ultrafiltration, J. Cereal Sci. 35 (2002) 327–335. [18] J.S. Wang, M.M. Zhao, Y. Bao, T. Hong, C.M. Rosella, Preparation and characterization of modified wheat gluten by enzymatic hydrolysis-ultrafiltration, J. Food Biochem. 32 (2008) 316–334. [19] J.S. Wang, M.M. Zhao, X.Q. Yang, Y.M. Jiang, Improvement on functional properties of wheat gluten by enzymatic hydrolysis and ultrafiltration, J. Cereal Sci. 44 (2006) 93–100. [20] I. Barackov, A. Mause, S. Kapoor, R. Winter, G. Schembecker, B. Burghoff, Investigation of structural changes of -casein and lysozyme at the gas–liquid interface during foam fractionation, J. Biotechnol. 161 (2012) 138–146. [21] B. Burghoff, Foam fractionation applications, J. Biotechnol. 161 (2012) 126–137. [22] A.P. Shea, C.L. Crofcheck, F.A. Payne, Y.L. Xiong, Foam fractionation of ␣-lactalbumin and -lactoglobulin from a whey solution, Asia-Pac. J. Chem. Eng. 4 (2009) 191–203. [23] C.E. Lockwood, P.M. Bummer, M. Jay, Purification of proteins using foam fractionation, Pharm. Res. 14 (1997) 1511–1515. [24] T. Nakabayashi, Y. Takakusagi, K. Iwabata, K. Sakaguchi, Foam fractionation of protein: correlation of protein adsorption onto bubbles with a ph-induced conformational transition, Anal. Biochem. 419 (2011) 173–179. [25] P. Dhordain, M. Bigan, M. Vanhoute, C. Pierlot, J.M. Aubry, P. Dhulster, D. Guillochon, R. Froidevaux, Optimization of peptide separation from complex peptide mixture in a foaming-draining system, Sep. Sci. Technol. 47 (2012) 654–662. [26] M. Vanhoute, R. Froidevaux, C. Pierlot, F. Krier, J.M. Aubry, D. Guillochon, Advancement of foam separation of bioactive peptides using an aeration column with a bubbling-draining method, Sep. Purif. Technol. 63 (2008) 460–465. [27] AOAC, Official Methods of Analysis. Method 990.03, Association of Official Analytical Chemists, Washington, DC, USA, 1995. [28] J. Adler-Nissen, Enzymic Hydrolysis of Food Proteins, Elsevier Applied Science Publishers, New York, USA, 1985, p. [29] P. Diermayr, L. Dehne, Controlled enzymatic hydrolysis of proteins at low ph values.1. Experiments with bovine serum-albumin, Z. Lebensm. Unters. Forsch. 190 (1990) 516–520. [30] P.M. Nielsen, D. Petersen, C. Dambmann, Improved method for determining food protein degree of hydrolysis, J. Food Sci. 66 (2001) 642–646. [31] P.W.J.R. Caessens, H. Gruppen, S. Visser, G.A. van Aken, A.G.J. Voragen, Plasmin hydrolysis of beta-casein: foaming and emulsifying properties of the fractionated hydrolysate, J. Agric. Food Chem. 45 (1997) 2935–2941. [32] I. Celus, K. Brijs, J.A. Delcour, Enzymatic hydrolysis of brewers’ spent grain proteins and technofunctional properties of the resulting hydrolysates, J. Agric. Food Chem. 55 (2007) 8703–8710. [33] X. Guan, H.Y. Yao, Z.X. Chen, L.A. Shan, M.D. Zhang, Some functional properties of oat bran protein concentrate modified by trypsin, Food Chem. 101 (2007) 163–170. [34] A.G.B. Wouters, I. Rombouts, E. Fierens, K. Brijs, J.A. Delcour, Relevance of the functional properties of enzymatic plant protein hydrolysates in food systems, Comp. Rev. Food Sci. Food Saf. 15 (2016) 786–800. [35] B.S. Murray, Rheological properties of protein films, Curr. Opin. in Colloid & Interface Sci. 16 (2011) 27–35.
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Foam fractionation as a tool to study the air-water interface structure-function relationship of wheat gluten hydrolysates Arno G.B. Wouters a,∗ , Ine Rombouts a , Nele Schoebrechts a , Ellen Fierens a , Kristof Brijs a , Christophe Blecker b , Jan A. Delcour a a Laboratory of Food Chemistry and Biochemistry and Leuven Food Science and Nutrition Research Center (LFoRCe), KU Leuven, Kasteelpark Arenberg 20, B-3001 Leuven, Belgium b Department of Food Sciences and Formulation, Gembloux Agro-Bio Tech, University of Liège, 5030 Gembloux, Belgium
a r t i c l e
i n f o
Article history: Received 11 October 2016 Received in revised form 2 December 2016 Accepted 20 December 2016 Available online 22 December 2016 Keywords: Wheat gluten Hydrolysate Foam fractionation Hydrophobicity Air-water interfacial properties
a b s t r a c t Enzymatic hydrolysis of wheat gluten protein improves its solubility and produces hydrolysates with foaming properties which may find applications in food products. First, we here investigated whether foam-liquid fractionation can concentrate wheat gluten peptides with foaming properties. Foam and liquid fractions had high and very low foam stability (FS), respectively. In addition, foam fractions were able to decrease surface tension more pronouncedly than un-fractionated samples and liquid fractions, suggesting they are able to arrange themselves more efficiently at an interface. As a second objective, foam fractionation served as a tool to study the structural properties of the peptides, causing these differences in air-water interfacial behavior. Zeta potential and surface hydrophobicity measurements did not fully explain these differences but suggested that hydrophobic interactions at the air-water interface are more important than electrostatic interactions. RP-HPLC showed a large overlap between foam and liquid fractions. However, a small fraction of very hydrophobic peptides with relatively high average molecular mass was clearly enriched in the foam fraction. These peptides were also more concentrated in un-fractionated DH 2 hydrolysates, which had high FS, than in DH 6 hydrolysates, which had low FS. These peptides most likely play a key role in stabilizing the air-water interface. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Commercial wheat gluten is the co-product of the industrial starch isolation. It mainly consists of wheat storage proteins [1]. It is predominantly used in bakery and animal feed applications, but there is also a clear industrial interest in alternative valorization routes [2,3]. One of the main obstacles for a wide application of gluten proteins in food is their low solubility in aqueous media [4]. Enzymatic hydrolysis not only strongly improves their solubility but also induces foaming properties [5]. In a food and beverage context, foams are important in e.g. meringues, beer and whipped dairy products. They consist of a gaseous phase dispersed in a liquid, usually in the form of closely packed air bubbles in an aqueous phase. While foams, which consist of many air-water (A-W) interfaces, are thermodynamically unstable, they can be stabilized by surface-active compounds [6,7]. Because of their amphiphilic nature, proteins and peptides have
∗ Corresponding author. E-mail address: [email protected] (A.G.B. Wouters). http://dx.doi.org/10.1016/j.colsurfb.2016.12.031 0927-7765/© 2016 Elsevier B.V. All rights reserved.
some affinity for A-W interfaces. They can adsorb to interfaces, thereby lowering the surface tension but also sterically preventing gas bubbles to approach and eventually merge with other gas bubbles [6,8]. After adsorption at the interface, proteins tend to interact and form a visco-elastic film which stabilizes the foam. Several authors have discussed the link between foaming and structural properties of wheat gluten hydrolysates. A recurring observation is that hydrolysates with a relatively low degree of hydrolysis (DH), which represents the percentage of peptide bonds cleaved (see below), lead to better foam stability than hydrolysates with a high DH [9–13]. Evidently, a higher DH implies a lower average molecular mass (MM). The importance of a high MM has also been illustrated by the improvement of foaming properties upon transglutaminase treatment [14,15]. Additionally, fractionation with membrane technology has shown that peptide fractions with high average MM have better foaming properties than fractions with lower MM. It is important to note that in different studies [16–19] the peptides in the high MM fractions were also the more hydrophobic ones. The relevance of hydrophobicity for foaming of peptides has also been suggested in a previous study from our group [13]. In this context, it is necessary to keep in mind
296
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
that protein hydrolysates contain a very heterogeneous mixture of peptides, of which it is not always clear to what extent certain peptides contribute to a hydrolysate’s overall functionality. Selectively enriching gluten hydrolysates in peptides which are more efficient in stabilizing A-W interfaces will help to gain insight in the mechanism whereby they stabilize foam. Ways to do so include separation with membranes, as already mentioned above, or by chromatography. However, these techniques are often laborintensive and time-consuming. In addition, they are not selective for peptide fractions with a higher affinity for A-W interfaces. In contrast, foam fractionation processes usually consist of a fairly simple setup [20–22]. They are mostly used industrially to isolate proteins from more complex waste-streams [21,23], such as e.g. from whey [22]. However, they can also be used to separate peptides or proteins with high affinity for an A-W interface from a mixture of peptides or proteins with a lower affinity for such interface. This concept has been used to selectively enrich either ␣-amylase or lysozyme from a mixture of both by foam fractionation at different pH values [24], and to enrich specific bio-active peptides from complex protein hydrolysates [25,26]. Thus far, the potential of foam fractionation to alter the functional properties of protein hydrolysates has not been investigated. In this work, a fairly simple foam fractionation was carried out to investigate its potential for concentrating peptides with improved foaming and interfacial properties. The peptide composition, structural properties and A-W interfacial properties of these fractions were assessed to gain insight in the mechanism of foam stabilization by wheat gluten peptides. 2. Materials and methods 2.1. Materials Commercial wheat gluten was from Tereos Syral (Aalst, Belgium). It contained 82.4% protein (N x 5.7) on dry matter basis when determined using an adaptation of the AOAC Official Method [27] to an EA1108 Elemental Analyzer (Carlo Erba/Thermo Scientific, Waltham, MA, USA). Trypsin (EC 3.4.21.4) from porcine pancreas and pepsin (EC 3.4.23.1) from porcine gastric mucosa were from Sigma-Aldrich (Bornem, Belgium), as were all other chemicals, solvents and reagents. 2.2. Enzymatic hydrolysis A 6.0% (wprotein /v) wheat gluten aqueous dispersion in 0.03% (w/v) NaCl was incubated with trypsin or pepsin at pH-stat conditions. For each enzyme, hydrolysis was performed until DH 2 and DH 6. Tryptic hydrolysis was at 50 ◦ C, pH 8.0 and using an enzyme to substrate ratio of 1:480 (DH 2) or 1:20 (DH 6) on protein mass basis. Peptic hydrolysis was performed at 37 ◦ C, pH 3.5 and using an enzyme to substrate ratio of 1:1200 (DH 2) or 1:300 (DH 6) on protein mass basis. When the desired DH (see Section 2.3) was reached, the pH was adjusted to 6.0 with 0.5 M NaOH and proteolysis was stopped by heating the protein suspension for 15 min at 95 ◦ C. The hydrolysates were then centrifuged (10 min, 12,100g) at room temperature and supernatants were filtered and freeze-dried. Tryptic DH 2 or DH 6 hydrolysates are further referred to as T2 and T6, respectively, and those of peptic DH 2 or DH 6 hydrolysates as P2 and P6, respectively. 2.3. Determination of degree of hydrolysis DH is the percentage of the number of peptide bonds hydrolysed (h) to the total number of peptide bonds per unit weight present in wheat gluten protein (htot ). DH was then calculated from
the amount of base (trypsin) or acid (pepsin) used to keep the pH constant during hydrolysis, using the formula: DH (%) =
h X.Mx .100 = htot ˛.Mp .htot
(1)
X is the consumption (ml) of acid or base needed to keep the pH constant during hydrolysis and Mx the molarity of the acid or base. The term ␣ is a measure for the degree of dissociation of ␣-NH3 + (neutral or alkaline conditions) or ␣-COOH (acidic conditions) groups. Under the conditions used, in tryptic hydrolysis ␣ is 0.89 [28], whereas in peptic hydrolysis it is 0.29 [29]. Mp is the mass of protein used, h are the hydrolysis equivalents [milli-equivalents (meq/g) protein] and htot is the theoretical number of peptide bonds per unit protein. For gluten protein, htot is 8.3 meq/g protein [28,30]. 2.4. Foam fractionation Aliquots (50 ml) of protein solutions [0.15% (wprot /v)] of T2, T6, P2 and P6 were temperature equilibrated in graduated glass cylinders (internal diameter 60.0 mm) in a water bath at 20 ◦ C. A standardized stirring test based on Caessens et al. [31] was performed. The protein solutions were stirred for 70 s using a rotating propeller (outer diameter 45.0 mm, thickness 0.4 mm) at 2000 rpm. After stirring, the propeller was immediately removed and the glass cylinder was sealed with a plastic paraffin film. After 15 min, the foam and liquid phases were freeze-dried separately. This yielded eight samples, with an extra letter in their code (F for foam or L for liquid fraction). 2.5. Analysis of foaming properties Foaming properties of protein solutions [0.05% (wprot /v)] of T2, T6, P2 and P6 and of their respective foam or liquid fractions were determined with the standardized stirring procedure described above (Section 2.4). The foaming capacity (FC) was defined as the foam volume 120 s after the start of stirring. The foam volume was also measured 4, 10, 15, 30, 45 and 60 min after the start of stirring. Foam volumes were calculated based on foam height and cylinder diameter, and expressed in ml. The decrease of foam volume over time was an indication for the foam stability (FS) of a given sample. 2.6. Analysis of zeta potential Protein solutions of T2, P2, T6 and P6 [0.15% (wprotein /v)] in deionized water and of their respective foam and liquid fractions were placed in a disposable capillary zeta cell (Malvern Instruments, Malvern, United Kingdom) to determine zeta potential in a Zetasizer Nano ZS (Malvern) based on laser Doppler microelectrophoresis. 2.7. Analysis of protein surface hydrophobicity The protein surface hydrophobicity of solutions of T2, T6, P2 and P6 and of their respective foam and liquid fractions was determined with 1-anilino-8-naphtalene sulfonic acid (ANS) as fluorescent probe. Samples containing between 0.18 and 0.90 mg protein/ml deionized water were prepared. Aliquots (200.0 l) of these samples were transferred to a 96-well plate, and 10.0 l 8 mM ANS in deionized water was added. The fluorescence intensity of the protein samples was measured with a Synergy Multi-Mode Microplate Reader (BioTek, Winooski, VT, USA), using 390 and 480 nm as excitation and emission wavelengths, respectively. The relative fluorescence intensity was then calculated as the difference in intensity of the protein-ANS mixture and the control sample (ANS in water), divided by the intensity of the control sample. The slope of the plot of relative fluorescence intensity as a function of
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
protein concentration for each sample was a measure for surface hydrophobicity.
2.8. Reversed phase high performance liquid chromatography Protein solutions (1.0 mg/ml in deionized water) of T2, T6, P2 and P6 and of their respective foam and liquid fractions were centrifuged (10 min, 10,000g), filtered (Millex-GP, 0.45 m, polyethersulfone; Millipore), loaded on a Jupiter C18 column (Phenomenex, Torrance, CA, USA) and analyzed at 60 ◦ C using a LC-20AT modular HPLC system (Shimadzu, Kyoto, Japan) with protein elution monitored at 214 nm. The eluents were milli-Q water (solvent A) and acetonitrile (solvent B), both containing 0.1% (v/v) trifluoroacetic acid. Peptides were eluted at 0.3 ml/min with the following gradient conditions: linear from 0 to 56% (v/v) eluent B in 35 min, linear from 56 to 90% (v/v) eluent B in 2 min, isocratic with 90% eluent B for 2 min, linear from 90 to 0% (v/v) eluent B in 2 min. Hydrophobicity distributions of all liquid and foam fractions were compared to those of their respective un-fractionated samples. While all liquid fractions were enriched in peptides eluting before 30 min, all foam fractions were enriched in peptides eluting after 30 min. Therefore, a distinction was made between more hydrophilic (eluting in the first 30 min) and more hydrophobic (eluting thereafter) peptides in the obtained profiles. The relative proportions of hydrophilic and hydrophobic peptides were deduced from the respective areas in the profile and expressed as percentages of the total area. A further separation was then made to divide the hydrophobic (HPHO) peptides in a HPHO-1 (eluting between 30 and 33 min) and a HPHO-2 fraction (eluting between 33 and 42 min).
2.9. Size exclusion high performance liquid chromatography HPHO-1 and HPHO-2 fractions of T2F, T6F, P2F and P6F were collected during RP-HPLC, as described in Section 2.8, after injecting 100 g in 100 l. The solvent (a mixture of milli-Q water and acetonitrile) was removed from these samples in a Rotational Vacuum Concentrator (Martin Christ, Osterode am Harz, Germany), after which the pellet was dissolved in 120 l 0.05 M sodium phosphate buffer (pH 6.8) containing 2.0% sodium dodecyl sulfate. An aliquot (90 l) of each sample was loaded on a BioSep SEC-S2000 column (separation range from 0.2 to 75 kDa, Phenomenex, Torrance, CA, USA) using a Shimadzu LC-2010 integrated system monitoring protein elution at 214 nm. The mobile phase was the above mentioned sodium phosphate buffer as mentioned above. Elution was at 0.5 ml/min and 30 ◦ C. As the protein content of the pellets varied, comparison of the peptide composition for the different samples was difficult. Therefore, all chromatograms were normalized to the same total area. The column was calibrated with MM markers [chicken egg ovalbumin (45 kDa), lysozyme (14.3 kDa), and aprotinin (6.5 kDa)].
2.10. Maximum bubble pressure method In the maximum bubble pressure method (MBP), solutions [0.05% (wprotein /v)] of T2, T6, P2 and P6 and of their respective foam and liquid fractions were used to determine concentrationdependent kinetics of diffusion to and adsorption at the A-W interface in a timescale up to 10 ms. In a BP 100 bubble pressure tensiometer (Krüss, Hamburg, Germany), air bubbles are generated at 20 ◦ C at a constant rate (depending on the desired bubble formation time) through a capillary (diameter 0.200 mm) in the liquid phase. When the bubble radius equals the capillary radius (rcap ), the pressure in the bubble is maximal (Pmax ) and measured. This
297
pressure can be used in the following equation (with P0 the initial hydrostatic pressure) to determine surface tension (␥): =
(Pmax − P0 ) .rcap 2
Surface tension was determined as a function of bubble formation times in a 10 ms to 10 s time frame. 2.11. (Oscillating) pendant drop measurements Solutions [0.05% (wprotein /v)] of T2, T6, P2 and P6 and of their respective foam and liquid fractions were used in a Theta optical tensiometer (Biolin Scientific Attension, Stockholm, Sweden) to create a hanging drop with a fixed volume (8 l). The surface tension at the A-W interface can be calculated by means of drop shape analysis and a Young-Laplace fit. For every drop, the decrease in surface tension was measured over 10 min to assess protein adsorption and rearrangement at the A-W interface. To this end, images were taken at a rate of 0.14 frames per second. After this, a sinusoidal oscillation (50 cycles) was performed at a frequency of 1 Hz with an amplitude set at 1.00 in the OneAttension software (Biolin Scientific Attension), which corresponded to a volume of about 1.0 l. During oscillation, images were recorded at a rate of 7 frames per second. The dilatational elastic modulus E’ of the protein interfacial film could be calculated from the drop shape analysis data during oscillation. 2.12. Statistical analysis MBP measurements were carried out at least in duplicate. All other experiments were performed at least fourfold. All data were analyzed using statistical software JMP Pro 11 (SAS Institute, Cary, NC, USA) with a Tukey multiple comparison test at a significance level ␣ = 0.05. 3. Results and discussion 3.1. Foaming properties Fig. 1 shows the foaming properties of T2, T6, P2 and P6 and of their respective foam and liquid fractions. The first data point (2 min after the start of whipping) in the foam volume vs. time curves indicated that all un-fractionated samples had comparable FC (Fig. 1). However, there were clear differences in FS between the different un-fractionated samples, assessed by the course of foam volume over time, as well as in the final foam volume after 60 min. Both DH 2 samples had better FS than their DH 6 counterparts. This is in line with observations by others that low DH hydrolysates have better FS than high DH hydrolysates [9,32–34,13]. Fractionation only caused minor alteration of the FC. T2F and T2 had comparable FC, whereas T6F, P2F and P6F had slightly higher FC than the un-fractionated samples. All liquid fractions had a FC comparable to that of their respective un-fractionated samples, except for T6L which showed slightly better FC than T6. In contrast, fractionation largely impacted FS. All foam fractions had FS as good or better than their respective un-fractionated samples. In addition, T6F and P6F displayed significantly higher (P < 0.05) foam volume after 60 min than did the un-fractionated samples, while this was not the case for the DH 2 hydrolysates. Liquid fractions had much poorer FS than both the un-fractionated samples and foam fractions. Thus, a relatively simple foam fractionation yields foam and liquid fractions with distinct foaming properties. The rate of drainage of a protein foam is related to the viscosity of the continuous phase [6]. However, viscosities of solutions of the un-fractionated samples did not differ significantly from that of pure water, even when tested in concentrations three times as high as the ones used in this work
298
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
Fig. 1. Foam volume versus time plots of un-fractionated samples of T2, T6, P2 and P6 and their liquid and foam fractions at a concentration of 0.05% wprot /v in deionized water. The first data point of every curve is the foam volume 2 min after the start of whipping and therefore corresponds to the foaming capacity (FC). The data correspond to the mean of four independent measurements ± standard deviations. The codes T and P refer to gluten trypsin and pepsin hydrolysates, the codes 2 and 6 refer to degrees of hydrolysis of 2 and 6.
(data not shown). Therefore, none of the differences in foaming properties observed here were caused by differences in viscosity of the continuous phase. In what follows, structural properties of the peptides in the foam and liquid fractions of all samples will be related to their potential to stabilize interfaces.
3.2. Zeta potential and surface hydrophobicity The capacity of proteins to stabilize foams is generally dictated by their ability to interact and form a cohesive film at the A-W interface [6]. The nature of these interactions can vary, but electrostatic and hydrophobic interactions are assumed to be most common [34]. Fig. 2A shows the zeta potential of T2, T6, P2 and P6 and of their respective foam and liquid fractions. All samples had a negative zeta potential, which indicated an overall negative charge of the peptide mixture in deionized water. T2, P2 and P6 had comparable zeta potential values, whereas that of T6 was much more negative. This difference does not seem to directly relate to differences in foaming properties, given e.g. that T6 and P6 had very similar FS. While the zeta potential varied for foam and liquid fractions, no general trend was observed. Zeta potential readings of T2L and T2 were similar, whereas that of T2F was more negative. In contrast, the zeta potentials of T6F and T6L were closer to zero than that of T6. For P2 and P6, the liquid fractions had more negative zeta potential values than, in that order, the foam fractions and the un-fractionated samples.
Fig. 2B shows the surface hydrophobicity of all un-fractionated and fractionated samples. P2 had the highest surface hydrophobicity, followed by P6, T2 and T6. All foam fractions had higher surface hydrophobicity than the corresponding un-fractionated samples and liquid fractions as a result of the high affinity of hydrophobic peptides for A-W interfaces. Thus, as generally accepted for other protein-stabilized foams, surface hydrophobicity seems important for foaming of gluten hydrolysates. However, not all observations in Section 3.1 can be explained based on surface hydrophobicity readings. After all, DH 2 hydrolysates had better FS than DH 6 hydrolysates, but e.g. P6 had higher surface hydrophobicity than T2. Of course, while the measured surface hydrophobicity results from the entire peptide mixture, certain peptides with a greater tendency to adsorb at A-W interfaces, may dominate the foam stabilizing effect. 3.3. Chromatography Fig. 3A shows RP-HPLC chromatograms of the foam and liquid fractions of T2, T6, P2 and P6. The chromatograms of the un-fractionated samples (not on the graph) were intermediate between those of their foam and liquid fractions. Even though large differences in foaming properties were noted, there was a large overlap between the chromatograms of foam and liquid fractions for each sample. Thus, foam fractionation did not drastically change the overall peptide composition. However, small changes evidently induced significant differences in foaming properties. An
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
299
Fig. 2. Zeta potential (A) and surface hydrophobicity (B) of the T2, T6, P2 and P6 foam (F) and liquid (L) fractions compared to the values for the un-fractionated (U) samples. Codes T2, T6, P2 and P6 as in Fig. 1.
Fig. 3. A. RP-HPLC chromatograms of the liquid and foam fractions of T2, T6, P2 and P6 (from bottom to top). Peptides eluting before 30 min considered as more hydrophilic, whereas those eluting after 30 min as more hydrophobic. B. Relative amounts of hydrophobic peptides in the un-fractionated and the liquid and foam fractions of T2, T6, P2 and P6. Numbers between brackets indicate, within each sample, relative changes in the percentages of hydrophobic peptides from those in each un-fractionated sample (of which the% of hydrophobic peptides is regarded as 1). Codes T2, T6, P2 and P6 as in Fig. 1.
enrichment in hydrophobic peptides (those eluting after 30 min) in the foam fractions was noted. Because the foam fractions had much better FS than the liquid fractions, the hydrophobic peptides are likely to be very efficient at stabilizing A-W interfaces. Fig. 3B shows the percentages of hydrophobic peptides present in each sample. Firstly, the share of hydrophobic peptides was higher in DH 2 hydrolysates (P2: 33%, T2: 25%) than in DH 6 hydrolysates
(P6: 21%, T6: 13%). Secondly, all foam fractions were enriched in these hydrophobic peptides, while all liquid fractions contained less hydrophobic peptides than the un-fractionated samples. Foam fractions of DH 6 hydrolysates contained 1.7 times the amount of hydrophobic peptides of their un-fractionated samples while for foam fractions of DH 2 hydrolysates this was only 1.4 times. This more pronounced enrichment could be related to T6F and P6F hav-
300
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
ing an even higher residual foam volume after 60 min than their un-fractionated samples, which was not the case for T2F and P2F (see Section 3.1). However, these observations still did not explain the overall better FS of DH 2 hydrolysates than of DH 6 hydrolysates, given that T2F and P6F both contained the same level of hydrophobic peptides (36%). In Fig. 4A, the RP-HPLC chromatograms of the liquid fractions were subtracted from those of the corresponding foam fractions. Negative peaks (areas below the gray reference line) correspond to peptides more abundantly present in the liquid fractions, while positive peaks correspond to peptides more abundantly present in the foam fractions. The foams of the DH 6 hydrolysates were especially enriched in peptides eluting between 30 and 33 min (HPHO-1 peptides), whereas the foams of DH 2 hydrolysates were especially enriched in even more hydrophobic peptides eluting between 33 and 42 min (HPHO-2 peptides). Indeed, comparing the distribution of hydrophobic peptides over the HPHO-1 and HPHO-2 fractions (Fig. 4B) allowed concluding that DH 2 hydrolysates had a higher share of HPHO-2 peptides while DH 6 hydrolysates had a higher share of HPHO-1 peptides. So, even if these hydrolysates contained a variety of peptides with varying structure, our results suggest that only a small portion of hydrophobic peptides, more in particular the HPHO-2 peptides, are of utmost importance for foam stabilization. Not only their hydrophobicity but also their size is important for foam stabilization by peptides. Fig. 5 shows the apparent MM distributions of the HPHO-1 and HPHO-2 fractions of T2F, T6F, P2F and P6F when analyzed with SE-HPLC. To enable comparison of peptide compositions, all chromatograms were normalized to the same total area. The HPHO-2 fractions of all samples contained some larger peptides (apparent MM exceeding 45 kDa) which were not or to a much lesser extent present in the HPHO-1 fraction. While hydrophobic interactions obviously contribute to A-W interfacial stabilization, it is very well possible that the higher average MM of these peptides also facilitates interfacial stabilization. However, although certain specific peptides are essential to the foam stabilizing ability of a protein hydrolysate, the mechanism of this process is not yet clear. Evaluation of the rate of diffusion and adsorption to the A-W interface, as well as the E’ of the different samples and their foam and liquid fractions may yield valuable insights. 3.4. Air-water interfacial properties Fig. 6 allows evaluating the surface tension decrease at the earliest stages of protein adsorption (MBP method) of T2, T6, P2 and P6, as well as of their foam and liquid fractions. The initial surface tension of all samples equaled 72 mN/m, indicating that no peptides were adsorbed at the A-W interface. The drop formation time at which surface tension started to decrease, here referred to as the lag time, is indicative for the rate of diffusion of peptides to the interface. Lag time values of all samples ranged between 37 and 53 ms, but were not significantly (P > 0.05) different. However, from that point onwards, differences in the slope of the decrease of surface tension with time, which is indicative for the rate of adsorption to the A-W interface, were observed. While each un-fractionated sample and its respective liquid fraction had a comparable rate of surface tension decrease, the adsorption rate for foam fractions was significantly (P < 0.05) higher. Thus, foam fractions showed substantially more affinity for the interface. This was in line with the fact that the overall surface hydrophobicity of all foam fractions was higher than that of their corresponding un-fractionated samples and liquid fractions (see Section 3.2). However, a higher adsorption rate did not in all cases imply an increased FC (see Section 3.1). For example, T6L had a FC comparable to that of T6F and higher than that of T6, while the adsorption rate of T6L was lower than those of T6 and T6F. It is possible that differences in FC were not
Table 1 Values for equilibrium surface tension of T2, T6, P2 and P6, as well as of their respec√ tive foam and liquid fractions, were estimated at t → ∞ in surface tension vs 1/ t plots with the same data as in Fig. 7. The data correspond to the mean of six independent measurements ± standard deviations. Capital letters indicate significant (␣ = 0.05) differences between the different fractions for a given samples, while lowercase letter indicate significant (␣ = 0.05) differences between different samples in a given phase. T2 Foam Un-fractionated Liquid
T6
47.9 ± 0.5 50.3 ± 0.5 B,bc 52.3 ± 0.8 A,c C,c
P2
49.1 ± 0.3 50.6 ± 0.5 B,ab 55.5 ± 0.8 A,a C,b
P6
47.1 ± 0.6 49.7 ± 0.1 B,c 52.8 ± 0.8 A,c C,d
50.3 ± 0.3 C,a 51.3 ± 0.7 B,a 54.1 ± 0.3 A,b
observed due to the limitations of the foam test in terms of the maximal amount of foam formed during whipping. All things considered, foam fractions were enriched in peptides that were able to decrease surface tension more pronouncedly. Measurements with a pendant drop device show a more pronounced decrease of surface tension over time for the foam fractions of all samples (Fig. 7), not only during the early stages of adsorption (as was also the case for the MBP method described earlier), but also after 10 min. At this point, equilibrium surface ten√ sion had not been reached. To this end, surface tension vs 1/ t plots (plots not shown here) were made of the same data as in Fig. 7, and equilibrium surface tension values were estimated at t → ∞ (Table 1). For any sample, estimated equilibrium surface tension value of its foam fraction was significantly (P < 0.05) higher than that of its un-fractionated sample and of its liquid fraction. This suggests that the peptides enriched in the foam fraction not only had a higher affinity for the A-W interface, they also seemed able to pack more closely or arrange themselves more efficiently at this interface. After subsequent sinusoidal volume oscillations of the equilibrated drops, E’ could be calculated for all samples. E’ is a measure of the elasticity of the protein film and a high E’ is often associated with high foam stability [35]. E’ values of both (un-fractionated) DH 2 hydrolysates were significantly (P < 0.05) higher than E’ of both (un-fractionated) DH 6 hydrolysates. This is in accordance with the observation that FS of DH 2 hydrolysates was in general better than that of DH 6 hydrolysates (see Section 3.1). The relationship between protein film elasticity and FS of enzymatic gluten hydrolysates has also been described elsewhere [13]. All foam fractions and their respective un-fractionated samples had similar E’ values, except for T2F which had significantly higher (P < 0.05) E’ than T2. All liquid fractions but P6L had significantly (P < 0.05) lower E’ than their respective un-fractionated samples. As discussed above, these liquid fractions all had very low FS (See section 3.1). It seems that, while fractionation led to some differences in E’ between the different fractions, these could not explain all observations regarding FS (see Section 3.1). For example, E’ of T2L was equal to that of T6F, but the latter had much better FS than the former. That foam fractionation only had a limited impact on overall peptide composition of wheat gluten hydrolysates (see Section 3.2) seems an important observation in this context. Although E’ and FS are related to some extent, E’ seems to depend on overall peptide composition while slight changes in peptide composition, such as those induced by the foam fractionation process, can strongly affect FS. 4. Conclusions While the choice of enzyme and hydrolysis conditions can to a certain extent determine which peptides are released, the resultant protein hydrolysate always consists of a heterogeneous mixture of peptides that are efficient at stabilizing interfaces and peptides that are not. This work has shown that a relatively simple foam fractionation process can have a drastic impact on the foaming
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
301
Fig. 4. A. Profiles resulting from subtracting the RP-HPLC chromatograms of the liquid fractions from those of the foam fractions for T2, T6, P2 and P6. The hydrophobic (HPHO) peptides are further divided in HPHO-1 and HPHO-2 peptides. B. Relative amounts of HPHO-1 (30–33 min) and HPHO-2 (33–42 min) peptides in the hydrophobic (30–42 min) part of the profiles for all samples. Codes T2, T6, P2 and P6 as in Fig. 1.
Fig. 5. Comparison of the average apparent molecular mass (MM) distribution of the HPHO-1 and HPHO-2 fractions of T2, T6, P2 and P6, after these fractions were collected from RP-HPLC (as described in 2.9). Dots above the curves indicate apparent MM marker proteins of 6.5, 14.3 and 45 kDa. In all cases, the HPHO-2 fraction contained a number of peptides with a higher apparent MM (>45 kDa), which was much less the case for the HPHO-1 fraction. Codes T2, T6, P2 and P6 as in Fig. 1.
302
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
Fig. 6. Surface tension in the early stages of adsorption at the air-water (A-W) interface (from 10 ms to 10 s after bubble formation in a protein solution) of T2, T6, P2 and P6 liquid and foam fractions, compared to their respective un-fractionated samples. Codes T2, T6, P2 and P6 as in Fig. 1.
Fig. 7. Surface tension at the air-water interface after forming a hanging drop with solutions of T2, T6, P2 and P6 and their respective foam and liquid fractions. After 10 min, the drops were sinusoidally oscillated to determine surface dilatational moduli (E’). Codes T2, T6, P2 and P6 as in Fig. 1.
properties of such a complex enzymatic protein hydrolysate. Freeze dried foam fractions had FS as good or better than those of the respective un-fractionated samples, and much better than those of
the liquid fractions, which all had very low FS. When comparing surface-active properties, the foam fractions seemed to have the higher affinity for the A-W interface and caused a larger decrease in
A.G.B. Wouters et al. / Colloids and Surfaces B: Biointerfaces 151 (2017) 295–303
surface tension when approaching equilibrium. This suggests that peptides in the foam fraction arrange themselves in a more efficient way at the interface than those in the liquid fraction. However, although there were some notable differences, most foam and liquid fractions did not have remarkably high or low E’, respectively, compared to those of their respective un-fractionated samples. This suggests that, although protein film elasticity is important for gluten hydrolysate foaming properties, other mechanisms are also in play. These will be addressed in follow-up work. Even more important is that foam fractionation is an excellent tool for studying the structural properties of the peptides which cause the differences in air-water interfacial behavior described above. When comparing the peptide composition of each sample before and after fractionation, only small differences were observed, despite large differences in FS. However, a small fraction of hydrophobic peptides with a relatively high apparent MM was enriched in foam fractions. These peptides were also present in higher levels in un-fractionated DH 2 hydrolysates than in DH 6 hydrolysates, which is in accordance with the better FS of the first. These peptides likely play a key role in efficiently stabilizing foam. In contrast to other studies, in which usually the overall properties of the entire hydrolysate were studied, our evidence suggests that it is rather the presence of certain specific peptides that determines the foam stabilizing ability of the entire hydrolysate. Acknowledgements I. Rombouts thanks the Research Foundation − Flanders (FWO, Brussels, Belgium) for financial support. K. Brijs acknowledges the Industrial Research Fund (KU Leuven, Leuven, Belgium) for a position as Industrial Research Manager. J.A. Delcour holds the W.K. Kellogg Chair in Cereal Science and Nutrition at KU Leuven. This work is part of the Methusalem program “Food for the Future” at KU Leuven. References [1] A. Van Der Borght, H. Goesaert, W.S. Veraverbeke, J.A. Delcour, Fractionation of wheat and wheat flour into starch and gluten: overview of the main processes and the factors involved, J. Cereal Sci. 41 (2005) 221–237. [2] L. Day, M.A. Augustin, I.L. Batey, C.W. Wrigley, Wheat gluten uses and industry needs, Trends Food Sci. Technol. 17 (2006) 82–90. [3] W.S. Veraverbeke, J.A. Delcour, Wheat protein composition and properties of wheat glutenin in relation to breadmaking functionality, Crit. Rev. Food Sci. Nutr. 42 (2002) 179–208. [4] J.A. Delcour, I.J. Joye, B. Pareyt, E. Wilderjans, K. Brijs, B. Lagrain, Wheat gluten functionality as a quality determinant in cereal-based food products, Annu. Rev. Food Sci. Technol. 3 (2012) 469–492. [5] J. Adler-Nissen, Enzymatic hydrolysis of proteins for increased solubility, J. Agric. Food Chem. 24 (1976) 1090–1093. [6] S. Damodaran, Protein stabilization of emulsions and foams, J. Food Sci. 70 (2005) R54–R66. [7] B.S. Murray, Stabilization of bubbles and foams, Curr. Opin. Colloid Interface Sci. 12 (2007) 232–241. [8] T.N. Hunter, R.J. Pugh, G.V. Franks, G.J. Jameson, The role of particles in stabilising foams and emulsions, Adv. Colloid Interface Sci. 137 (2008) 57–81. [9] S.R. Drago, R.J. González, Foaming properties of enzymatically hydrolysed wheat gluten, Innov. Food Sci. Emerg. Technol. 1 (2000) 269–273. [10] X.Z. Kong, H.M. Zhou, H.F. Qian, Enzymatic preparation and functional properties of wheat gluten hydrolysates, Food Chem. 101 (2007) 615–620.
303
[11] E. Linares, C. Larre, M. Lemeste, Y. Popineau, Emulsifying and foaming properties of gluten hydrolysates with an increasing degree of hydrolysis: role of soluble and insoluble fractions, Cereal Chem. 77 (2000) 414–420. [12] B. Mimouni, J. Raymond, A.M. Merle-Desnoyers, J.L. Azanza, A. Ducastaing, Combined acid deamidation and enzymic hydrolysis for improvement of the functional properties of wheat gluten, J. Cereal Sci. 20 (1994) 153–165. [13] A.G.B. Wouters, I. Rombouts, M. Legein, E. Fierens, K. Brijs, C. Blecker, J.A. Delcour, Air–water interfacial properties of enzymatic wheat gluten hydrolyzates determine their foaming behavior, Food Hydrocolloid 55 (2016) 155–162. [14] K.K. Agyare, K. Addo, Y.L. Xiong, Emulsifying and foaming properties of transglutaminase-treated wheat gluten hydrolysate as influenced by ph, temperature and salt, Food Hydrocolloid 23 (2009) 72–81. [15] E.E. Babiker, N. Fujisawa, N. Matsudomi, A. Kato, Improvement in the functional properties of gluten by protease digestion or acid hydrolysis followed by microbial transglutaminase treatment, J. Agric. Food Chem. 44 (1996) 3746–3750. [16] S. Berot, Y. Popineau, J.P. Compoint, C. Blassel, B. Chaufer, Ultrafiltration to fractionate wheat polypeptides, J. Chromatogr. B 753 (2001) 29–35. [17] Y. Popineau, B. Huchet, C. Larre, S. Berot, Foaming and emulsifying properties of fractions of gluten peptides obtained by limited enzymatic hydrolysis and ultrafiltration, J. Cereal Sci. 35 (2002) 327–335. [18] J.S. Wang, M.M. Zhao, Y. Bao, T. Hong, C.M. Rosella, Preparation and characterization of modified wheat gluten by enzymatic hydrolysis-ultrafiltration, J. Food Biochem. 32 (2008) 316–334. [19] J.S. Wang, M.M. Zhao, X.Q. Yang, Y.M. Jiang, Improvement on functional properties of wheat gluten by enzymatic hydrolysis and ultrafiltration, J. Cereal Sci. 44 (2006) 93–100. [20] I. Barackov, A. Mause, S. Kapoor, R. Winter, G. Schembecker, B. Burghoff, Investigation of structural changes of -casein and lysozyme at the gas–liquid interface during foam fractionation, J. Biotechnol. 161 (2012) 138–146. [21] B. Burghoff, Foam fractionation applications, J. Biotechnol. 161 (2012) 126–137. [22] A.P. Shea, C.L. Crofcheck, F.A. Payne, Y.L. Xiong, Foam fractionation of ␣-lactalbumin and -lactoglobulin from a whey solution, Asia-Pac. J. Chem. Eng. 4 (2009) 191–203. [23] C.E. Lockwood, P.M. Bummer, M. Jay, Purification of proteins using foam fractionation, Pharm. Res. 14 (1997) 1511–1515. [24] T. Nakabayashi, Y. Takakusagi, K. Iwabata, K. Sakaguchi, Foam fractionation of protein: correlation of protein adsorption onto bubbles with a ph-induced conformational transition, Anal. Biochem. 419 (2011) 173–179. [25] P. Dhordain, M. Bigan, M. Vanhoute, C. Pierlot, J.M. Aubry, P. Dhulster, D. Guillochon, R. Froidevaux, Optimization of peptide separation from complex peptide mixture in a foaming-draining system, Sep. Sci. Technol. 47 (2012) 654–662. [26] M. Vanhoute, R. Froidevaux, C. Pierlot, F. Krier, J.M. Aubry, D. Guillochon, Advancement of foam separation of bioactive peptides using an aeration column with a bubbling-draining method, Sep. Purif. Technol. 63 (2008) 460–465. [27] AOAC, Official Methods of Analysis. Method 990.03, Association of Official Analytical Chemists, Washington, DC, USA, 1995. [28] J. Adler-Nissen, Enzymic Hydrolysis of Food Proteins, Elsevier Applied Science Publishers, New York, USA, 1985, p. [29] P. Diermayr, L. Dehne, Controlled enzymatic hydrolysis of proteins at low ph values.1. Experiments with bovine serum-albumin, Z. Lebensm. Unters. Forsch. 190 (1990) 516–520. [30] P.M. Nielsen, D. Petersen, C. Dambmann, Improved method for determining food protein degree of hydrolysis, J. Food Sci. 66 (2001) 642–646. [31] P.W.J.R. Caessens, H. Gruppen, S. Visser, G.A. van Aken, A.G.J. Voragen, Plasmin hydrolysis of beta-casein: foaming and emulsifying properties of the fractionated hydrolysate, J. Agric. Food Chem. 45 (1997) 2935–2941. [32] I. Celus, K. Brijs, J.A. Delcour, Enzymatic hydrolysis of brewers’ spent grain proteins and technofunctional properties of the resulting hydrolysates, J. Agric. Food Chem. 55 (2007) 8703–8710. [33] X. Guan, H.Y. Yao, Z.X. Chen, L.A. Shan, M.D. Zhang, Some functional properties of oat bran protein concentrate modified by trypsin, Food Chem. 101 (2007) 163–170. [34] A.G.B. Wouters, I. Rombouts, E. Fierens, K. Brijs, J.A. Delcour, Relevance of the functional properties of enzymatic plant protein hydrolysates in food systems, Comp. Rev. Food Sci. Food Saf. 15 (2016) 786–800. [35] B.S. Murray, Rheological properties of protein films, Curr. Opin. in Colloid & Interface Sci. 16 (2011) 27–35.