Advertisement

Ionic strength and buffering capacity of emulsifying salts determine denaturation and gelation temperatures of whey proteins

Open AccessPublished:July 22, 2022DOI:https://doi.org/10.3168/jds.2021-21738

      ABSTRACT

      Ionic conditions affect the denaturation and gelling of whey proteins, affecting the physical properties of foods in which proteins are used as ingredients. We comprehensively investigated the effect of the presence of commonly used emulsifying salts on the denaturation and gelling properties of concentrated solutions of β-lactoglobulin (β-LG) and whey protein isolate (WPI). The denaturation temperature in water was 73.5°C [coefficient of variation (CV) 0.49%], 71.8°C (CV 0.38%), and 69.9°C (CV 0.41%) for β-LG (14% wt/wt), β-LG (30% wt/wt), and WPI (30% wt/wt), respectively. Increasing the concentration of salts, except for sodium hexametaphosphate, resulted in a linear increase in the denaturation temperature of WPI (kosmotropic behavior) and an acceleration in its gelling rate. Sodium chloride and tartrate salts exhibited the strongest effect in protecting WPI against thermal denaturation. Despite the constant initial pH of all solutions, salts having buffering capacity (e.g., phosphate and citrate salts) prevented a decrease in pH as the temperature increased above 70°C, resulting in a decline in denaturation temperature at low salt concentrations (≤0.2 mol/g). When pH was kept constant at denaturation temperature, all salts except sodium hexametaphosphate, which exhibited chaotropic behavior, exhibited similar effects on denaturation temperature. At low salt concentration, gelation was the controlling step, occurring up to 10°C above denaturation temperature. At high salt concentration (>3% wt/wt), thermal denaturation was the controlling step, with gelation occurring immediately after. These results indicate that the ionic and buffering properties of salts added to milk will determine the native versus denatured state and gelation of whey proteins in systems subjected to high temperature, short time processing (72°C for 15 s).

      Key words

      INTRODUCTION

      β-Lactoglobulin, a major whey protein in milk, has a growing and vital role as a food ingredient (
      • Petit J.
      • Moreau A.
      • Ronse G.
      • Debreyne P.
      • Bouvier L.
      • Blanpain-Avet P.
      • Jeantet R.
      • Delaplace G.
      Role of whey components in the kinetics and thermodynamics of β-lactoglobulin unfolding and aggregation.
      ). The denaturation, aggregation, and gelling behavior of β-LG under a range of environmental conditions has received considerable attention over the past few decades (
      • Wagner J.
      • Biliaderis C.G.
      • Moschakis T.
      Whey proteins: Musings on denaturation, aggregate formation and gelation.
      ). The required temperature and heating conditions for denaturation and gelling to occur—for pure β-LG or other pure main whey proteins (α-LA and BSA) and for their mixtures as whey protein concentrates or isolates—have been investigated as a function of several variables, including pH (
      • Hoffmann M.A.M.
      • van Mil P.J.J.M.
      Heat-induced aggregation of beta-lactoglobulin as a function of pH.
      ;
      • Haug I.J.
      • Skar H.M.
      • Vegarud G.E.
      • Langsrud T.
      • Draget K.I.
      Electrostatic effects on β-lactoglobulin transitions during heat denaturation as studied by differential scanning calorimetry.
      for pure β-LG;
      • Yamasaki M.
      • Yano H.
      • Aoki K.
      Differential scanning calorimetric studies on bovine serum albumin: III. Effect of sodium dodecyl sulphate.
      ;
      • Boye J.I.
      • Alli I.
      • Ismail A.A.
      Interactions involved in the gelation of bovine serum albumin.
      ,
      • Boye J.I.
      • Alli I.
      • Ismail A.A.
      Use of differential scanning calorimetry and infrared spectroscopy in the study of thermal and structural stability of α-lactalbumin.
      ;
      • Homer S.
      • Lundin L.
      • Dunstan D.E.
      A detailed investigation of whey protein isolate solutions and gels reveals a number of novel characteristics.
      ;
      • Boye J.I.
      • Alli I.
      Thermal denaturation of mixtures of α-lactalbumin and β-lactoglobulin: A differential scanning calorimetric study.
      for pure α-LA or BSA or protein mixtures), protein concentration (
      • Qi X.L.
      • Brownlow S.
      • Holt C.
      • Sellers P.
      Thermal denaturation of β-lactoglobulin: Effect of protein concentration at pH 6.75 and 8.05.
      ), and solution ionic composition (
      • Haug I.J.
      • Skar H.M.
      • Vegarud G.E.
      • Langsrud T.
      • Draget K.I.
      Electrostatic effects on β-lactoglobulin transitions during heat denaturation as studied by differential scanning calorimetry.
      for pure β-LG;
      • Xiong Y.L.
      Influence of pH and ionic environment on thermal aggregation of whey proteins.
      ;
      • Fitzsimons S.M.
      • Mulvihill D.M.
      • Morris E.R.
      Denaturation and aggregation processes in thermal gelation of whey proteins resolved by differential scanning calorimetry.
      ;
      • Hussain R.
      • Gaiani C.
      • Jeandel C.
      • Ghanbaja J.
      • Scher J.
      Combined effect of heat treatment and ionic strength on the functionality of whey proteins.
      for pure α-LA or BSA or protein mixtures). Typical reported β-LG denaturation temperatures obtained via differential scanning calorimetry (DSC) at high protein concentrations and pH 6.5 to 7.0 range from 73.3 to 75.5°C (
      • Damodaran S.
      Influence of protein conformation on its adaptability under chaotropic conditions.
      ;
      • Verheul M.
      • Roefs S.P.F.M.
      • De Kruif K.G.
      Aggregation of β-lactoglobulin and influence of D2O.
      ;
      • Haug I.J.
      • Skar H.M.
      • Vegarud G.E.
      • Langsrud T.
      • Draget K.I.
      Electrostatic effects on β-lactoglobulin transitions during heat denaturation as studied by differential scanning calorimetry.
      ). Higher temperatures are needed to denature β-LG at acidic pH (
      • Haug I.J.
      • Skar H.M.
      • Vegarud G.E.
      • Langsrud T.
      • Draget K.I.
      Electrostatic effects on β-lactoglobulin transitions during heat denaturation as studied by differential scanning calorimetry.
      ;
      • Blake S.
      • Amin S.
      • Qi W.
      • Majumdar M.
      • Lewis E.N.
      Colloidal stability & conformational changes in β-lactoglobulin: Unfolding to self-assembly.
      ).
      The denaturation temperature of β-LG depends on the presence of salts in the solution (
      • Damodaran S.
      Influence of protein conformation on its adaptability under chaotropic conditions.
      ). Generally, at low salt concentrations, the native protein is stabilized by the salt, and denaturation temperature increases, even if the increase is different for different salts. At higher salt concentrations, different behavior is observed: whereas some salts (e.g., sodium chloride) increase the denaturation temperature with increasing salt concentration, others (e.g., sodium thiocyanate) decrease the denaturation temperature, to a point where the β-LG denatures at a lower temperature compared with water.
      Authors have linked the salt specificity on the thermal denaturation of whey proteins to the Hofmeister series (e.g.,
      • Yamasaki M.
      • Yano H.
      • Aoki K.
      Differential scanning calorimetric studies on bovine serum albumin: III. Effect of sodium dodecyl sulphate.
      ;
      • Bowland E.L.
      • Foegeding E.A.
      Effects of anions on thermally induced whey protein isolate gels.
      ).
      • Haug I.J.
      • Skar H.M.
      • Vegarud G.E.
      • Langsrud T.
      • Draget K.I.
      Electrostatic effects on β-lactoglobulin transitions during heat denaturation as studied by differential scanning calorimetry.
      and
      • Damodaran S.
      Influence of protein conformation on its adaptability under chaotropic conditions.
      classified a series of salts as chaotropic (e.g., lithium iodide, sodium thiocyanate) or kosmotropic (e.g., potassium fluoride, sodium sulfate), depending on whether the salts decreased or increased the denaturation temperature of β-LG, respectively. However,
      • Damodaran S.
      Influence of protein conformation on its adaptability under chaotropic conditions.
      , working with several sodium salts and a wide range of concentrations, showed that there was no single effect in these salts. At low salt concentrations, both kosmotropic and chaotropic salts stabilized native β-LG against thermal denaturation, suggesting the presence of a nonspecific stabilizing effect that was proportional to ionic strength. When the experiment was repeated with BSA and low salt concentrations (up to ∼0.75 M), the opposite effect was observed; that is, the highest BSA stability was observed in salts considered chaotropic and the lowest in salts considered kosmotropic.
      We were interested in how the denaturation and gelling of β-LG in highly concentrated solutions of β-LG or whey protein isolate (WPI) would be affected by the presence of emulsifying salts, as listed in the Code of Federal Regulations (21CFR133.179). These are salts normally used to improve the quality of processed cheese and to stabilize protein-rich beverages. Given that it is hard to predict how the addition of the emulsifying salts will affect denaturation, aggregation, and gelling of the protein of interest because of the variability and multiplicity of effects shown above, the aim of this study was to determine the effect of ionic strength (IS) and buffering capacity created by commonly used salts on the thermal denaturation and gelation of β-LG and WPI.

      MATERIALS AND METHODS

      Bovine β-LG (>90%, lyophilized powder, CAS#9045-23-2) was obtained from Sigma-Aldrich; WPI (∼91% protein, 60% β-LG) was obtained from Agropur; monosodium phosphate [molecular weight (MW): 119.98 Da; CAS#7558-80-7], sodium hexametaphosphate (MW: 611.77 Da; CAS#10124-56-8), and citric acid (MW: 192.12 Da; CAS#77-92-9) were obtained from Carlo Erba; dipotassium phosphate (MW: 174.18 Da, CAS#7758-11-4) was obtained from Mallinckrodt; tartaric acid (MW: 150.09 Da, CAS#87-69-4) and disodium phosphate (MW: 141.96 Da; CAS#7558-79-4) were obtained from JT Baker; and sodium hydroxide (NaOH; MW: 40.0 Da, CAS#1310-73-2), sodium citrate (MW: 294.1 Da; CAS#6132-04-3), sodium chloride (NaCl; MW: 58.44 Da, CAS#7647-14-5), potassium hydroxide (KOH; MW: 56.11 Da, CAS#1310-58-3) and monopotassium phosphate (MW: 136.09, CAS#7778-77-0) were obtained from Merck. All reagents were of analytical grade (except sodium hexametaphosphate, which was pure) and were used without further purification. No animals were used in this study, and ethical approval for the use of animals was thus deemed unnecessary.

      Sample Preparation

      Salt solutions were prepared by mixing deionized (DI) water (Nortesur), NaOH, and KOH with salts (sodium hexametaphosphate) or acids (citric acid and KOH; tartaric acid and NaOH), or by mixing acids with salts (citric acid and sodium citrate, tartaric acid and sodium potassium tartrate) or by mixing 2 salts (mono- and disodium or potassium phosphates) in proportions to render solutions with the desired final salt concentration at a pH of 6.8 ± 0.1. Salts with no buffering capacity (i.e., NaCl, sodium thiocyanate, potassium iodide) were dissolved in water only, at the desired concentration.
      Sample solutions were prepared by adding dry β-LG or WPI to water or the salt solution to the desired final concentration (not corrected for protein water or content of other small impurities). The β-LG and WPI solutions in water had a pH of approximately 6.8 and were not adjusted. Once the protein and solvent were mixed, the samples were left for 24 h at 4°C to completely dissolve and equilibrate. Sample pH was checked before use in all cases for WPI, and in a few cases for β-LG, to confirm that it remained at 6.8 (in the case of β-LG solutions that were not checked, we assumed that the pH remained close enough to 6.8 for our purposes). The IS of all solutions was calculated on a molal basis (moles of ions per kg of water).

      Protein Denaturation by DSC

      The native to denatured state of the samples was determined by DSC using a DSC-60A Plus calorimeter equipped with a TA-60WS thermal analysis workstation and software (Shimadzu). An aluminum sealed pan (Cat# 201-53090) with about 20 mg of accurately weighed protein solution was used in each test; an empty pan was used as a blank.
      Water or salt solutions (NaCl, sodium tartrate and sodium potassium tartrate, sodium and potassium citrates, sodium and potassium phosphates, and sodium hexametaphosphate) at concentrations of 3.5, 2.8, 2.1, 1.4 and 0.7 (wt/wt), and 30% (wt/wt) β-LG or WPI, or 14% (wt/wt) β-LG in a similar range of salt concentrations were analyzed. Thirty percent protein content was the highest concentration rendering a clear, liquid solution; 14% was the lowest concentration that consistently formed a gel during rheological testing.
      The reported denaturation temperatures correspond to the minima in the endothermal peaks, as determined by the software, after correction for baseline drift. Because we were dealing with complex kinetic processes, it should be noted that at the denaturation temperature, part of the protein is denatured and part is not, and that denaturation temperatures change if the experiment conditions change (as detailed below for the heating rate). The thermograms were run from 25°C to 90°C, followed by 2 min at 90°C. Table 1 shows a selection of protein samples that were dispersed in water and run for several independently prepared sample replications to demonstrate the good reproducibility of the method.
      Table 1Preliminary tests on the reproducibility of β-LG and whey protein isolate (WPI) denaturation temperature obtained by differential scanning calorimetry
      SourceConcentration (% wt/wt)No. of replicationsDenaturation temperature (°C)SD (°C)CV (%)
      β-LG301771.80.280.38
      β-LG14873.50.360.49
      WPI301569.90.290.41
      The denaturation peak temperature depends on the heating rate (
      • de Wit J.N.
      • Swinkels G.A.
      A differential scanning calorimetric study of the denaturation of bovine β-lactoglobulin.
      ;
      • Hoffmann M.A.M.
      • Van Miltenburg J.C.
      • Van Mil P.J.J.M.
      The suitability of scanning calorimetry to investigate slow irreversible protein denaturation.
      ). In a preliminary experiment using β-LG (30% wt/wt in water) at different heating rates (0.5, 1, 2, 5, 10, 15, and 20°C/min), the peak denaturation temperatures ranged from 73.3°C at a heating rate of 20°C/min to 67.8°C at a heating rate of 0.5°C/min, a temperature variation in accordance with previous reports (
      • de Wit J.N.
      • Swinkels G.A.
      A differential scanning calorimetric study of the denaturation of bovine β-lactoglobulin.
      ). To keep results comparable, afterward all thermograms were run at 10°C/min (peak temperature slightly below 72°C in the previously mentioned experiment). Indium (melting point at 156.6°C) was used for calibration.

      Effect of Temperature on pH

      For pH measurements at temperatures higher than room temperature, a beaker containing the solution was heated slowly, with stirring, on a heated plate. The temperature and pH were measured using a AD1020 pH meter (ADWA Instruments), with the probes immersed in the solution. The pH electrode was calibrated before each run. The pH meter equipment made the necessary compensation for the different temperatures.

      Rheological Properties

      A rheometer (MCR 302, Anton Paar) equipped with a Peltier heating element and concentric cylinders (bob diameter: 26.66 mm, cup diameter: 28.92 mm) was used for all measurements. An oscillatory test was conducted with amplitude of 0.1%, frequency of 0.5 Hz, and temperature from 25 to 65°C at a rate of 8°C/min, 65 to 85°C at a rate of 1.0°C/min, followed by a holding time of 20 min at 85°C. In a standard experiment, 19 mL of 14% (wt/wt) WPI in water or different salt solutions were tested. Whey protein isolate at 14% (wt/wt) was the minimum concentration that consistently yielded a gel in our experimental conditions. Viscoelastic properties were determined by monitoring the storage (G′, Pa) and loss (G″, Pa) moduli as a response to the applied dynamic deformation. The temperature of gelation was arbitrarily defined as the temperature where G′ = 1 Pa (
      • Kavanagh G.M.
      • Clark A.H.
      • Gosal W.S.
      • Ross-Murphy S.B.
      Heat-induced gelation of β-lactoglobulin/α-lactalbumin blends at pH 3 and pH 7.
      ). Table 2 shows the results of several independently prepared sample replication runs of 14% (wt/wt) WPI dispersed in water (solvent A) or a 4.3% (wt/wt) sodium phosphate solution (solvent B) to demonstrate the good reproducibility of the method.
      Table 2Reproducibility of whey protein isolate gelling temperature (i.e., temperature where storage modulus G′ = 1 Pa) under the experimental conditions of this study
      Solvent
      Solvent A = deionized water; solvent B = sodium phosphate in water (4.3% wt/wt). Whey protein isolate content was 14% (wt/wt) in all cases.
      No. of replicationsGelling temperature (°C)SD (°C)CV (%)
      A681.30.130.15
      B675.40.861.14
      1 Solvent A = deionized water; solvent B = sodium phosphate in water (4.3% wt/wt). Whey protein isolate content was 14% (wt/wt) in all cases.

      Statistical Analysis

      Linear regression was used, when needed, to describe the relationship between denaturation temperatures and salt concentration. In Table 1, Table 2, standard definitions of standard deviation and coefficient of variation were employed.

      RESULTS AND DISCUSSION

      Denaturation temperatures of β-LG, either alone or as part of WPI, using DSC have been abundantly reported in the literature. However, very few precedents of denaturation temperatures at high protein concentration and in the presence of salts are found (
      • Damodaran S.
      Influence of protein conformation on its adaptability under chaotropic conditions.
      ;
      • Qi X.L.
      • Brownlow S.
      • Holt C.
      • Sellers P.
      Thermal denaturation of β-lactoglobulin: Effect of protein concentration at pH 6.75 and 8.05.
      ;
      • Verheul M.
      • Roefs S.P.F.M.
      • De Kruif K.G.
      Aggregation of β-lactoglobulin and influence of D2O.
      ;
      • Haug I.J.
      • Skar H.M.
      • Vegarud G.E.
      • Langsrud T.
      • Draget K.I.
      Electrostatic effects on β-lactoglobulin transitions during heat denaturation as studied by differential scanning calorimetry.
      ). Figure 1 shows the denaturation temperature for solutions containing 14% (wt/wt) β-LG (Figure 1A), 30% (wt/wt) β-LG (Figure 1B), and 30% (wt/wt) WPI (Figure 1C), in the presence of different concentrations of the tested salts. The average denaturation temperature in water was 73.6°C [coefficient of variation (CV) 0.49%], 71.8°C (CV 0.38%), and 70.0°C (CV 0.41%), respectively, for 14% and 30% β-LG and 30% WPI. Figure 1A, B shows a quasi-linear denaturation temperature increase with increasing salt concentrations for most salts except those with buffering capacity. Sodium chloride and sodium or sodium potassium tartrates, added to 3.5% (wt/wt), increased the denaturation temperature of 30% β-LG and 30% WPI by 7 to 8°C and 5 to 6°C, respectively. It must be noted that denaturation, aggregation, and gelling are kinetic processes; thus, lower temperatures and longer processing times will yield equivalent net whey protein denaturation. The Code of Federal Regulations does not establish limits for the use of many of these salts in beverages (e.g., CFR21.I.B.184.1751), but the overall salt content is usually <2%. In the case of process cheese, salts can be used to a maximum of 3% (wt/wt; CFR21.I.B.133.169). Despite maximum salt content, dairy plants following Pasteurized Milk Ordinance regulations and safeguards choose to pasteurize milk at >72°C. In such cases, the protective effect of salts at low concentration against denaturation may allow for higher processing temperatures while preserving native whey proteins. Salts having buffering capacity (i.e., sodium phosphates and potassium phosphates) showed an initial reduction of 1 to 2°C in denaturation temperature of 14% β-LG, followed by a linear increase similar to that of salts having no buffering capacity. The same was observed for citrate salts but with only a 0.5°C decrease in denaturation temperature. The deviation from a linear increase in denaturation temperature observed for citrates and phosphates at low concentration is explained by different pH behavior at elevated temperatures (discussed below). We hypothesized that if all samples were prepared to have the same pH (i.e., a similar protein charge density) at the denaturation temperature, this initial lag in the protective effect of salts against thermal denaturation would not occur.
      Figure thumbnail gr1
      Figure 1Effect of salt concentration (conc.; expressed in % wt/wt) on the thermal denaturation of (A) β-LG at 14% (wt/wt), (B) β-LG at 30% (wt/wt), and (C) whey protein isolate (WPI) at 30% (wt/wt) concentration.
      When 30% WPI was analyzed (Figure 1C), lower denaturation temperatures were apparent for salt solutions having buffering capacity, compared with nonbuffering salts, but the differences were smaller than in the case of β-LG. An absolute decrease in the denaturation temperature at low salt concentration, as was the case with β-LG, was not observed. The presence of several other components (e.g., naturally occurring salts and proteins such as α-LA and BSA) probably masked, to some extent, the buffering capacity of the salts. Several mechanisms could be suggested, including better WPI buffering capacity compared with β-LG, calcium present in small amounts in the WPI interacting with the solvent salts, and BSA and (mostly) α-LA modifying the β-LG aggregation and gelling steps and indirectly modifying its denaturation temperature.
      We suggest different ways by which a salt interacts with solvent and proteins along the thermal denaturation pathway. The near linearity between denaturation temperature and salt concentration and the similar slope for all salts suggest that all salts (except sodium hexametaphosphate) interacted with the whey protein similarly and behaved as kosmotropic salts based on the Hofmeister series (
      • Zhang Y.
      • Cremer P.S.
      Interactions between macromolecules and ions: the Hofmeister series.
      ). When we tested salts from the chaotropic side of the Hofmeister series (e.g., sodium thiocyanate), we found that after an initial increase in denaturation temperature with increasing salt concentration (up to ∼2.0% wt/wt), the denaturation temperature curve turned downward, and at ∼4.5% (wt/wt), the denaturation temperature was lower than that in DI water (data not shown), in agreement with previous reports (
      • Damodaran S.
      Influence of protein conformation on its adaptability under chaotropic conditions.
      ). Potassium iodide, a less chaotropic salt, showed intermediate behavior, and after the initial usual increase in denaturation temperature with increasing salt concentration, it leveled off and maintained an almost constant denaturation temperature (at least up to ∼8% wt/wt salt concentration; data not shown). The denaturation temperature response to sodium hexametaphosphate was similar to that of potassium iodide, behaving as a pseudo-chaotropic salt.
      Figure 2 shows denaturation temperature versus IS (the portion contributed by the salt), which is a more meaningful relationship from a chemical point of view. Based on the IS, sodium chloride and the tartrate salts exhibited an important protective effect against the thermal denaturation of β-LG and WPI, and the protection against denaturation linearly increased with increasing IS. The denaturation temperature for 30% β-LG was 71.84°C in water, which increased by an estimated 8°C for every 1 mol/kg increase in IS of sodium chloride and the tartrate salts (r = 0.99; P < 0.001). As mentioned before, this is remarkable because many processing plants pasteurize milk at 74 to 75°C, exactly the point where the presence of salts may ultimately determine the native versus denatured state of the whey proteins. The citrate and phosphate salts exhibited a slightly destabilizing effect at IS ≤0.2 mol/g and then, above 0.2 mol/kg, an increase in the denaturation temperature at a rate of 3.6 to 6.0°C for every 1 mol/kg increase in IS, depending on the salt.
      Figure thumbnail gr2
      Figure 2Effect of salt ionic strength (IS) expressed on a molal basis (IS molal) on the thermal denaturation of (A) β-LG at 14% (wt/wt), (B) β-LG at 30% (wt/wt), and (C) whey protein isolate (WPI) at 30% (wt/wt) concentration.
      It is well known that a buffer pH, as in any equilibrium reaction, is temperature dependent, and also that the pH, modifying the protein charge, has a strong influence on the β-LG stability in terms of denaturation; that is, higher denaturation temperatures are observed at lower pH (pH < 6.8;
      • Verheul M.
      • Roefs S.P.F.M.
      • De Kruif K.G.
      Kinetics of heat-induced aggregation of β-lactoglobulin.
      ;
      • Haug I.J.
      • Skar H.M.
      • Vegarud G.E.
      • Langsrud T.
      • Draget K.I.
      Electrostatic effects on β-lactoglobulin transitions during heat denaturation as studied by differential scanning calorimetry.
      ). Whereas all starting solutions had a pH of 6.8, due to the pH-temperature dependence, the pH of β-LG solutions (β-LG itself having buffer capacity) decreased as the temperature increased, probably due to the 2 temperature-sensitive histidine AA residues (
      • Reijenga J.C.
      • Gagliardi L.G.
      • Kenndler E.
      Temperature dependence of acidity constants, a tool to affect separation selectivity in capillary electrophoresis.
      ;
      • Gnaiger E.
      pH measurement and temperature dependence of pH.
      ;
      • Reineke K.
      • Mathys A.
      • Knorr D.
      Shift of pH-value during thermal treatments in buffer solutions and selected foods.
      ).
      The first group of salts; that is, sodium chloride and sodium and potassium tartrates, had negligible buffering capacity at the pH of these experiments. The second group—citrates and phosphates—were good buffers at the pH tested. The phosphoric acid second proton and the citric acid third proton have an acid dissociation constant (pKa) close to 6.8. At the highest concentration used, both salts had a concentration higher than that of the lactoglobulin histidine residues (around 0.1–0.2 M vs. 0.03–0.015 M) and did not show a significant pH change when heated. As a result, even when all solutions had the same initial pH 6.8 at room temperature, their pH was different at 70 to 80°C, the temperature range when denaturation of β-LG occurred.
      Thus, the salts used in these experiments were expected to affect denaturation temperature in 2 ways: (1) the IS effect, where tested salts stabilized β-LG, in proportion to the amount added, measured through their contribution to IS, with no large differences among the tested salts (except sodium hexametaphosphate); and (2) the pH effect, where the change in temperature affected the final pH of the salt–β-LG system, depending on whether the salt had buffer capacity (no decrease in pH at 70–80°C) or did not have buffer capacity (decrease in pH at 70–80°C), affecting the stability of the protein toward thermal denaturation through the pH-related protein charge density.
      β-Lactoglobulin in DI water and sodium chloride and tartrate solutions had lower final pH when the temperature increased (pH ∼6.1 at 75°C). In the case of citrate– and phosphate–β-LG solutions, the situation was mixed: at very low salt concentration, the buffering capacity of β-LG was stronger than that of the salts, whereas at higher salt concentrations, the salt's buffering capacity became dominant. As a result, the pH at the denaturation temperature increased with the addition of salt, rapidly with the initial additions, and more slowly when more salt was added and the system was fully buffered, and from pH ∼6.1, when the β-LG was dissolved in DI water, to pH approaching 6.8 when β-LG was suspended in salts with buffering capacity (i.e., phosphate or citrate salts). Because the phosphate salts had better buffering capacity at pH 6.8 than the citrate salts under our conditions (i.e., at the same IS, phosphate salts were more concentrated on a molar basis), the pH at denaturation temperature was closer to 6.8 for phosphates at sufficient concentration.
      Because a higher pH means a lower denaturation temperature, the pH effect of the citrate and phosphate salts reduced the denaturation temperature as salt was added. This drop in denaturation temperature was steep at first and less pronounced when the buffering capacity was complete. The suggested mechanism is depicted in Figure 3, where the net effect of salt over the denaturation temperature of β-LG results from the addition of a pH effect (i.e., buffering capacity) and an IS effect.
      Figure thumbnail gr3
      Figure 3Representation of salt competing effects: ionic strength (IS) effect and buffering capacity effect, on the denaturation (Denat.) temperature of β-LG.
      To test the suggested mechanism, several pH determinations were made on several WPI and salt solutions at different temperatures (Figure 4). Whey protein isolate dispersed in water showed a strong pH decrease to pH ∼6.15 when the temperature increased to 75°C, whereas phosphate, citrate, and tartrate salts in water solutions (5% wt/wt, without WPI) showed little change in pH as the temperature increased to 75°C (a small increase in pH was observed for the citrate salt solution; Figure 4A). When WPI (10% wt/wt) was mixed with salts (Figure 4B), the tartrate–WPI dispersion (tartrates having little buffering capacity at pH 6.8) behaved similarly to the water–WPI solution, with a strong drop in pH to ∼6.25 when the temperature increased to 75°C. However, the phosphate or citrate salts and WPI solutions, having strong buffering capacity, showed a small pH decrease to 6.68 and 6.54, respectively, when the temperature was raised to 75°C. Figure 5 shows the pH change (ΔpH) when each solution containing 10% (wt/wt) WPI and different salt concentrations was heated from 25 to 75°C (ΔpH = pH 6.8 at 25°C − pH at 75°C), for sodium tartrate, sodium phosphate, and sodium citrate. As expected, the pH of the dispersions containing sodium tartrate (salts with little buffering capacity) decreased by about 0.6 pH units when heated to 75°C, regardless of the concentration of sodium tartrate. However, the ΔpH of the phosphate solutions, with strong buffering capacity, dropped sharply with increasing salt concentration, to reach a plateau once the phosphate buffer capacity was already higher than that of the β-LG. The citrate solutions, having somewhat less buffering capacity compared with the phosphate salts at the same IS, showed the same behavior but less markedly.
      Figure thumbnail gr4
      Figure 4Effect of temperature on the pH of salt solutions and whey protein isolate (WPI) dispersions. (A) WPI dispersions (10% and 2.5% wt/wt in water) and 3 salt solutions (5% wt/wt salt in water), and (B) WPI dispersion (10% wt/wt) in water and 3 salt solutions (3.57 salt wt/wt). DI = deionized.
      Figure thumbnail gr5
      Figure 5Change in pH (ΔpH = pH 6.8 at 25°C – pH at 75°C) in whey protein isolate (WPI) dispersions (10% wt/wt) containing increasing concentrations [ionic strength (IS) molal] of sodium tartrate, sodium citrate, or sodium phosphate.
      The thermal denaturation of β-LG (Figure 1, Figure 2) was consistent with the observed changes in buffering capacity of the salt–protein systems. The addition of salts with no buffering capacity (i.e., sodium chloride and tartrate salts) increased the denaturation temperature in proportion to the IS of the salt. In these salts, the IS effect (Figure 3) determined the final denaturation temperature. The addition of salts with buffering capacity to a β-LG dispersion is represented by the “combined effect” curve in Figure 3. At a low concentration of buffering salt, the pH effect was stronger than the IS effect, and the denaturation temperature decreased. Once the salt concentration was high enough so that the pH was mostly controlled by the buffering capacity of the salt, further salt addition resulted in little change in the pH. From this point on, the pH effect remained almost constant and the IS effect increased linearly, thus the denaturation temperature increased almost linearly too.
      An additional experiment was conducted to corroborate whether the effect of buffering salts promoted the denaturation of β-LG by preventing a pH decrease at denaturation temperature. β-Lactoglobulin (14% wt/wt) was prepared in sodium and potassium citrate and phosphate salt solutions with pH adjusted to 6.2 (the pH of the salt solutions before the addition of β-LG). The pH was not further adjusted after the β-LG addition. The buffering capacity of these salts would keep the pH at 6.2 once the sample was heated to 75°C, which was the estimated pH of β-LG dispersion in water when heated to ∼75°C. The overall objective of this experiment was to eliminate the “pH effect” by adjusting the pH at room temperature, at every concentration, so that the pH was 6.2 once the samples were heated to denaturation temperature (∼75°C). Figure 6 shows the results for the phosphate salts (similar results were observed with the citrate salts), with sodium chloride added for comparison. As expected, the β-LG samples dispersed in buffering salts solutions with pH previously adjusted to pH 6.2 at room temperature showed a linear increase in denaturation temperature, similar to that of β-LG samples containing the nonbuffering salts (e.g., sodium chloride and tartrates).
      Figure thumbnail gr6
      Figure 6Denaturation temperature of 14% β-LG solutions in water containing increasing concentrations [ionic strength (IS) molal] of buffering salts with initial pH set to 6.2 and 6.8. Sodium chloride is shown as an example of a salt with no buffering capacity.
      Table 3 shows the gelation temperature, arbitrarily defined as the oscillatory temperature sweep where G′ = 1 Pa, of 14% (wt/wt) WPI dispersions in water containing 4.3% (wt/wt) salt. The observed gelation temperatures were comparable to the corresponding denaturation temperatures obtained by DSC, except in the case of DI water, where the gelation temperature was much higher. Results were compatible with those of other authors working in diluted system (
      • Verheul M.
      • Roefs S.P.F.M.
      Structure of particulate whey protein gels: Effect of NaCl concentration, ph, heating temperature, and protein composition.
      ;
      • Puyol P.
      • Pérez M.D.
      • Horne D.S.
      Heat-induced gelation of whey protein isolates (WPI): Effect of NaCl and protein concentration.
      ;
      • Fitzsimons S.M.
      • Mulvihill D.M.
      • Morris E.R.
      Denaturation and aggregation processes in thermal gelation of whey proteins resolved by differential scanning calorimetry.
      ).
      Table 3Gelation temperature (temperature where storage modulus G′ > 1 Pa) and denaturation temperature (differential scanning calorimetry) of whey protein isolate (WPI) dispersed in water and water salt solutions
      SolutionGelation temperature (°C) (WPI 14%, 4.37 wt/wt salt)Denaturation temperature (°C) (WPI 30%, 3.5% wt/wt salt)Difference (°C)
      Water81.369.911.4
      Sodium phosphate74.373.80.5
      Sodium citrate74.374.30.0
      Sodium tartrate76.175.60.5
      Sodium hexametaphosphate74.273.01.2
      Sodium chloride77.277.20.0
      Potassium phosphate75.073.41.6
      Potassium citrate74.173.70.4
      Potassium, sodium tartrate75.675.60.0
      The formation of a whey protein gel is a 2-step process where denaturation occurs first, followed by aggregation and gelation (
      • Fitzsimons S.M.
      • Mulvihill D.M.
      • Morris E.R.
      Denaturation and aggregation processes in thermal gelation of whey proteins resolved by differential scanning calorimetry.
      ). Assuming that the presence of ions in solution destabilized mostly denatured protein due to their exposed hydrophobic groups, it follows that the relatively concentrated 4.3% salt solutions, as well as increasing the denaturation temperature of the protein (where the denatured protein was a “product”), decreased the protein gelling temperature (where the denatured protein was a “reagent”). As a result, in systems with high salt concentration, gelation was controlled by the first step (i.e., the denaturation of the protein), and gelation followed immediately. In contrast, in the absence of salt (WPI in water), protein–protein electrostatic repulsion made gelation occur at temperatures >10°C above the denaturation temperature. Note that the observed results were dependent on the experimental conditions. A different heating rate in the oscillatory run, for example, would change the 10°C difference between denaturation and gelation temperatures. What our experiments show is that the addition of salts increased the gelation:denaturation rate ratio.
      We hypothesized that a linear increase in the concentration of nonbuffering salts would result in a linear increase in the denaturation temperature of β-LG (i.e., in an increase in the denaturation reaction rate) followed by a comparable reduction in the gelation temperature (i.e., a decrease in the aggregation and gelation rates). The gelling rate-controlling step, as salt concentration increased, would switch from gelation to denaturation. We expected that the gelation temperature would decrease at first, as the gelation-controlling temperature decreased, reach a minimum in an intermediate stage where both denaturing and gelation phenomena had comparable rates, and then increase as the now-controlling denaturation temperature increased.
      Whey protein isolate was dispersed in solutions containing increased concentrations of sodium chloride, sodium citrate, and sodium phosphate. Each salt concentration was evaluated for denaturation temperature (DSC; WPI at 30% wt/wt) and gelation temperature (G′ > 1 Pa; WPI at 14% wt/wt). Figure 7 shows the results of overall gelation temperatures versus salt concentration for sodium chloride (Figure 7A), sodium citrate (Figure 7B), and sodium phosphate (Figure 7C). As expected, at low salt concentrations, the gelation phenomenon was the controlling step, and gelation occurred at temperature ∼10°C higher than the denaturation temperature. At high salt concentration, thermal denaturation was the controlling step, and both gelation and denaturation temperature were closer and tended to coincide, as initially hypothesized. Note that these results should be taken qualitatively, as the concentration of WPI was different due to experimental constraints, denaturation temperature was arbitrarily defined as the temperature of the DSC peak, and gelation temperature was arbitrarily defined as a given storage modulus value (1 Pa). Moreover, denaturation, aggregation, and gelling are processes that start at temperatures lower than the denaturation and gelling temperatures, as defined previously. Using the defined denaturation and gelling temperatures is a good way to compare how easily these processes reach a similar extent of reaction under different reaction conditions.
      Figure thumbnail gr7
      Figure 7Denaturation temperature of 30% whey protein isolate (WPI) and gelation temperature of 14% WPI in increasing concentrations (conc.) of (A) sodium chloride, (B) sodium citrate, and (C) sodium phosphate.
      Figure 8 shows the suggested response if pure β-LG were used in similar experiments of gelation and denaturation temperature using buffering and nonbuffering salts. In the case of the sodium chloride solutions (nonbuffering salt, Figure 8B), no qualitative difference would be expected. The absence of the other whey proteins would change, to some extent, the gelation behavior, but β-LG (the most abundant protein) is usually the protein that determines gelation in WPI (
      • Homer S.
      • Lundin L.
      • Dunstan D.E.
      A detailed investigation of whey protein isolate solutions and gels reveals a number of novel characteristics.
      ). In the case of the buffering citrate and phosphate salts (Figure 8A), an initial reduction in denaturation temperature of β-LG would be expected, but with no strong effect on gelation temperature, because at low salt concentration, gelation would be controlled by the aggregation-gelation step. Once the buffering of the system was complete, then both gelation and denaturation temperature would increase linearly to finally reach the same temperature.
      Figure thumbnail gr8
      Figure 8Suggested denaturation temperature (orange) and gelation temperature (blue) profiles for β-LG suspended in a salt solution (A) with and (B) without buffering capacity.

      CONCLUSIONS

      The addition of emulsifying salts at pH values close to neutral resulted in a significant increase in denaturation temperature and gelling rate of β-LG. Except for sodium hexametaphosphate, the near linearity of the increase in denaturation temperature and the comparable slope shown by kosmotropic-type salts suggests similar salt–protein interactions. The addition of salts with good buffering capacity showed that even if the pH is the same for all solutions at room temperature, the pH at denaturation temperature changes according to the solution buffering capacity (which is affected by the protein and the salt). Our results also demonstrated that the native versus denatured state of whey proteins in protein drinks subjected to thermal processing close to HTST (72°C for 15 s) will be largely determined by the IS and buffering capacity of the system.

      ACKNOWLEDGMENTS

      The authors acknowledge the support from Universidad Tecnológica del Uruguay (Montevideo, Uruguay). The authors have not stated any conflicts of interest.

      REFERENCES

        • Blake S.
        • Amin S.
        • Qi W.
        • Majumdar M.
        • Lewis E.N.
        Colloidal stability & conformational changes in β-lactoglobulin: Unfolding to self-assembly.
        Int. J. Mol. Sci. 2015; 16 (26247930): 17719-17733
        • Bowland E.L.
        • Foegeding E.A.
        Effects of anions on thermally induced whey protein isolate gels.
        Food Hydrocoll. 1995; 9: 47-56
        • Boye J.I.
        • Alli I.
        Thermal denaturation of mixtures of α-lactalbumin and β-lactoglobulin: A differential scanning calorimetric study.
        Food Res. Int. 2000; 33: 673-682
        • Boye J.I.
        • Alli I.
        • Ismail A.A.
        Interactions involved in the gelation of bovine serum albumin.
        J. Agric. Food Chem. 1996; 44: 996-1004
        • Boye J.I.
        • Alli I.
        • Ismail A.A.
        Use of differential scanning calorimetry and infrared spectroscopy in the study of thermal and structural stability of α-lactalbumin.
        J. Agric. Food Chem. 1997; 45: 1116-1125
        • Damodaran S.
        Influence of protein conformation on its adaptability under chaotropic conditions.
        Int. J. Biol. Macromol. 1989; 11 (2489053): 2-8
        • de Wit J.N.
        • Swinkels G.A.
        A differential scanning calorimetric study of the denaturation of bovine β-lactoglobulin.
        Biochim. Biophys. Acta. 1980; 624: 40-50
        • Fitzsimons S.M.
        • Mulvihill D.M.
        • Morris E.R.
        Denaturation and aggregation processes in thermal gelation of whey proteins resolved by differential scanning calorimetry.
        Food Hydrocoll. 2007; 21: 638-644
        • Gnaiger E.
        pH measurement and temperature dependence of pH.
        Water. 2010; 5: 1-5
        • Haug I.J.
        • Skar H.M.
        • Vegarud G.E.
        • Langsrud T.
        • Draget K.I.
        Electrostatic effects on β-lactoglobulin transitions during heat denaturation as studied by differential scanning calorimetry.
        Food Hydrocoll. 2009; 23: 2287-2293
        • Hoffmann M.A.M.
        • van Mil P.J.J.M.
        Heat-induced aggregation of beta-lactoglobulin as a function of pH.
        J. Agric. Food Chem. 1999; 47: 1898-1905
        • Hoffmann M.A.M.
        • Van Miltenburg J.C.
        • Van Mil P.J.J.M.
        The suitability of scanning calorimetry to investigate slow irreversible protein denaturation.
        Thermochim. Acta. 1997; 306: 45-49
        • Homer S.
        • Lundin L.
        • Dunstan D.E.
        A detailed investigation of whey protein isolate solutions and gels reveals a number of novel characteristics.
        Food Hydrocoll. 2018; 77: 566-576
        • Hussain R.
        • Gaiani C.
        • Jeandel C.
        • Ghanbaja J.
        • Scher J.
        Combined effect of heat treatment and ionic strength on the functionality of whey proteins.
        J. Dairy Sci. 2012; 95 (22939789): 6260-6273
        • Kavanagh G.M.
        • Clark A.H.
        • Gosal W.S.
        • Ross-Murphy S.B.
        Heat-induced gelation of β-lactoglobulin/α-lactalbumin blends at pH 3 and pH 7.
        Macromolecules. 2000; 33: 7029-7037
        • Petit J.
        • Moreau A.
        • Ronse G.
        • Debreyne P.
        • Bouvier L.
        • Blanpain-Avet P.
        • Jeantet R.
        • Delaplace G.
        Role of whey components in the kinetics and thermodynamics of β-lactoglobulin unfolding and aggregation.
        Food Bioprocess Technol. 2016; 9: 1367-1379
        • Puyol P.
        • Pérez M.D.
        • Horne D.S.
        Heat-induced gelation of whey protein isolates (WPI): Effect of NaCl and protein concentration.
        Food Hydrocoll. 2001; 15: 233-237
        • Qi X.L.
        • Brownlow S.
        • Holt C.
        • Sellers P.
        Thermal denaturation of β-lactoglobulin: Effect of protein concentration at pH 6.75 and 8.05.
        Biochim. Biophys. Acta. 1995; 1248 (7711056): 43-49
        • Reijenga J.C.
        • Gagliardi L.G.
        • Kenndler E.
        Temperature dependence of acidity constants, a tool to affect separation selectivity in capillary electrophoresis.
        J. Chromatogr. A. 2007; 1155 (17046005): 142-145
        • Reineke K.
        • Mathys A.
        • Knorr D.
        Shift of pH-value during thermal treatments in buffer solutions and selected foods.
        Int. J. Food Prop. 2011; 14: 870-881
        • Verheul M.
        • Roefs S.P.F.M.
        Structure of particulate whey protein gels: Effect of NaCl concentration, ph, heating temperature, and protein composition.
        J. Agric. Food Chem. 1998; 46: 4909-4916
        • Verheul M.
        • Roefs S.P.F.M.
        • De Kruif K.G.
        Aggregation of β-lactoglobulin and influence of D2O.
        FEBS Lett. 1998; 421 (9468321): 273-276
        • Verheul M.
        • Roefs S.P.F.M.
        • De Kruif K.G.
        Kinetics of heat-induced aggregation of β-lactoglobulin.
        J. Agric. Food Chem. 1998; 46: 896-903
        • Wagner J.
        • Biliaderis C.G.
        • Moschakis T.
        Whey proteins: Musings on denaturation, aggregate formation and gelation.
        Crit. Rev. Food Sci. Nutr. 2020; 60 (31997648): 3793-3806
        • Xiong Y.L.
        Influence of pH and ionic environment on thermal aggregation of whey proteins.
        J. Agric. Food Chem. 1992; 40: 380-384
        • Yamasaki M.
        • Yano H.
        • Aoki K.
        Differential scanning calorimetric studies on bovine serum albumin: III. Effect of sodium dodecyl sulphate.
        Int. J. Biol. Macromol. 1992; 14 (1476985): 305-312
        • Zhang Y.
        • Cremer P.S.
        Interactions between macromolecules and ions: the Hofmeister series.
        Curr. Opin. Chem. Biol. 2006; 10 (17035073): 658-663