ARTICLE pubs.acs.org/JAFC
Behavior of Protein in the Presence of Calcium during Heating of Whey Protein Concentrate Solutions Emmanuelle Riou,†,‡ Palatasa Havea,*,‡ Owen McCarthy,† Philip Watkinson,‡ and Harjinder Singh§ †
Institute of Food, Nutrition and Human Health, Massey University, Private Bag 11222, Palmerston North, New Zealand Fonterra Research Centre, Private Bag 11029, Dairy Farm Road, Palmerston North, New Zealand § Riddet Institute, Massey University, Private Bag 11222, Palmerston North, New Zealand ‡
ABSTRACT: The effect of added CaCl2 on heat-induced changes in whey protein (WP) solutions prepared from whey protein isolate (WP1), acid whey protein concentrate (WP2), and cheese whey protein concentrate (WP3) was investigated. The loss of native-like, proteins, aggregation, and gel firmness of WP were maximum at certain levels of added CaCl2. These levels were different for different WP products. The effect of added CaCl2 on these changes appeared to be related to the initial calcium concentrations of these solutions. The higher the calcium content of the product, the less available sites for added CaCl2 to bind. It was considered that addition of CaCl2 changed the types of protein interactions that formed the protein aggregates during heating. Added calcium caused dramatic decreases in fracture stress of WP gels due to the formation of large protein aggregates. KEYWORDS: whey protein concentrate, whey protein isolate, calcium, heating
’ INTRODUCTION Whey protein (WP) products have been used as functional ingredients in many food applications, mainly as gelling agents.1 The effect of minerals on heat-induced WP aggregation has been studied extensively over the past two decades. The mineral content of commercial WP products such as whey protein concentrate (WPC) has been shown to play a major role in the heat-induced aggregation of WP solutions prepared from cheese WPC and acid WPC.2,3 Havea et al.3 suggested that the monovalent cations and the divalent cations influenced protein aggregation via different mechanisms. The high calcium content of cheese WPC promoted relatively rapid losses of native proteins and the formation of large aggregates (>0.22 μm) during heating. The rapid formation of large protein aggregates was suggested to be due to the formation of calcium bridges between negatively charged residues, as well as to rapid formation of hydrophobic interactions. In contrast, the low calcium content of acid WPC limited the formation of calcium bridges and led to the formation of small, predominantly disulfide-linked, aggregates. The role of calcium in the thermal denaturation/aggregation of WP has been reported to involve three phenomena: calcium bridging,46 electrostatic shielding of negative charges on the protein,7 and ion-induced conformational changes that could lead to hydrophobic interactions at elevated temperatures.8,9 The effect of calcium on heat-induced changes in WPs has been the subject of numerous publications in recent years.5,6,10,11 These studies were carried out predominantly on model systems with well-defined protein compositions, usually pure β-lactoglobulin (β-lg) at a low concentration of 1 or 2%, and defined mineral buffer systems. It is usually difficult to meaningfully extrapolate the findings from these studies to commercial WP products in which many factors, such as mixtures of different protein components with different thermal properties and heterogeneity of the mineral environment, are at play. r 2011 American Chemical Society
Although the above studies, among many others, provide some information about the impact of various minerals on the aggregation and/or gelation of WPs, the controlling mechanisms of the effects of different minerals on the denaturation, aggregation, and gelation of WP in more complex protein systems (e.g., WPC) are not fully understood. The objective of this study was to investigate the effect of added calcium on the loss of native-like proteins, aggregation, and gelation of WP solutions prepared from commercial WP products with different mineral contents: whey protein isolate (WPI) (WP1), acid WPC (WP2) and cheese WPC (WP3). It should be noted that the current study was orientated to industry, where there is much interest in the functional properties of overall protein systems. In this context, the actual impact of treatments such as the addition of CaCl2 to protein systems at macro levels is of significant technological importance. These products may contain denatured whey proteins,12 but the effect of added CaCl2 on the whole protein system is of significant interest. Scientific insights from recent studies on model systems may help to explain the findings from the current study.
’ MATERIALS AND METHODS Materials. Commercial WP powders (WP1, WP2, and WP3) were obtained from Fonterra Co-operative Group (formerly Anchor Products), Hamilton, New Zealand. These products were typical of their commercial specifications. The chemicals used for the preparation of the electrophoresis buffers (obtained from Bio-Rad Laboratories, Richmond, CA) were of analytical grade. Received: May 11, 2011 Accepted: November 8, 2011 Revised: November 7, 2011 Published: November 08, 2011 13156
dx.doi.org/10.1021/jf203610k | J. Agric. Food Chem. 2011, 59, 13156–13164
Journal of Agricultural and Food Chemistry Composition of WP Powders. The total protein contents of the WP powders were determined using the Kjeldahl method13 with a nitrogen conversion factor of 6.38. The fat content was determined using the Soxhlet extraction method, as described by Russell et al.14 The moisture content was determined by oven drying preweighed duplicate samples at 102 °C for 5 h, cooling in a desiccator for 2 h, and reweighing. The lactose contents were as reported by the manufacturer. The mineral analyses were carried out at the New Zealand Pastoral Agricultural Research Laboratory, Palmerston North, by inductively coupled argon-plasma emission spectrometry using the method described by Lee et al.15 Preparation of WP Solutions. Appropriate quantities of WP powder were dissolved in water (purified using a Milli-Q system; Millipore Corp., Bedford, MA) so that the final solutions had the desired protein concentrations (10, 40, 48, or 120 g/kg). The solutions were stirred for 1 h at room temperature using a magnetic stirrer, and the pH was adjusted where necessary using 1.0 M NaOH or HCl. The WP solutions were then used for various analyses and experiments. To analyze the effect of added salts (CaCl2 or NaCl) on the aggregation or gelation of WP solutions, appropriate volumes of a 2 M CaCl2 solution were added to give 0, 2, 4, 6, 8, 11, 14, 17, 20, 50, 80, 110, 140, 170, 200, and 230 mM added CaCl2 concentrations. To maintain constant protein contents, volumes of CaCl2 were added to solutions at higher protein concentrations than required, and the protein concentrations were then corrected by the addition of appropriate volumes of Milli-Q water. For NaCl (2 M), volumes were added at levels simulating the same ionic strengths of samples with added CaCl2, and then diluted the same way with Milli-Q water. Heat Treatment of WP Solutions. Aliquots (3 mL) of WP solutions (40 g/kg) with various levels of added CaCl2 were heated at 75, 80, or 85 °C for 5 min in glass tubes (7.5 cm in length, 1.2 cm in outside diameter, and 1.0 cm in inside diameter). The tubes were immediately immersed in ice water (∼0 °C) for 1 h to stop denaturation/ aggregation of the proteins. Samples (0.05 mL) of each heated solution were diluted with appropriate sample buffers so that a final protein concentration of 2 g/kg was obtained. The samples were then analyzed using either alkaline polyacrylamide gel electrophoresis (PAGE) or sodium dodecyl sulfate (SDS)PAGE as described by Havea et al.12 Polyacrylamide Gel Electrophoresis. The control and heattreated solutions were analyzed using a Mini-Protean II dual cell system (Bio-Rad Laboratories). After preparing the appropriate (alkaline or SDS) gel, a 10 μL sample of each solution was injected into the sample wells, and the proteins were then separated using electrophoresis. After the gels had been stained with Coomassie Blue dye and destained, as described by Havea et al.,12 they were scanned using a computing laser densitometer (Molecular Dynamics, Sunnyvale, CA; model P.D.). The integrated intensities of the WP bands were determined using the associated integration software. The changes in the WP solutions as affected by heating and calcium addition were determined by comparing the residual WP band intensities of the treated WP sample with those of an untreated sample (control). Zeta Potential Measurements. WP solutions (10 g/kg, pH 6.7) for zeta potential measurements were prepared. The pH was adjusted to 6.0 or 5.5 using 1.0 M HCl solution. These solutions were not heat treated, and samples were filtered with a 0.45 μm Minisart filter (Sartorius Stedim) before analysis. The zeta potentials of the WP solutions were determined using a Malvern Zetasizer Nano instrument, the associated Malvern Multi-8 64 channel correlator, and the clear disposable electrophoresis cell, which incorporates a folded capillary (Malvern Instruments Ltd., Malvern, Worcestershire, U.K.). The temperature of the electrophoresis cell was maintained at 25 °C by a water jacket that was temperature controlled by the Peltier system associated with the cell. An applied voltage of 150 V was used in all experiments.
ARTICLE
Triplicate samples were measured five times each; the measurements were significant at P e 0.05. Rheological Measurements. WP solutions (48 g/kg, pH 6.7), with 0, 4, 10, 20, 80, 140, and 200 mM added CaCl2, for rheological measurements were also prepared. (Note: As heating of 40 g/kg solutions did not result in smooth curves, a slightly higher concentration of 48 g/kg was used.) To account for the different protein compositions of the WP products, appropriate quantities of pure β-lg (obtained from Sigma Chemical Co., St. Louis, MO) were added to solutions of WP2 and WP3 so that their β-lg to α-lactalbumin (α-la) ratios were similar to that of the WP1 solutions. Oscillatory rheological measurements were made using a stress-controlled Paar Physica rheometer model UDS200 (Anton Paar GmbH, Graz, Austria) with a bob and cup geometry, a frequency of 0.1 Hz and a shear strain of 1%. Samples (19 mL) were heated from 20 to 85 °C at a rate of 1 °C/min, held at 85 °C for 30 min, then cooled to 20 °C at a rate of 1 °C/min, and held at 20 °C for 10 min, while measuring the storage modulus (G0 ) of the solutions. After the heating/cooling phase, a frequency sweep (10 to 0.01 Hz) followed by a strain sweep (0.1 to 10%) was carried out on the WP gels. The data (results not shown) showed that the conditions used for these measurements were well within the linear viscoelastic regions of the WP gels. Compression Testing. WP solutions (120 g/kg, pH 6.7) with various levels of added CaCl2 (0, 12, 30, 60, 240, 420, and 600 mM) were also prepared. They were placed in medium-walled polycellulosic plastic tubes (400 cm in length and 30 mm in diameter), and the ends were closed off using rubber bands to make stiff “sausages”. These sausages were then placed in a thermostatically controlled water bath (85 ( 0.5 °C) for 30 min. It took approximately 48 s to heat the centre of each tube to 84.7 °C. After heating, the tubes were removed from the water bath and immediately placed under cold running water (∼22.5 °C) for 30 min before being stored at 4 °C overnight. The gels were then analyzed using a range of techniques. Compression tests were carried out as described by Havea et al.16 The results were used to describe the textural nature of each gel according to the methods described by Hamann and MacDonald17 and Truong and Daubert.18 Fracture strain and fracture stress were used to broadly characterize texture. Transmission Electron Microscopy. Samples of each gel used for the compression tests were prepared for transmission electron microscopy (TEM). The samples were fixed, stained, and photographed using a transmission electron microscope (Philips CM10 equipped with an SIS Morada high resolution digital imaging system, Eindhoven, The Netherlands) as described by Langton and Hermansson.19 Statistical Analysis. Data collected from duplicate experiments were used to calculate the standard errors in the quantification of proteins using PAGE, the rheological measurements, and the compression tests using SigmaPlot (2002 for Windows version 8.02; SPSS Inc., Chicago, IL).
’ RESULTS AND DISCUSSION Composition of WP Powders. The compositional analyses of the WP powders showed that WP1 had a significantly higher protein content (931 g/kg) and lower ash (16 g/kg), fat (5 g/kg), and lactose (2 g/kg) contents than WP2 or WP3, which had similar compositions (Table 1). The main differences between WP2 and WP3 were that (1) WP3 contained ∼150 g of glycomacropeptide/kg (GMP), which was also measured as part of the total protein content (i.e., 18% of the total protein accounted for GMP); and (2) WP2 contained more (485 mmol/kg protein) total monovalent cation (K+ and Na+) but a third less (82 mmol/kg protein) total divalent cation (Ca2+ and Mg2+) 13157
dx.doi.org/10.1021/jf203610k |J. Agric. Food Chem. 2011, 59, 13156–13164
Journal of Agricultural and Food Chemistry
ARTICLE
Table 1. Composition of WP Products Used composition
WP1
WP2
WP3
protein (g/kg)
931
804
830
fat (g/kg)
5
54
60
ash (g/kg)
16
38
30
lactose (g/kg)
2
41
40
moisture (g/kg)
49
50
50
Ca
19
74
111
K Mg
13 2
471 8
176 28
Na
259
14
147
minerals (mmol/kg) protein
than WP3 (323 and 139 mmol/kg protein, respectively). WP1 and WP2 did not contain any GMP. Using a reversed-phase high performance liquid chromatography (HPLC) method,20 it was estimated that, for 48 g/kg protein solutions prepared from WP1, WP2, and WP3 powders, the β-lg contents were 34, 25, and 18 g/kg, respectively, and the α-la contents were 7.6, 7.4, and 7.8 g/kg, respectively. Thus, the WP products had similar α-la contents but very different β-lg contents. These results also indicated that the WP1, WP2, and WP3 powders had β-lg to α-la ratios of 4.4, 3.3, and 2.3, respectively. Effect of Added CaCl2 on the Aggregation of WPs Loss of Native-like Proteins During Heating. Two terms have been used (Havea et al.12) to describe the likely changes to proteins during heating. When heated whey protein solutions were analyzed using native PAGE, there were protein bands that coincided with the native proteins in the unheated samples. These are referred as “native-like monomeric” proteins. Running the same heated solutions on SDSPAGE gave more intense bands that ran as though they were monomeric. These are referred as “SDS-monomeric” proteins. These terms are also used in this manuscript to describe protein changes in heated WP samples. The loss of native-like protein during heating was investigated in 40 g/kg WP solutions (a common protein concentration used in many food applications). Native-like β-lg had decreased markedly in the heated (85 °C) sample with 4 mM added CaCl2 (Figure 1a, lane 4) and had disappeared completely with an increase in added CaCl2 up to 80 mM (Figure 1a, lane 12). Faint bands corresponding to native-like β-lg appeared in the sample with 110 mM added CaCl2 (Figure 1a, lane 13), and the intensities of these bands appeared to increase with further increase in the level of added CaCl2. These results suggest that all levels of added CaCl2 (2230 mM) enhanced the aggregation of β-lg compared with the control (lane 2) but that some inhibition of aggregation took place at higher levels (g110 mM) of added CaCl2. The enhancement of the denaturation/aggregation of β-lg by calcium during heating has been reported previously.6,10,11,21 It appears that added CaCl2 prevents the unfolded regions from refolding into the native state, hence promoting aggregation, subsequent to unfolding, during heating. A considerable amount of native-like α-la remained after heating the control solution (compare lanes 1 and 2 in Figure 1a). The loss of native-like α-la increased with an increase in the level of added CaCl2 to a maximum at around 2050 mM added CaCl2 (lanes 11 and 12) and then decreased with further increases in the level of added CaCl2. These results suggest that
Figure 1. (a) Alkaline PAGE and (b) SDSPAGE patterns of heated (85 °C, 5 min) WP1 solutions (40 g/kg) with added CaCl2. Lane 1, unheated, no added CaCl2; lane 2, heated, no added CaCl2; lanes 317, heated with 2, 4, 6, 8, 11, 14, 17, 20, 50, 80, 110, 140, 170, 200, and 230 mM added CaCl2, respectively.
addition of CaCl2 to the protein solutions enhanced the loss of native-like α-la, the effect being optimal at ∼50 mM added CaCl2. It appeared that high levels of added CaCl2 (>110 mM) stabilized the α-la structure. Bovine serum albumin (BSA) disappeared completely after heating for 5 min, even in the heated control sample without any added CaCl2 (compare lanes 1 and 2 in Figure 1a). However, in the sample with 110 mM added CaCl2, the BSA band reappeared (lane 13) and continued to increase in intensity with an increase in the level of added CaCl2 up to 230 mM. The results indicated that BSA is sensitive to heat treatment, as reported previously,12 but that its aggregation is somewhat inhibited in the presence of >110 mM added CaCl2. Quantification of the PAGE band intensities provided typical loss profiles for the native-like proteins (β-lg, α-la, or BSA) during heating (75, 80, or 85 °C) as affected by increasing levels of added CaCl2. It is clear that aggregation of these proteins was greatest at ∼5080 mM added CaCl2 and decreased with further increases in added CaCl2. The same trends were observed for the three WP systems (results not shown). Sherwin and Foegeding21 reported that the stoichiometric relationship between the concentrations of calcium and β-lg was the important factor in determining the rate of protein aggregation, that there was a maximum aggregation rate at CaCl2 (mM) to protein (%, w/v) ratios of between 3.33 and 23.3, and that excess CaCl2 appeared to have an inhibitory effect on protein aggregation. A concentration of 3 mM added CaCl2 was reported to induce maximal aggregation of denatured β-lg (1%, w/w) when heated at 78 °C for 10 min.11,22 In these studies (which were conducted predominantly on pure protein model systems), observations were interpreted on the basis of the effect of added CaCl2 on interactions (e.g., the formation of calcium bridges), giving rise to the formation of protein aggregates. The effects of added calcium on the loss of proteins when heating complex systems in this study are consistent with the above findings. Loss of SDS-Monomeric Proteins During Heating. The losses of SDS-monomeric proteins (Figure 1b) followed similar patterns to the losses of native-like proteins (Figure 1a). There appeared to be some faint bands corresponding to BSA in the 13158
dx.doi.org/10.1021/jf203610k |J. Agric. Food Chem. 2011, 59, 13156–13164
Journal of Agricultural and Food Chemistry heated control, suggesting that some noncovalently associated BSA aggregates may have been dissociated under SDS conditions (Figure 1b, lane 2). Otherwise, the loss of SDS-monomeric BSA in the samples with added CaCl2 was very similar to the loss of native-like BSA. Likewise, the loss of SDS-monomeric α-la (Figure 1b) was similar to the loss of native-like α-la (Figure 1a); the α-la band intensities in Figures 1a and 1b were similar. These results suggest that disulfide bonding of BSA and α-la was predominantly involved in aggregate formation during heat treatment. Possible involvement of BSA and α-la in noncovalent interactions was not apparent from these results. Although the loss of SDS-monomeric β-lg was similar to the loss of native-like β-lg, it was clear that considerable amounts of noncovalently linked aggregates dissociated under SDS conditions to give more intense bands (Figure 1b). The intensity of the β-lg bands decreased to a minimum at ∼5080 mM added CaCl2 and then increased with further increases in added CaCl2. Trends similar to all those discussed above were observed for samples heated at 75 and 80 °C, and for WP2 and WP3 solutions heated under the same conditions (results not shown). α-La is widely known for its calcium-binding properties.23,24 The role of bound calcium ions appears to be stabilization of the tertiary structure.25 The added CaCl2 probably limited the extent of the unfolding of α-la molecules, suppressing the exposure of some of the hydrophobic residues, and hence hydrophobic aggregations, and the aggregation of any denatured α-la was predominantly by disulfide linkages. The effect of calcium on the denaturation/aggregation of BSA is not fully understood. Because BSA contains more disulfide bonds than β-lg and contains a free thiol group, its denatured form will be more susceptible to the formation of intermolecular disulfide-linked aggregates (Figure 1). Effect of a Shift in pH on the Loss of Native-like WPs. When CaCl2 at levels from 0 to 230 mM was added to 40 g/kg WP solutions, the pH shifted from 6.70 to 6.08, 6.70 to 6.11, and 6.70 to 6.36 for WP1, WP2, and WP3, respectively. To investigate the effect of this shift in pH, caused by the added CaCl2, on WP aggregation, three sets of samples (40 g/kg) with the same range of added CaCl2 (0230 mM) were prepared for each of the WP systems: (1) samples with added CaCl2 without adjusting the pH; (2) samples with added CaCl2 but with the pH adjusted back to the original pH of the solutions without added CaCl2 by the addition of 1.0 M NaOH; (3) samples without added CaCl2 but with the pH adjusted to reflect the corresponding pHs of the samples with added CaCl2 by the addition of 1.0 M HCl. These sets were then heated (85 °C, 5 min), and the losses of proteins were determined using alkaline PAGE or SDS PAGE. The results (not shown) demonstrated that heat-induced losses of protein in the samples of set 1 and in the samples of set 2 were the same. It was also clear that the heat-induced loss of proteins in the samples of set 3 did not vary with decreasing pH over the range of the shift in pH for each WP system. We concluded that the shifts in pH in these samples due to the addition of CaCl2 did not significantly alter the trends observed due to the addition of CaCl2. Therefore, we conducted further experiments without adjusting the pH of the solutions after the addition of CaCl2. Effect of Added NaCl on the Loss of Native-like WPs. In these experiments, WP solutions (40 g/kg, pH 6.7) were heated under the same conditions but NaCl was added at levels that simulated the ionic strength of each of the samples with added CaCl2 (results shown in Figure 1). Figure 2 shows the effect of added
ARTICLE
Figure 2. (a) Alkaline PAGE and (b) SDSPAGE patterns of heated (85 °C, 5 min) WP1 solutions (40 g/kg) with added NaCl. Lane 1, unheated, no added NaCl; lane 2, heated, no added NaCl; lanes 317, heated with 6, 12, 18, 24, 33, 42, 51, 60, 150, 240, 330, 420, 510, 600, and 690 mM added NaCl, respectively.
NaCl on the aggregation of WPs. The trends observed in Figure 1 were also observed in Figure 2. However, the effect of added NaCl (Figure 2) on the loss of native monomeric and SDSmonomeric proteins was much less than the effect of added CaCl2 (Figure 1). For example, no native-like β-lg remained after heating (85 °C, 5 min) the sample with 6 mM added CaCl2 (Figure 1a, lane 5), whereas 26% of the original native-like β-lg remained after heating (85 °C, 5 min) the sample with added NaCl of equivalent ionic strength (Figure 2a, lane 5). This indicated that, although there may be an ionic strength effect, the effect of added CaCl2 is more than just this. Veerman et al.26 reported that both monovalent cations and divalent cations screen electrostatic interactions but that the effect is greater with divalent cations. Effect of Added CaCl2 on the Rheological Properties of Heated WP Solutions. In order to obtain smooth curves for the changes in storage modulus, G0 , with heating time, solutions with protein concentrations of 48 g/kg were used, instead of 40 g/kg as used for the PAGE analyses (Figures 1 and 2). The effect of added CaCl2 on the gel stiffness (storage modulus, G0 ) of the WP1 solutions (48 g/kg, pH 6.7) during heating is shown in Figure 3a. Figure 4 shows the effect of added CaCl2 on the final storage modulus (G0 final) of 48 g/kg WP solutions prepared from the three WP powders. Table 2 summarizes specific rheological characteristics of the three systems. When the WP (WP1, WP2, and WP3) solutions with no added CaCl2 were heated, no measurable changes in the values of G0 were observed, i.e., no gel networks were formed (Table 2). When 4 mM CaCl2 was added, all the WP solutions formed gel networks (see also Figure 3b). The effect of added CaCl2 on the gel point, defined as either the heating time or the heating temperature at which measurable values of G0 were first registered, followed similar trends in all WP systems. The lowest gel points were achieved in heated protein solutions with 10 or 20 mM added CaCl2. Beyond these levels, the gel point increased with an increase in added CaCl2. At any level of added CaCl2, the gel point for WP2 was higher than that for either WP1 or WP3 (Table 2). For WP1, the strongest gel (highest G0 final) was 13159
dx.doi.org/10.1021/jf203610k |J. Agric. Food Chem. 2011, 59, 13156–13164
Journal of Agricultural and Food Chemistry
ARTICLE
Figure 4. Effect of the total residual calcium on the G0 final of the heatinduced gels (48 g/kg) prepared from WP1 (b), WP2 (O), or WP3 (1). The WP solutions were heated from 20 to 85 °C at 1 °C/min, held for 30 min, cooled from 85 to 20 °C at 1 °C/min, and then held for 10 min. Error bars: (standard error, n = 2.
Figure 3. Effect of added CaCl2 on thermally induced changes in the storage modulus (G0 ) of (a) the WP1 solutions and (b) the WP solutions with 4 mM added CaCl2: 48 g/kg, heated from 20 to 85 at 1 °C/min, held at 85 °C for 30 min, cooled to 20 at 1 °C/min, and held at 20 °C for 10 min. G0 was measured every minute. (a) b, no added CaCl2; O, 4 mM added CaCl2; 1, 10 mM added CaCl2; 9, 20 mM added CaCl2; 3, 80 mM added CaCl2; 0, 140 mM added CaCl2; [, 200 mM added CaCl2. (b) b, control (no added CaCl2, WP1); O, WP1; 1, WP2; 3, WP3.
formed by the solution with 4 mM added CaCl2 (Figure 4 and Table 2). As the level of added CaCl2 increased, G0 final decreased, indicating weaker gels. The observation that weaker gels were formed from solutions with higher levels of added CaCl2, and that the extent of aggregation was lower, could probably be attributed to the effect of CaCl2 on protein aggregation (Figure 1). How calcium suppresses the aggregation of WP during heating is a question that demands an answer. For both WP2 and WP3, G0 final increased with increasing added CaCl2 up to 20 mM, at which the gels were stiffest (Table 2). Further increases in the level of added CaCl2 resulted in a decrease in G0 final. The stiffest WP1 gel (846 Pa) was more than twice as stiff as the stiffest WP2 gel (332 Pa) and almost four times stiffer than the stiffest WP3 gel (221 Pa). The G0 final values were high at low levels of added CaCl2 (e25 mM), but then decreased with increasing levels of added CaCl2. The protein solutions (48 g/kg) made from these WP products required higher protein concentrations of around 80 to 100 g/kg to form heat induced gels. The total calcium concentrations in the original WP solutions (48 g/kg) were 2.1, 8.4, and 11.2 mM for WP1, WP2, and WP3, respectively. It is clear that at 48 g/kg added CaCl2 induced gel formation in all WP solutions regardless of the original calcium concentration.
Although the WP3 solutions had the highest calcium content (11.2 mM), they still needed added CaCl2 to induce gel formation during heating. In contrast, although the WP1 solution had a low initial calcium content (2.1 mM), a strong gel was formed when 4 mM CaCl2 was added (giving an overall residual calcium content of 6.1 mM, which was lower than the initial calcium content of the WP3 solutions). It is also clear that the initial calcium content had an effect on G0 final; the lower the initial calcium content, the higher was the G0 final at 4 mM added CaCl2 (Figures 3b and 4). It appears that the calcium initially present in the WP solutions was probably bound to the proteins. As such, it would be ineffective in inducing aggregation of the WPs during heating. Simons et al.6 suggested that calcium needs to be bound specifically to carboxylate groups with a threshold affinity to trigger β-lg aggregation. Protein aggregation itself is then driven by exposure of, and interaction between, denatured parts of the protein molecules without hindrance from electrostatic repulsion between these molecules. We observed different gel firmness (G0 ) values for the three protein systems (Figure 3b). It is unclear if at least part of these differences could probably be attributable to the different degrees of calcium binding to the protein in the systems. When the system contains less calcium (e.g., WP1), more sites could be available for calcium binding, hence the effect would be more pronounced. The opposite is true for when the calcium content is high (e.g., WP3). The zeta potential results at pH 6.7 (Table 3) suggested that the available carboxylate groups were limited in the WP2 and WP3 systems, compared with the WP1 system, so that the binding of calcium might have been significant in this protein system, resulting in more interactions and much stiffer gels (Figure 3b), and the formation of much larger protein aggregates (see Figure 6) than those formed in either the WP2 system or the WP3 system. The differences between WPC2 and WPC3 are probably attributable to the different effects of high sodium (WP2) and high calcium (WP3) in these products (Table 1).3 Effect of Added CaCl2 on the Fracture Properties and Microstructure of Heat-Induced WP Gels. The use of WP products as functional ingredients normally involves heat 13160
dx.doi.org/10.1021/jf203610k |J. Agric. Food Chem. 2011, 59, 13156–13164
Journal of Agricultural and Food Chemistry
ARTICLE
Table 2. Rheological Characteristics of Heated 48 g/kg WP Solutions ((Standard Error, n = 2) added CaCl2 (mM) specific characteristics
0
4
10
20
80
140
200
gel pointa (°C)
no gel
76 ( 0.0
74 ( 0.0
75 ( 0.5
gel pointb (min)
no gel
56 ( 0.0
54 ( 0.0
55 ( 0.7
78 ( 0.0
80 ( 0.0
81 ( 0.0
58 ( 0.0
60 ( 0.0
G0 85c (Pa)
no gel
53 ( 1
39 ( 3
32 ( 5
27 ( 4
18 ( 2
61 ( 0.0 11 ( 1
G0 finalc (Pa)
no gel
846 ( 9
683 ( 32
627 ( 50
610 ( 0.0
496 ( 25
359 ( 25
gel pointa (°C)
no gel
82 ( 0.5
76 ( 0.0
76 ( 0.0
81 ( 0.5
83 ( 0.0
84 ( 0.5
gel pointb (min) G0 85 (Pa)
no gel no gel
62 ( 0.5 7(0
56 ( 0.0 21 ( 0
56 ( 0.0 22 ( 0
61 ( 0.5 11 ( 0
63 ( 0.0 5(0
64 ( 0.5 3(0
G0 final (Pa)
no gel
98 ( 15
295 ( 2
332 ( 3
221 ( 11
173 ( 5
143 ( 4
gel pointa (°C)
no gel
78 ( 0.0
76 ( 0.5
75 ( 0.5
78 ( 0.0
79 ( 0.0
80 ( 0.0
gel pointb (min)
no gel
58 ( 0.0
56 ( 0.5
55 ( 0.5
58 ( 0.0
59 ( 0.0
60 ( 0.0
G0 85 (Pa)
no gel
12 ( 2
14 ( 0
15 ( 2
14 ( 0
13 ( 1
11 ( 1
G0 final (Pa)
no gel
139 ( 2
167 ( 6
210 ( 9
205 ( 2
221 ( 13
208 ( 9
WP1
WP2
WP3
Temperatures at which measurable increases in G0 were first observed. b Times at which measurable increases in G0 were first observed. c G0 85 and G0 final are, respectively, the value of G0 when 85 °C was reached and the value of G0 at the end of the heating cycle. a
Table 3. Zeta Potentials of the Unheated 10 g/kg WP Solutions pH
WP1
WP2
WP3
6.7
28
24
24
6.0
21
18
18
5.5
14
14
15
treatment of reconstituted powders at reasonably high concentrations (e.g., 80120 g/kg). To understand the effect of added CaCl2 on the fracture and textural properties of the WP gels in this high protein concentration range, heat-induced WP gels (120 g/kg) were prepared from the WP products. These gels were then analyzed using large deformation compression tests and TEM. The large deformation test results (Figure 5) reveal more information about the functional properties of the three WP systems. Unlike the heated 48 g/kg WP solutions (Figures 3 and 4), the 120 g/kg WP solutions without added CaCl2 formed gel networks on heating. The fracture stress of the WP1 gels appeared to follow similar trends to those observed for G0 in the small deformation tests (Figures 3a and 4). The gels without added CaCl2 had a fracture stress of ∼4 kPa (Figure 5a). The gels with 12 mM added CaCl2 had more than double the initial fracture stress (∼10 kPa). The fracture stress then decreased with further increases in the level of added CaCl2, to ∼3.5 kPa for the WP1 gel containing 600 mM added CaCl2. For the WP2 gels, the highest fracture stress (4.8 kPa) and fracture strain (∼0.9) were observed in the control gels with no added CaCl2 (Figure 5b). Both the fracture stress and the fracture strain decreased with increasing levels of added CaCl2. For the WP3 gels, the compression test results (Figure 5c) showed that the addition of CaCl2 resulted in negligible change in the fracture stress but that there appeared to be some decrease in the fracture strain. It is difficult to determine any effect of the added CaCl2 on the WP3 gels.
When the gels were cut for compression testing, the control WP1 gels were observed to be strong and clear, and could be described as being rubbery.16,27 When CaCl2 was added, the gels became very opaque and displayed considerable syneresis. Syneresis appeared to increase with an increase in the level of added CaCl2. The gels with high levels of added CaCl2 (e.g., 240 mM) were similar to sponges: water could readily be squeezed out of the gels. Mulvihill and Kinsella28 reported similar behavior in β-lg gels (17%, w/v, pH 8) heated at 90 °C for 30 min in the presence of high (>4000 mM) concentrations of NaCl. In contrast, the control WP2 gels (with no added CaCl2) were slightly more opaque than the control WP1 gels. They were relatively strong and rubbery. The gels became slightly more opaque and less rubbery with an increase in the level of added CaCl2 up to ∼240 mM. With further increases in the level of added CaCl2, the gels became “soft” and “mushy”.27 WP2 and WP3 gels released relatively small amounts of serum. Serum release from the protein gels during the compression test is reported here only as an obvious difference between the WP1 gels and the WP2 and WP3 gels. A proper analysis of this phenomenon following the method of van den Berg et al.29 would have been interesting. However, this analysis was not performed when the experiments were carried out and will be the subject of a future publication. For the purposes of this study, the measured strains and stresses for the protein gels are reported here as “relative stress” and “relative strain” (Figure 5). Zirbel and Kinsella30 reported maximum hardness (force (N) at 70% compression) at 20 mM added CaCl2 in gels formed from 20% protein WPI solutions. The difference between their results and our results can probably be attributed to the different protein concentrations and the different methods of preparing the WP solutions. Kuhn and Foegeding31,32 suggested that a minimal amount of calcium was required for optimal gelation and that the initial calcium concentration played a beneficial role in promoting gelation unless this concentration was too high. Mulvihill and Kinsella28 and Schmidt et al.33 reported that the gel strength increased to a maximum and then decreased with increasing 13161
dx.doi.org/10.1021/jf203610k |J. Agric. Food Chem. 2011, 59, 13156–13164
Journal of Agricultural and Food Chemistry
Figure 5. Effect of added CaCl2 on the fracture stress and fracture strain of the WP gels: 120 g/kg, heated at 85 °C for 30 min, 0 < added CaCl2 < 600 mM. Fracture stress and fracture strain were measured by a 90% compression test on the gel. (a) WP1; (b) WP2; (c) WP3; b, no added CaCl2; O, 12 mM added CaCl2; 1, 30 mM added CaCl2; 3, 60 mM added CaCl2; 9, 240 mM added CaCl2; 0, 420 mM added CaCl2; [ 600 mM added CaCl2. Error bars: (standard error, n = 2.
calcium concentration. This could have been due to an excessive amount of intrachain calcium bridging, with the protein matrix collapsing into large, dense, noncontinuous aggregates surrounded by an aqueous medium. We observed similar results in our microstructure analysis (Figure 6). Recent work16,27 on the impact of different types of protein interactions on the functional properties of WP gels suggested that the intermolecular disulfide linkages between the denatured protein molecules were responsible for the rubberiness (fracture strain) of heat-induced WP gels. The degree of noncovalent association determines the brittleness (fracture stress) of WP gels. It appears that the addition of 4.0 mM CaCl2 to the WP1 systems resulted in a considerable degree of noncovalent intermolecular interaction, resulting in high fracture stress (Figure 5a). However, further addition of CaCl2 resulted in
ARTICLE
decreasing fracture stress with no apparent changes in fracture strain. This suggests that a specific phenomenon, possibly calcium bridging, was probably responsible for the intermolecular interactions. This specific interaction did not affect the degree of disulfide linkages, and hence there was no change in fracture strain (Figure 5a). Addition of CaCl2 to WP2 solutions (Figure 5b) resulted in a decrease in both fracture stress and fracture strain. This indicated that the addition of CaCl2 resulted in the formation of more noncovalent interactions at the expense of disulfide bonds, in comparison with WP1. It should be noted that the WP1 solutions contained significantly more β-lg, which is more susceptible to the formation of noncovalent associations when heated in the presence of added CaCl2 (Figure 1). Increasing the level of added CaCl2 resulted in β-lg forming more noncovalent interactions that dominated the gelling behavior of the WP1 gels (Figure 5a). The TEM micrographs show that the control WP1 gel (Figure 6Ia) had a very fine homogeneous structure (difficult to see even at 25,000 times magnification). With 60 mM added CaCl2, round aggregates of ∼500 μm were formed (Figure 6Ib). With 240 and 600 mM added CaCl2 (Figures 6Ic and 6Id), even bigger aggregates, >2000 μm, were formed. The control WP2 gel had relatively fine structure (Figure 6IIa), similar to that reported previously.3 With 60 mM added CaCl2, relatively larger protein aggregates (100200 μm) were formed (Figure 6IIb). The aggregate size did not seem to increase further with an increase in the level of added CaCl2 (Figures 6IIc and 6IId). The control WP3 gel had a relatively coarse aggregate structure (∼50 μm, Figure 6IIIa). With 60 mM added CaCl2, larger protein aggregates (∼300 μm) were formed (Figure 6IIIb). This structure did not seem to change with the addition of greater amounts of CaCl2 (Figures 6IIIc and 6IIId). Verheul and Roefs34 described gelation as a two-phase process: the formation of the primary spatial structure and then an increase in the amount and/or stiffness of the bonds in the gel, which does not change the spatial structure. It appears that calcium promoted the formation of coarser gels3,3537 in terms of primary spatial structure and that, as the calcium concentration increased, the amount and/or stiffness of the bonds in the gel diminished, leading to weaker gels. Added CaCl2 appeared to have less effect on the microstructure of the WP2 and WP3 gels (Figures 6II and 6III) than on the microstructure of the WP1 gels (Figure 6I). The structures of the control gels (Figure 6a, II and III) were relatively coarse. When CaCl2 was added, larger aggregates were formed, although the aggregates were not as large as those formed in the WP1 gels with added CaCl2. The large coarse structure of the WP1 gels probably explains the observed lack of water-holding capacity. Mulvihill and Kinsella28 observed a maximum gel strength (17% β-lg, pH 8, 90 °C, 30 min) at 10 mM CaCl2, with the microstructures being described as evenly dispersed aggregates that appeared to be linked by stronger well-defined, aggregatetype, strands. At higher levels of CaCl2 (25100 mM), the dispersed matrix collapsed into very large, densely packed protein aggregates in large areas of free aqueous space. These results were consistent with those of the current study. In summary, the effect of added CaCl2 on the heat-induced aggregation and gelation of WP solutions prepared from commercial WP products (WP1, WP2, and WP3) is complex and was maximum at certain levels. It appeared that addition of calcium interferes with the types of interactions that form the protein 13162
dx.doi.org/10.1021/jf203610k |J. Agric. Food Chem. 2011, 59, 13156–13164
Journal of Agricultural and Food Chemistry
ARTICLE
Figure 6. Effect of added CaCl2 on the structure of WP gels. TEM micrographs of 120 g/kg protein solutions heated at 85 °C for 30 min. I, WP1; II, WP2; III, WP3. (a) 0 mM added CaCl2; (b) 60 mM added CaCl2; (c) 240 mM added CaCl2; (d) 600 mM added CaCl2.
aggregates during heating. It was considered that addition of CaCl2 resulted in the formation of more noncovalent associations, which dominated the properties of the heat-induced WP1 gels. Addition of CaCl2 to WP2 solutions resulted in the formation of noncovalent associations at the expense of disulfide linkages during heating. Addition of CaCl2 to WP3 solutions did not have any apparent effect, probably because of a limited number of sites available for calcium binding.
’ AUTHOR INFORMATION Corresponding Author
*Ingredient Functionality, Fonterra Research Centre, Private Bag 11029, Palmerston North, New Zealand. Phone: 64 6 350 4649. Fax: 64 6 350 4663. E-mail:
[email protected]. Funding Sources
The authors gratefully acknowledge the New Zealand Foundation for Research, Science and Technology for funding E.R.’s PhD scholarship.
’ ACKNOWLEDGMENT The authors gratefully acknowledge Dr. Barbara KuhnSherlock for her assistance with the statistical analyses.
’ REFERENCES (1) Kinsella, J. E.; Whitehead, D. M. Proteins in whey: chemical, physical, and functional properties. Adv. Food Nutr. Res. 1989, 33, 343–438. (2) Schmidt, R. H.; Illingworth, B. L.; Ahmed, E. M.; Richter, R. L. The effect of dialysis on heat-induced gelation of whey protein concentrate. J. Food Process. Preserv. 1978, 2, 111–120. (3) Havea, P.; Singh, H.; Creamer, L. K. Heat-induced aggregation of whey proteins: comparison of cheese WPC with acid WPC and relevance of mineral composition. J. Agric. Food Chem. 2002, 50, 4674–4681. (4) Bryant, C. M.; McClements, D. J. Molecular basis of protein functionality with special consideration of cold-set gels derived from heat-denatured whey. Trends Food Sci. Technol. 1998, 9, 143–151. (5) Hongsprabhas, P.; Barbut, S.; Marangoni, A. C. The structure of cold-set whey protein isolate gels prepared with Ca2+. Lebensm.-Wiss. Technol. 1999, 32, 196–202. (6) Simons, J.-W. F. A.; Kosters, H. A.; Visschers, R. W.; de Jongh, H. H. J. Role of calcium as trigger in thermal β-lactoglobulin aggregation. Arch. Biochem. Biophys. 2002, 406, 143–152. (7) Hongsprabhas, P.; Barbut, S. Structure-forming processes in Ca2+-induced whey protein isolate cold gelation. Int. Dairy J. 1997, 7, 827–834. (8) Kinsella, J. E.; Whitehead, D. M.; Brady, J.; Bringe, N. A. Milk proteins: possible relationships of structure and function. In Developments in Dairy Chemistry, Vol. 4, Functional Milk Proteins; Fox, P. F., Ed.; Elsevier Applied Science: New York, NY, 1989; pp 5595. 13163
dx.doi.org/10.1021/jf203610k |J. Agric. Food Chem. 2011, 59, 13156–13164
Journal of Agricultural and Food Chemistry (9) Wang, C.-H.; Damodaran, S. Thermal gelation of globular proteins: influence of protein conformation on gel strength. J. Agric. Food Chem. 1991, 39, 433–438. (10) Xiong, Y. L.; Dawson, K. A.; Wan, L. Thermal aggregation of β-lactoglobulin: effect of pH, ionic environment, and thiol reagent. J. Dairy Sci. 1993, 76, 70–77. (11) O’Kennedy, B. T.; Mounsey, J. S. The dominating effect of ionic strength on the heat-induced denaturation and aggregation of β-lactoglobulin in simulated milk ultrafiltrate. Int. Dairy J. 2009, 19, 123–128. (12) Havea, P.; Singh, H.; Creamer, L. K.; Campanella, O. H. Electrophoretic characterization of the protein products formed during heat treatment of whey protein concentrate solutions. J. Dairy Res. 1998, 65, 79–91. (13) Williams, S. (Ed.). Official Methods of Analysis, 14th ed.; Association of Official Analytical Chemists: Arlington, VA, 1984. (14) Russell, C. E.; Matthews, M. E.; Gray, I. K. A comparison of methods for the extraction of the fat from soluble whey protein concentrate powders. N. Z. J. Dairy Sci. Technol. 1980, 15, 239–244. (15) Lee, J.; Sedcole, J. R.; Pritchard, M. W. Matrix interactions in an inductively coupled argon plasma optimized for simultaneous multielement analysis by atomic emission spectrometry. Spectrochim. Acta 1986, 41B, 217–225. (16) Havea, P.; Watkinson, P.; Kuhn-Sherlock, B. Heat-induced whey protein gels: proteinprotein interactions and functional properties. J. Agric. Food Chem. 2009, 57, 1506–1512. (17) Hamann, D. D.; MacDonald, G. A. Rheology and texture properties of surimi and surimi-based foods. In Surimi Technology; Lanier, T. C., Lee, C. M., Eds.; Marcel Dekker, Inc.: New York, NY, 1992; pp 429500. (18) Truong, V. D.; Daubert, C. R. Textural characterization of cheeses using vane rheometry and torsion analysis. J. Food Sci. 2001, 66, 716–721. (19) Langton, M.; Hermansson, A.-M. Fine-stranded and particulate gels of β-lactoglobulin and whey protein at varying pH. Food Hydrocolloids 1992, 5, 523–539. (20) Elgar, D. F.; Norris, C. S.; Ayers, J. S.; Pritchard, M.; Otter, D. E.; Palmano, K. P. Simultaneous separation and quantitation of the major bovine whey proteins including proteose peptone and caseinomacropeptide by reversed-phase high-performance liquid chromatography on polystyrene-divinylbenzene. J. Chromatogr., A 2000, 878, 183–196. (21) Sherwin, C. P.; Foegeding, E. A. The effects of CaCl2 on aggregation of whey proteins. Milchwissenschaft 1997, 52, 93–96. (22) Mounsey, J. S.; O’Kennedy, B. T. Conditions limiting the influence of thioldisulphide interchange reactions on the heat-induced aggregation kinetics of β-lactoglobulin. Int. Dairy J. 2007, 17, 1034– 1042. (23) Bernal, V.; Jelen, P. Effect of calcium binding on thermal denaturation of bovine α-lactalbumin. J. Dairy Sci. 1984, 67, 2452–2454. (24) Brew, K. α-Lactalbumin. In Advanced Dairy Chemistry, Vol. 1, Proteins, 3rd ed.; Fox, P. F., McSweeney, P. L. H., Eds.; Kluwer Academic/Plenum Publishers: New York, NY, 2003; pp 387419. (25) Relkin, P. Thermal unfolding of β-lactoglobulin, α-lactalbumin, and bovine serum albumin. A thermodynamic approach. Crit. Rev. Food Sci. Nutr. 1996, 36, 565–601. (26) Veerman, C.; Baptist, H.; Sagis, L. M. C.; van der Linden, E. A new multistep Ca2+-induced cold gelation process for β-lactoglobulin. J. Agric. Food Chem. 2003, 51, 3880–3885. (27) Havea, P.; Carr, A. J.; Creamer, L. K. The roles of disulphide and non-covalent bonding in the functional properties of heat-induced whey protein gels. J. Dairy Res. 2004, 71, 330–339. (28) Mulvihill, D. M.; Kinsella, J. E. Gelation of β-lactoglobulin: effects of sodium chloride and calcium chloride on the rheological and structural properties of gels. J. Food Sci. 1988, 53, 231–236. (29) van den Berg, L.; van Vliet, T; van der Linden, E.; van Boekel, M. A. J. S.; van de Velde, F. Serum release: The hidden quality in fracturing composites. Food Hydrocolloids 2007, 21, 420–432.
ARTICLE
(30) Zirbel, F.; Kinsella, J. E. Factors affecting the rheological properties of gels made from whey protein isolate. Milchwissenschaft 1988, 43, 691–694. (31) Kuhn, P. R.; Foegeding, E. A. Mineral salt effects on whey protein gelation. J. Agric. Food Chem. 1991, 39, 1013–1016. (32) Kuhn, P. R.; Foegeding, E. A. Factors influencing whey protein gel rheology - dialysis and calcium chelation. J. Food Sci. 1991, 56, 789–791. (33) Schmidt, R. H.; Illingworth, B. L.; Deng, J. C.; Cornell, A. J. Multiple regression and response surface analysis of the effects of calcium chloride and cysteine on heat-induced whey protein gelation. J. Agric. Food Chem. 1979, 27, 529–532. (34) Verheul, M.; Roefs, S. P. F. M. Structure of whey protein gels, studied by permeability, scanning electron microscopy and rheology. Food Hydrocolloids 1998, 12, 17–24. (35) Morr, C. V.; Josephson, R. V. Effect of calcium, N-ethylmaleimide and casein upon heat-induced whey protein aggregation. J. Dairy Sci. 1968, 51, 1349–1355. (36) Parris, N.; Anema, S. G.; Singh, H.; Creamer, L. K. Aggregation of whey proteins in heated sweet whey. J. Agric. Food Chem. 1993, 41, 460–464. (37) Ju, Z. Y.; Kilara, A. Textural properties of cold-set gels induced from heat-denatured whey protein isolates. J. Food Sci. 1998, 63, 288–292.
13164
dx.doi.org/10.1021/jf203610k |J. Agric. Food Chem. 2011, 59, 13156–13164
