Journal of Dairy Science
Volume 92, Issue 12 , Pages 5843-5853, December 2009

Interactions between globular proteins and procyanidins of different degrees of polymerization

  • S.V.E. Prigent

      Affiliations

    • WageningenUniversity, Laboratory of Food Chemistry, PO Box 8129, 6700 EV Wageningen, the Netherlands
    • Corresponding Author InformationCorresponding authors.
    • ,
  • A.G.J. Voragen

      Affiliations

    • WageningenUniversity, Laboratory of Food Chemistry, PO Box 8129, 6700 EV Wageningen, the Netherlands
    • ,
  • G.A. van Koningsveld

      Affiliations

    • WageningenUniversity, Laboratory of Food Chemistry, PO Box 8129, 6700 EV Wageningen, the Netherlands
    • ,
  • A. Baron

      Affiliations

    • INRA, UR117 Cidricoles et Biotransformation des Fruits et Légumes, 35653 Le Rheu, France
    • ,
  • C.M.G.C. Renard

      Affiliations

    • INRA, UMR408 Sécurité et Qualité des Produits d’Origine Végétale, Université d’Avignon, 84914 Avignon, France
    • ,
  • H. Gruppen

      Affiliations

    • WageningenUniversity, Laboratory of Food Chemistry, PO Box 8129, 6700 EV Wageningen, the Netherlands
    • Corresponding Author InformationCorresponding authors.

Received 30 March 2009; accepted 11 August 2009.

Article Outline

Abstract 

Interactions of proteins with phenolic compounds occur in food products containing vegetable sources, such as cocoa, cereals, or yogurts containing fruit. Such interactions can modify protein digestion and protein industrial properties. Noncovalent interactions between globular proteins (proteins important in industry) and procyanidins (phenolic compounds present in large quantity in fruits) were studied. The affinity constants between procyanidins of various average degrees of polymerization () and lysozyme or α-lactalbumin were measured by isothermal titration calorimetry. The effects of these interactions on protein solubility and foam properties were examined using α-lactalbumin and BSA. Weak interactions were found with epicatechin and procyanidin dimers. Procyanidins of n=5.5 and n=7.4 showed medium (1.5×105 M−1) and high (8.69×109 M−1) affinities, respectively, for α-lactalbumin at pH 5.5, with n the average number of subunits per oligomer. A positive cooperativity of binding at low procyanidin:protein molar ratios was observed. The affinities of α-lactalbumin and lysozyme for procyanidins increased when the pH was close to the isoelectric pH. Solubility of lysozyme was strongly decreased by procyanidins of n=5.5, whereas α-lactalbumin and BSA were less affected. Protein solubility in the presence of procyanidins was not affected by increased ionic strength but increased slightly with temperature. Procyanidins of n=5.5 and n=7.4 stabilized the average bubble diameter of foam formed with α-lactalbumin but had no effect on foam made from BSA. These results indicate that procyanidins of medium can lead to an undesirable decrease of protein solubility, but may play a positive role in foam stability.

Key words: protein, polyphenol, proanthocyanidin, isothermal titration calorimetry

 

Back to Article Outline

Introduction 

During the past few decades, consumers have shown increasing interest in drinkable milk and fruit juice mixes and, recently, food industries have launched probiotic yogurts with added polyphenols from wine or from green tea. Like traditional hot chocolate in Western countries and masala chai (i.e., spicy tea latte) in India, these products contain dairy proteins and phenolic compounds (PC). The effect of PC on some sensory properties of dairy products is known: they induce flavors and off-flavors in milk, cheese, and butter, and induce discoloration in cheeses (O’Connell and Fox, 2001). They can be used as antioxidants in milk (Serafini et al., 2009), as natural pigments for yogurts (Wallace and Giusti, 2008), and as an agent against pathogenic microorganisms (O’Connell and Fox, 2001). An important characteristic of condensed PC is that they are tannins; that is, they are able to interact with and precipitate proteins. Tannins have adverse effects on digestion and reduce the nutritional value of dairy beverages (Boor et al., 1986) and of forage for cattle (Burns, 1978). Protein–tannin interactions are also responsible for the perception of astringency by precipitating proline-rich proteins from saliva (Baxter et al., 1997). Furthermore, these interactions can modify food functional properties: they often result in a decrease of protein solubility (Kumar and Horigome, 1986); increase the heat stability of skimmed milk by addition of tea, cocoa, or coffee extracts; and modify flavor attributes (e.g., decrease flavor of the phenol vanillin in ice cream; O’Connell and Fox, 2001). In contrast to the effects of tannins on astringency and on nutrition, especially on protein digestion, the effect of proanthocyanidins on food functional properties has been less investigated. This investigation requires understanding of the nature and the extent of the interactions of proanthocyanidins with the important proteins in food (i.e., globular proteins).

Proanthocyanidins were the PC chosen for this study because they are common in fruits, cocoa, tea, some cereals, some legume seeds, and wine (Santos-Buelga and Scalbert, 2000). In foods, the most common proanthocyanidins are procyanidins, which are composed of units of catechin or epicatechin or both, and prodelphinidins, which are composed of units of gallocatechin or epigallocatechin or both (Santos-Buelga and Scalbert, 2000).

The protein precipitation induced by PC reaches a maximum close to or below the isoelectric point of the proteins (Hagerman and Butler, 1981; Naczk et al., 1996). The precipitation is also influenced by the degree of polymerization () of proanthocyanidins; the protein-precipitating effectiveness of mixtures of proanthocyanidins and proanthocyanidin gallates seems to increase with increasing degrees of polymerization until an optimal of 12 units of catechin (De Freitas and Mateus, 2001). Above this , the precipitating capacity decreases because of steric hindrance (De Freitas and Mateus, 2001).

Proanthocyanidins are believed to bind mainly to proteins via noncovalent interactions. Covalent bonds may also be formed at low pH (Torres and Bobet, 2001) and under oxidation conditions (Pierpoint, 1969). The precise nature of the noncovalent interactions remains unclear. Both hydrophobic interactions and hydrogen bonding have been suggested (Haslam, 1996; Hagerman et al., 1998). Relatively open proteins, such as nonstructured and proline-rich proteins, seem to have a higher affinity for proanthocyanidins than the more closed, globular proteins (Hagerman and Butler, 1981; De Freitas and Mateus, 2001).

Studies of protein interactions with proanthocyanidins including gallocatechins have been often restricted to turbidimetry, protein precipitation, and ultrafiltration measurements (Oh et al., 1980; Hagerman and Butler, 1981; Asquith and Butler, 1986; Kumar and Horigome, 1986; Artz et al., 1987; Naczk et al., 1996; De Freitas and Mateus, 2001). Using a protein precipitation assay, relatively polar PC, such as procyanidins, have been reported to interact with BSA via hydrogen bonds rather than via hydrophobic interactions (Hagerman et al., 1998). However, because protein-precipitating measurements are related to the solubility of the aggregates and not only to protein–procyanidin affinity, this technique may not be a suitable technique for measuring the real protein–procyanidin affinity. Another technique commonly used is ultrafiltration. According to ultrafiltration measurements, dimers and trimers of catechin were reported to interact mainly with BSA via hydrophobic interactions (Artz et al., 1987). However, because a decrease of solubility could prevent measurement of the true free phenolic concentrations, ultrafiltration might not be a suitable technique for characterizing the affinity and the nature of the main forces driving procyanidin–protein interactions when insolubility occurs.

In this study, isothermal titration calorimetry was used to measure the affinity between procyanidins of various degrees of polymerization and α-lactalbumin and lysozyme at various pH levels. An additional protein, BSA, was used to study the effect of procyanidins on the functional properties of protein. α-Lactalbumin, lysozyme, and BSA were chosen as models for globular protein because α-lactalbumin and lysozyme have highly similar AA sequences but a different isoelectric point (pI) and hydrophobicity. On the other hand, α-lactalbumin and BSA have similar pI values but BSA possesses a higher molecular mass than α-lactalbumin. Such comparisons of procyanidin interactions using 3 globular proteins that possess similarities and differences have not been carried out previously, apart from measuring their effects on protein precipitation capacity (Hagerman and Butler, 1981). The present study investigated the effects of these interactions on protein solubility as a function of pH, temperature, and ionic strength, and on protein foam properties.

Back to Article Outline

Materials and Methods 

Materials 

Bovine α-lactalbumin (type I, holo-α-lactalbumin, molecular weight=14.2 kDa, pI=4.2–4.5), lysozyme (from chicken egg white, molecular weight=14.3 kDa, pI=10.5–11.3), BSA (fractionated by cold alcohol precipitation and essentially fatty acid-free, molecular weight=66.4 kDa, pI=4.7–4.9) and (−)-epicatechin were purchased from Sigma Chemical Company (St. Louis, MO) and used without further purification. The BSA consisted of approximately 86% monomers, 12% noncovalent dimers, and 2% higher noncovalent oligomers in 20-mM sodium phosphate buffer, pH 7.0, as measured by gel permeation chromatography (Superdex 200, GE Healthcare, Diegem, Belgium).

Purification of Procyanidins 

A methanolic extract (320g) from apples var. ‘Jeanne Renard’ (using 640g of fresh parenchyma) and an aqueous acetone extract (8g) from apples var. ‘Marie Ménard’ (using 750g of fresh parenchyma) were obtained by solvent extraction of freeze-dried pulp from cider apples according to Guyot et al. (2001). These fractions were purified using solid-phase extraction. The Jeanne Renard fractions were further purified using normal-phase HPLC. From the eluates, a fraction enriched in dimers and several fractions of medium were collected. The dimers-enriched fraction was further purified using reversed-phase HPLC. The Marie Ménard fractions, which were obtained using solid-phase extraction, were further purified on a Toyopearl TSK HW-40 column (Tosohaas, Japan). These fractions were characterized after thiolysis on HPLC using calibration with phenolics standards, as described in Guyot et al. (2001). The freeze-dried fractions were stored in a vacuum dessicator.

Isothermal Titration Calorimetry 

The heat effects of successive additions of procyanidins to protein solutions were measured with an MCS isothermal titration calorimeter (MicroCal Inc., Northampton, MA). Solutions of procyanidins and proteins were prepared by weighing both procyanidins and buffer solutions and were degassed during 25min under vacuum. The reference cell contained degassed water. The sample cell (1.3mL) contained protein [(0.067–0.27g/L (i.e., 0.0047 to 0.0187mM)]. A 250-μL syringe was used to perform successive injections of a solution of procyanidins [3.16mM (i.e., 0.927g/L for epicatechin, 1.820g/L for n=2, 4.550g/L for n=5.5; and 6.734g/L for n=7.4)]. The initial delay was 600s before the first injection, which consisted of 2μL of procyanidin solution added to the protein solution for a duration of 5 s. Every 800 s, the syringe injected 5μL for a duration of 12.6 s for the second and the third injections and 10μL for a duration of 25.1s for the fourth until the twenty-sixth injections. To reduce experiment time length, the delay between the injections was reduced to 500 s once at least 26 procyanidin molecules per protein molecule were injected because at these ratios such delay was sufficient to get back to equilibrium. The reference offset was 50%. Samples with only protein or procyanidin were used as blanks, and the data of the sample containing only procyanidins were subtracted from the data of protein with procyanidins. Heat changes were analyzed with the use of Origin software (MicroCal Software, Inc.); the curves were fitted with an iterative modeling (according to the manufacturer protocol) to determine the apparent binding affinities (K) and the number of binding sites (N). Briefly, the MicroCal software proceeds as follows for the “two sets of interacting sites” model. A broad estimation of the number of K and N are guessed from the curves by the user and entered into the software, and then the following equations are used:

From these equations, [X] (the concentration of free procyanidins) and all Fb (the fraction of total proteins having b-bound procyanidins) are obtained. Then ΔQ(i) (the heat released from the ith injection) is calculated using the following equation and compared with the measured heat for the corresponding experimental injection:

Q=Mt Vo [F1 ΔH1+F2 (ΔH1+ΔH2) + … + Fb (ΔH1+ΔH2+ΔH3+ …… + ΔHb)], where M is the concentration of proteins in V, the working volume; Fb is the fraction of total proteins having b-bound procyanidins; and ΔH is the molar heat of ligand binding. Finally, the initial values of K and N are improved by the Levenberg-Marquardt minimization routine.

Solutions of 3.16 mM of procyanidins ( n=5.5), α-lactalbumin, and lysozyme were prepared in buffers with an ionic strength (I) of 0.023: 26 mM of sodium phosphate buffer (pH 3.0), 27mM of sodium acetate buffer (pH 5.5), and 10 mM of sodium phosphate buffer (pH 7.5). These protein solutions were diluted to concentrations of 0.0187 mM at pH 5.5 and 7.5 and 0.0047 mM at pH 3.0 and were then injected into the cell sample. The effect of the procyanidin on the interactions was studied at pH 5.5 using α-lactalbumin with (−)-epicatechin and procyanidin preparations with n=2, 5.5, and 7.4. Experiments were performed at 25°C.

Protein Content 

The nitrogen content of the supernatant was measured using the Dumas combustion method (Marcó et al., 2002) on an NA 2100 nitrogen and protein analyzer (Thermo Scientific, Rodano, Italy). After calibration with urea or methionine, the protein concentration was calculated using the nitrogen–protein conversion factors of 6.25 for α-lactalbumin, 5.29 for lysozyme, and 6.02 for BSA.

Protein Solubility 

α-Lactalbumin [0.50% (wt/vol)] and n=5.5 procyanidins [0.29% (wt/vol)] were incubated at a ratio of 5 mol of procyanidins per mol of protein in 26.6mM (I=0.023), 50mM (I=0.043), or 100mM (I=0.087) of sodium acetate buffer (pH 5.5) at 10°C, room temperature, and 40°C. After 2h of incubation, samples were homogenized by mixing; aliquots (0.15mL) were immediately collected and centrifuged for 1 min (12,700×g at 25°C) and the supernatants were analyzed for their protein contents using the Dumas combustion method. In the case of long incubation time (72h) at room temperature, 1μL of 10% (wt/vol) of sodium azide was added per mL of solution sodium, and the nitrogen content was accordingly corrected. Blanks consisted of centrifuged and noncentrifuged samples of protein without procyanidins and procyanidins without protein. The protein concentration of the supernatant of the protein blank was set at 100% and defined as 100% solubility. Each ionic strength and temperature was studied in triplicate.

α-Lactalbumin, lysozyme, and BSA [0.5% (wt/vol)], in the absence or presence of procyanidins of n=7.4, were dissolved in 0.01 M sodium phosphate buffer (pH 7.5; I=0.023). The protein solutions were subsequently adjusted to various pH levels (pH 2.0–pH 10.0) with NaOH or HCl (1 M or 6 M). After 2h, the samples were centrifuged for 15 min (12,700×g at 25°C). Each pH was studied in triplicate. The protein contents of the samples were measured as described previously and were corrected for the dilution with NaOH or HCl.

Air–Water Interfaces and Foam Properties 

The air–water interfacial and foam properties of α-lactalbumin and BSA were studied in the presence of procyanidins of n=2.0, 5.0, and 7.4 at pH 7.0, and with procyanidins of n=5.0 at pH 4.0. The ratios used were 2 and 5mol of procyanidins per mol of protein. Buffers used were 17mM sodium phosphate buffer (pH 7.0; I=0.030) and 100mM sodium acetate buffer containing 13mM of sodium chloride (pH 4.0; I=0.030).

The surface tension of samples was determined using an automatic drop tensiometer (Tracker, I.T. Concept Teclis, Longessaigne, France) according to Benjamins et al. (1996). The shapes of air bubbles in rising configuration in a protein solution [0.01% (wt/vol)] in the absence or presence of procyanidins were analyzed digitally. After 3,600s, dynamic oscillations of the area of the bubble with a period of 10s and a relative area deformation of 6.56% were applied. The elastic modulus was calculated from the changes in surface tension and surface area (Benjamins et al., 1996) using the average of 10 oscillations. Each sample was studied in duplicate or triplicate.

Foam was prepared according to the whipping method of Caessens and colleagues (Caessens et al., 1997). A protein solution [1% (wt/vol)] in the absence or presence of procyanidins was whipped for 3min at 2,500rpm for samples at pH 4.0 and at 3,500rpm for samples at pH 7.0. Next, the samples were poured into a cuvette (45×57×134mm). The mean bubble diameter (d21) was determined from images of the cuvettes taken in reflection mode via a prism (every 60s for the duration of 1h). The scheme of this experiment, set up by TNO Zeist (Bertus Dunnewind, Wim Lichtendonk, and Martin Bos), is described into detail by Wierenga (2005). Each sample was studied in duplicate.

Back to Article Outline

Results and Discussion 

Procyanidin Characterization 

From the Jeanne Renard variety, a fraction enriched in dimers and 2 fractions of medium were purified. The fraction of purified dimers consisted of 91.5% procyanidin dimers, 6.0% procyanidin oligomers, and 1.8% epicatechin and minor impurities [caffeoylquinic acid (i.e., chlorogenic acid) and phloridzin (i.e., dihydrochalcones)] (Table 1). The 2 fractions of medium contained only epicatechins and procyanidins and had n=5.0 and 5.5. The purified Marie Ménard fraction contained only epicatechins and procyanidins and had n=7.4 (Table 1). As expected for extracts of cider apple (Sanoner et al., 1999), the procyanidin fractions mainly consisted of (−)-epicatechin units and the dimeric fraction mainly consisted of procyanidin B2 (epicatechin-(4β→8)-epicatechin).

Table 1. Weight and composition of the purified phenolic fractions
Jeanne Renard (methanolic extract)Marie Ménard (acetonic extract)
ItemFraction 1Fraction 2Fraction 3Fraction 4
Fraction weight (% of the fresh parenchyma weight)0.0110.0510.0350.111
2.05.05.57.4
n1
Composition (% of the fraction)
Epicatechins, catechins, and procyanidins99.3100100100
Caffeoylquinic acid0.6ND2NDND
Phloridzin<0.2NDNDND

1Average degree of polymerization.

2ND=not detectable.

Effect of Procyanidin Size on Protein–Procyanidin Interactions 

Protein–procyanidin interactions were studied using isothermal titration calorimetry by measuring the heat released by procyanidin additions to solutions of α-lactalbumin and lysozyme. The biphasic shape of the curves indicated the existence of at least 2 sets of sites. A “two sets of interacting sites” model could fit the data, and not a “two independent sites” model. Such model takes into account interactivity between binding sites (i.e., it permits cooperative effects and allows protein structural changes). When procyanidins of n=5.5 and 7.4 were injected to a cell containing only buffer, a positive enthalpy change (ΔHobs) was measured, reaching a maximum of about 300 J per mol of procyanidins (data not shown). Because procyanidins are known to be able to aggregate with each other, presumably via hydrophobic interactions (Riou et al., 2002), such a positive (endothermic) enthalpy change upon dilution into the buffer-containing cell may indicate disaggregation of PC, as also proposed for hydrolysable tannins by Frazier et al. (2003). Enthalpy changes caused by protein–procyanidin interactions were corrected for this positive enthalpy.

The titration curves showing the enthalpy changes upon titration of α-lactalbumin with epicatechin, procyanidin dimers, and procyanidins of n=5.5 and 7.4 are presented in Figure 1. Procyanidins of n=5.5 and 7.4 induced enthalpy changes, which occurred in 2 stages (Figure 1). In the first stage (i.e., at ratios lower than about 5 mol of procyanidins/mol of protein), the enthalpy change became more and more negative with each addition of procyanidins, indicating that more and more energy was released (Figure 1). This first stage represents the strongest binding sites. In the second stage of protein–procyanidin interactions, the exothermic effect decreased until reaching a plateau, indicating saturation of the protein (Figure 1).

  • View full-size image.
  • Figure 1. 

    Isothermal titration calorimetry (25°C, pH 5.5) of α-lactalbumin titrated with procyanidins of various average degrees of polymerization ( n): epicatechin (×); procyanidin dimer (Δ); procyanidins of n=5.5 (○); procyanidins of n=7.4 (▴).

The K and N values obtained from the titration curves are presented in Table 2. The affinities corresponding to the first stage of interactions (i.e., at low molar ratios) are not presented because too few data points were available to obtain a satisfactory curve fitting; however, the strength and the N values of the second stage reflected the state of the first stage (e.g., a high N in the second stage was correlated with a high N in the first stage). α-Lactalbumin had a medium affinity for procyanidins of n=5.5 (1.105 M−1 at pH 5.5) and a high affinity for procyanidins of n=7.4 (9.109 M−1 at pH 5.5). The sigmoidal curve for n=7.4 reveals a specific interaction with α-lactalbumin. In contrast, the affinities for epicatechin and procyanidin dimers were aspecific and were too low to be quantified. To compare these values with other macromolecule–ligand interactions, one can notice that the affinity between α-lactalbumin and procyanidins of n=7.4 is in the same range of magnitude as the one between nicotine and neuronal receptors (2.109 M−1; Flammia et al., 1999), whereas α-lactalbumin binds Ca2+ with 1 binding site of 3.106 M−1 and 1 binding site of 3.104 M−1 (Kronman et al., 1981). The very low affinity of globular proteins for monomers and small procyanidins is in agreement with studies in which BSA aspecifically interacted with epicatechin with an affinity <103 M−1 at pH 5.0, as measured by capillary electrophoresis (Papadopoulou and Frazier, 2004) and an affinity of 3.103 M−1 for procyanidin dimers at pH 7.0, according to ultrafiltration measurements (Artz et al., 1987). The higher ability of large procyanidins to bind to proteins compared with small procyanidins may be explained by their multidentate character, which allows them to simultaneously bind several protein sites (Haslam, 1996). However, whereas affinity and specificity increase with larger procyanidins, fewer sites are available for these procyanidins (probably because of steric hindrance; Table 2). This means that, at high procyanidin:protein ratio, procyanidins of medium might affect protein solubility as much (or maybe even more) than procyanidins of high .

Table 2. Apparent binding affinity (K) constant and number of binding sites per protein molecule (N) of the second stage of interactions between proteins and procyanidins of various degrees of polymerization (; n=2, 5.5, and 7.4) and epicatechin ( n=1)1
Lysozymeα-Lactalbumin
pH 3.0pH 5.5pH 7.5pH 3.0pH 5.5pH 7.5
Item5.55.55.55.5125.57.45.5
K (M−1)<1×1035×105±2×1059×105±3×105ND2ND3ND31.5×105 ± 0.5×1058.7×109 ± 1×102ND3
NND324.3±0.411.6±0.6ND3ND3ND319.4±0.611±3ND3

1The SD correspond to the discrepancy between experimental data points and the theoretical fitting of the curve.

2Not determined because of an insufficient number of data points for curve fitting.

3Not determined because of a low slope (very low affinity).

At low molar procyanidin:protein ratios (lower than 5), the changes in enthalpy became more negative with increasing molar procyanidin:protein ratios; that is, once 1 procyanidin molecule is bound to the protein, the affinity of the protein to bind another procyanidin molecule increases. The mechanism, therefore, shows positive cooperativity, as has also been observed for the interactions between gelatin and gallotannins and ellagitannins (Frazier et al., 2003). A cooperativity mechanism may be the result of a change in protein conformation upon ligand binding (Fersht, 1999; e.g., binding of a ligand on one site could unmask a neighboring site). Whether procyanidins affect only conformational freedom (e.g., favoring one conformation) or modify secondary structure of globular proteins remains to be clarified.

Effect of pH on Protein–Procyanidin Interactions 

Titration curves showing the enthalpy changes with procyanidins of n=5.5 at pH 3.0, 5.5, and 7.5 are presented in Figure 2a for α-lactalbumin and in Figure 2b for lysozyme. As observed previously, the changes in enthalpy occurred in 2 stages.

  • View full-size image.
  • Figure 2. 

    Isothermal titration calorimetry (25°C) of proteins titrated with procyanidins of an average degree of polymerization of 5.5 at different pH values for (a) α-lactalbumin and (b) lysozyme: pH 3.0 (); pH 5.5 (○); pH 7.5 (■).

When α-lactalbumin was titrated with procyanidins, the exothermic effect, and therefore the affinity, at low ratios increased upon lowering the pH (Figure 2a; Table 2). In contrast, the affinities between lysozyme and procyanidins become stronger at higher pH levels, with a specific binding observed at pH 7.5 (Figure 2b; Table 2). These 2 observations are in agreement with Naczk et al. (1996), who have found that the optimum pH for complex formation generally is 0.3 to 3 units below the pI of the protein. However, the higher affinity at high pH corresponded to fewer binding sites in lysozyme (Figure 2b; Table 2).

To reach saturation of α-lactalbumin by procyanidins, the protein concentration used at pH 3.0 was 4 times lower than at higher pH levels. Because more phenolics are required to precipitate protein in diluted protein solutions than in concentrated protein solutions (Haslam, 1996), the affinity observed at pH 3.0 is likely to be underestimated compared with the affinity measured at higher pH levels.

Effect of Procyanidins as a Function of Protein Type 

Protein solubility was studied at a procyanidin:protein molar ratio (i.e., a ratio at which the proteins were not saturated) of 5 (Figure 3). Protein solubility decreased upon the addition of procyanidins (Figure 3). The solubility of BSA decreased by a maximum of 50% and only a narrow pH range (pH 4.0–6.0) was affected (Figure 3). The most drastic effects were obtained for α-lactalbumin and lysozyme: only 0 to 10% of α-lactalbumin remained soluble between pH 3.0 and pH 6.0 and lysozyme became totally insoluble at pH ≤ 6.0. Because the solubility of the complexes can differ as a function of various factors (e.g., size of the aggregates, nature of the protein, and type of the interactions), isothermal titration calorimetry was chosen to clarify pH effect on the interactions. This was performed on α-lactalbumin and lysozyme, which possess very similar sequences. At pH 5.5 and pH 7.5, lysozyme had a higher affinity for procyanidins than α-lactalbumin and, at pH 5.5, possessed more binding sites per protein molecule for n=5.5 than α-lactalbumin (24 vs. 19 binding sites; Figure 2; Table 2). At these pH levels, lysozyme (pI=10.5–11.3) and α-lactalbumin (pI=4.2–4.5) are positively and negatively charged, respectively, whereas procyanidins are weakly acidic (Vernhet et al., 1996); therefore, electrostatic interactions cannot directly explain the stronger affinity of procyanidins for lysozyme than for α-lactalbumin. Because lysozyme has a much lower hydrophobicity value than α-lactalbumin (Li-Chan, 1990), this contributes to the identification of hydrogen bonds as the main forces driving the interactions between procyanidins and globular proteins rather than hydrophobic interactions. The involvement of hydrogen bonds was proposed for epicatechin–BSA interactions by Frazier et al. (2006). As proposed by Haslam (2006) for interactions of PC with proline-rich proteins, it is possible that, at least when specific binding is observed, initial association is driven by hydrophobic interactions in hydrophobic regions and later enhanced by hydrogen bonds.

  • View full-size image.
  • Figure 3. 

    Protein [0.5% (wt/vol)] solubility as a function of pH, in the absence or presence of 5mol of procyanidins per mol of protein: α-lactalbumin (■)), lysozyme (•), and BSA (▴) in the absence of procyanidins; α-lactalbumin (□), lysozyme (○), and BSA (▵) in the presence of procyanidins of an average degree of polymerization of 7.4.

Procyanidins decrease the solubility of BSA to a lesser extent than the solubility of α-lactalbumin, which possesses a pI similar to that of BSA (Figure 3). However, BSA has a higher affinity for proanthocyanidins than α-lactalbumin (Hagerman and Butler, 1981). Therefore, the changes in solubility cannot entirely be ascribed to the strength of the interactions between proteins and procyanidins. To reach insolubility, it is likely that many more procyanidin molecules are required to mask hydrophilic binding sites of the high-molecular weight BSA than the low-molecular weight α-lactalbumin. This is in agreement with the fact that BSA can bind up to 178 molecules of tara tannins (i.e., mainly ellagitannins) at pH 5.0 as observed by Frazier et al. (2006), whereas α-lactalbumin was saturated by 19 molecules of procyanidins of n=5.5, which possess similar molecular weight to tara tannins (pH 5.5; Table 1).

Effect of Procyanidins on Solubility as a Function of Ionic Strength and Temperature 

No differences were observed after incubation of α-lactalbumin with procyanidins of n=5.5 mixtures at 3 ionic strengths (Table 3). After 2h, the solubility of α-lactalbumin was decreased to 18 to 21% by the presence of procyanidins (from 95–100% in the absence of procyanidins). Even after 3 d of incubation, increasing the ionic strength had no effect on the solubility of α-lactalbumin in the presence of procyanidins. Such a long incubation time was tested because it has been observed in apple juice that a higher ionic strength decreases protein solubility after several days of incubation (Tajchakavit et al., 2001). A first step of polymerization or activation (i.e., oxidation) of polyphenols, before reaction with nitrogenous material, has been proposed to explain this lag phase in the development of haze in beverages (Siebert, 2006). No lag for polyphenol–protein interactions is observed in our model, probably because this model does not contain any molecules (e.g., carbohydrates) that slow down protein–polyphenol interactions (Mateus et al., 2004).

Table 3. Protein solubility (%)1 of α-lactalbumin in the presence of procyanidins of an average degree of polymerization (; n=5.5) as a function of temperature and ionic strength
10°C25°C40°C
Buffer molarity2h2h72h2h
0.02716±120±221±527±1
0.050ND218±118±4ND
0.100ND20±118±1ND

1Percentage as a function of protein concentration (defined as 100% in the absence of procyanidins).

2ND=not determined.

Increasing the temperature to 40°C had no effect on α-lactalbumin solubility in the absence of procyanidins (as expected at this protein concentration). When α-lactalbumin was incubated with procyanidins of n=5.5 for 2h at 10, 25, or 40°C, its solubility increased with temperature from 16% at 10°C to 27% at 40°C (Table 3). This difference could indicate that α-lactalbumin interacts less strongly with procyanidins at higher temperatures. Because hydrophobic interactions increase with higher temperature, whereas hydrophilic interactions decrease, this result would indicate that hydrophilic interactions are the main driving force in α-lactalbumin–procyanidin interactions. However, because only solubility was measured and not direct interactions, another possible explanation may be that the solubility of the aggregates increases with temperature.

Effect of Procyanidins on Air–Water Interface and Foam 

The low solubility of lysozyme at pH 7.0 in the presence of procyanidins prevented foam properties from being studied with lysozyme. The effects of procyanidins on air–water interface properties and foam properties of α-lactalbumin and BSA were studied at pH 4.0 and 7.0, with 2 or 5mol of procyanidins per mol of protein. Proteins were whipped at lower speed at pH 4.0 than at pH 7.0 because of their higher foamability at the lower pH. At both pH levels, the presence of procyanidins did not modify the initial foams; the average bubble diameter of foam was not affected (data not shown). Procyanidins also had no effect on surface tension and elasticity at air–water interfaces (data not shown). However, the kinetics of foam disappearance were modified for α-lactalbumin. In the absence of procyanidins or in the presence of low- procyanidins ( n=2), the total number of bubbles and the average bubble diameter decreased over time (Figure 4). This indicates an overall coarsening of the foam as a result of coalescence of larger bubbles. On the contrary, procyanidins of n=5.0 and 7.4 were able to provide a rather constant average bubble diameter over time, especially at the high procyanidin:protein molar ratio (5mol of procyanidin:protein; Figure 4b). This extrastabilization of foam is even clearer at pH 4.0, as presented in Figures 5 and 6. The effect of medium- procyanidins on foam stability presumably is an effect against Ostwald ripening. In our system, low- procyanidins did not seem to possess this property. However, catechin improves foamability and volume stability of protein foam, possibly because of the cross-linking of protein molecules by catechin (pH 7.0; Sarker et al., 1995). Such a mechanism was optimal at a molar catechin:protein ratio of 0.1. The fact that catechin improved foam properties (Sarker et al., 1995), whereas procyanidin dimers did not, can be ascribed to the molar ratio that was used in the former study. But it can also be ascribed to the nature of the protein–procyanidin couple studied or to the presence of Tween 20 in the study of Sarker et al. (1995).

  • View full-size image.
  • Figure 4. 

    Cumulative number of bubbles as a function of bubble diameter of foam from α-lactalbumin at pH 7.0, after 15 and 49min of foam formation, in the absence (—) or in the presence of procyanidins of various degrees of polymerization (): n=2.0 (…); n=5.0 (— … —); and n=7.4 (– –). (a) Two moles of procyanidin/protein; (b) five moles of procyanidin/protein.

  • View full-size image.
  • Figure 5. 

    Images of foam from α-lactalbumin in the absence (left) and in the presence (right) of 2mol of procyanidins (with an average degree of polymerization of 5.0) per mol of α-lactalbumin at pH 4.0 (after 49min of foam formation).

  • View full-size image.
  • Figure 6. 

    Cumulative number of bubbles as a function of bubble diameter, in the absence (—) or in the presence (– –) of 2mol of procyanidins (with an average degree of polymerization of 5.0) per mol of α-lactalbumin at pH 4.0 (after 15 and 49min of foam formation).

Back to Article Outline

Conclusions 

Procyanidins of medium were able to stabilize protein foam, whereas those of smaller had no effect on protein foam properties. This can be related to the higher affinity of protein for procyanidins of medium than for smaller . Procyanidins of medium can decrease the solubility of proteins but may have a positive role in foam stability. In contrast to noncovalent interactions with hydroxycinnamic acids (i.e., PC that are present in large quantities in food and that interact poorly with globular proteins; Prigent et al., 2003), noncovalent interactions with oligomeric procyanidins affect protein functional properties. By modulating these interactions, food products may be created or improved, especially in the dairy industry.

Back to Article Outline

Acknowledgments 

This research was supported by the Ministry of Economic Affairs through the program IOP-Industrial Proteins (project IIE980019A), Senter, The Hague, the Netherlands. We thank Nathalie Marnet and Sylvain Guyot (INRA, Le Rheu, France) for their advice and help with procyanidin purification, and Martin Bos, Bertus Dunnewind, and Wim Lichtendonk (TNO Nutrition and Food Research, Zeist, the Netherlands) for their advice and work on foam experiments.

Back to Article Outline

Supplementary data 

Download file

download text

Interpretive summary.

Back to Article Outline

References 

  1. Artz WE, Bishop PD, Dunker AK, Schanus EG, Swanson BG. Interaction of synthetic proanthocyanidin dimer and trimer with bovine serum albumin and purified bean globulin fraction G-1. J. Agric. Food Chem. 1987;35:417–421
  2. Asquith TN, Butler LG. Interactions of condensed tannins with selected proteins. Phytochemistry. 1986;25:1591–1593
  3. Baxter NJ, Lilley TH, Haslam E, Williamson MP. Multiple interactions between polyphenols and a salivary proline-rich protein repeat result in complexation and precipitation. Biochemistry. 1997;36:5566–5577
  4. Benjamins J, Cagna A, Lucassens-Reynders EH. Viscoelastic properties of triacylglycerol/water interfaces covered by proteins. Colloids Surf. A Physicochem. Eng. Asp. 1996;20:245–254
  5. Boor KJ, Amundson CH, Brown DL. Protein conversion efficiencies of four test diets based on milk, two milk and tea treatments, and casein. J. Dairy Sci. 1986;69:979–982
  6. Burns JC. Antiquality factors as related to forage quality. J. Dairy Sci. 1978;61:1809–1820
  7. Caessens PWJR, Gruppen H, Visser S, Van Aken GA, Voragen AGJ. Plasmin hydrolysis of β-casein: Foaming and emulsifying properties of the fractionated hydrolysate. J. Agric. Food Chem. 1997;45:2935–2941
  8. De Freitas V, Mateus N. Nephelometric study of salivary protein-tannin aggregates. J. Sci. Food Agric. 2001;82:113–119
  9. Fersht A. Structure and Mechanism in Protein Science: A Guide to Enzyme Catalysis and Protein Folding. New York, NY: W.H. Freeman & Co.; 1999;
  10. Flammia D, Dukat M, Damaj MI, Martin B, Glennon RA. Structure−affinity investigation of nicotinic acetylcholinergic receptor binding. J. Med. Chem. 1999;42:3726–3731
  11. Frazier RA, Papadopoulou A, Green RJ. Isothermal titration calorimetry study of epicatechin binding to serum albumin. J. Pharm. Biomed. Anal. 2006;41:1602–1605
  12. Frazier RA, Papadopoulou A, Mueller-Harvey I, Kissoon D, Green RJ. Probing protein-tannin interactions by isothermal titration microcalorimetry. J. Agric. Food Chem. 2003;51:5189–5195
  13. Guyot S, Marnet N, Drilleau J-F. Thiolysis-HPLC characterization of apple procyanidins covering a large range of polymerization states. J. Agric. Food Chem. 2001;49:14–20
  14. Hagerman AE, Butler LG. The specificity of proanthocyanidin-protein interactions. J. Biol. Chem. 1981;256:4494–4497
  15. Hagerman AE, Rice ME, Ritchard NT. Mechanisms of protein precipitation for two tannins, pentagalloyl glucose and epicatechin16 (4→8) catechin (procyanidin). J. Agric. Food Chem. 1998;46:2590–2595
  16. Haslam E. Natural polyphenols (vegetable tannins) as drugs: Possible modes of action. J. Nat. Prod. 1996;59:205–215
  17. Kronman MJ, Sinha SK, Brew K. Characteristics of the binding of Ca2+ and other divalent metal ions to bovine α-lactalbumin. J. Biol. Chem. 1981;256:8582–8587
  18. Kumar R, Horigome T. Fractionation, characterization, and protein-precipitating capacity of the condensed tannins from Robinia pseudo acacia L. leaves. J. Agric. Food Chem. 1986;34:487–489
  19. Li-Chan E. Hydrophobicity in food protein systems. In:  Hui YH editors. Encyclopedia of Food Science and Technology. New York, NY: Wiley-Interscience; 1990;p. 1429–1439
  20. Marcó A, Rubio R, Compañó R, Casals I. Comparison of the Kjeldahl method and a combustion method for total nitrogen determination in animal feed. Talanta. 2002;57:1019–1026
  21. Mateus N, Carvalho E, Luis C, de Freitas V. Influence of the tannin structure on the disruption effect of carbohydrates on protein-tannin aggregates. Anal. Chim. Acta. 2004;513:135–140
  22. Naczk M, Oickle D, Pink D, Shahidi F. Protein precipitating capacity of crude canola tannins: Effect of pH, tannin, and protein concentrations. J. Agric. Food Chem. 1996;44:2144–2148
  23. O’Connell JE, Fox PF. Significance and applications of phenolic compounds in the production and quality of milk and dairy products: A review. Int. Dairy J. 2001;11:103–120
  24. Oh HI, Hoff JE, Armstrong GS, Haff LA. Hydrophobic interaction in tannin-protein complexes. J. Agric. Food Chem. 1980;28:394–398
  25. Papadopoulou A, Frazier RA. Characterization of protein-polyphenol interactions. Trends Food Sci. Technol. 2004;15:186–190
  26. Pierpoint WS. o-Quinones formed in plant extracts. Their reaction with bovine serum albumin. Biochem. J. 1969;112:619–629
  27. Prigent SVE, Gruppen H, Visser AJWG, van Koningsveld GA, de Jong GAH, Voragen AGJ. Effects of non-covalent interactions with 5-O-caffeoylquinic acid (chlorogenic acid) on the heat denaturation and solubility of globular proteins. J. Agric. Food Chem. 2003;51:5088–5095
  28. Riou V, Vernhet A, Doco T, Moutounet M. Aggregation of grape seed tannins in model wine—Effect of wine polysaccharides. Food Hydrocoll. 2002;16:17–23
  29. Sanoner P, Guyot S, Marnet N, Molle D, Drilleau J-F. Polyphenol profiles of French cider apple varieties (Malus domestica sp.). J. Agric. Food Chem. 1999;47:4847–4853
  30. Santos-Buelga C, Scalbert A. Review. Proanthocyanidins and tannin-like compounds—Nature, occurrence, dietary intake and effects on nutrition and health. J. Sci. Food Agric. 2000;80:1094–1117
  31. Sarker DK, Wilde PJ, Clark DC. Control of surfactant-induced destabilization of foams through polyphenol-mediated protein-protein interactions. J. Agric. Food Chem. 1995;43:295–300
  32. Serafini M, Testa MF, Villaño D, Pecorari M, van Wieren K, Azzini E, et al. Antioxidant activity of blueberry fruit is impaired by association with milk. Free Radic. Biol. Med. 2009;46:769–774
  33. Siebert KJ. Haze formation in beverages. Lebenson. Wiss. Technol. 2006;39:987–994
  34. Tajchakavit S, Boye JL, Bélanger D, Couture R. Kinetics of haze formation and factors influencing the development of haze in clarified apple juice. Food Res. Int. 2001;34:431–440
  35. Torres JL, Bobet R. New flavanol derivatives from grape (Vitis vinifera) byproducts. Antioxidant aminoethylthio-flavan-3-ol conjugates from a polymeric waste fraction used as a source of flavanols. J. Agric. Food Chem. 2001;49:4627–4634
  36. Vernhet A, Pellerin P, Prieur C, Osmianski J, Moutounet M. Charge properties of some grape and wine polysaccharide and polyphenolic fractions. Am. J. Enol. Vitic. 1996;47:25–30
  37. Wallace TC, Giusti MM. Determination of color, pigment, and phenolic stability in yogurt systems colored with nonacylated anthocyanins from Berberis boliviana L. as compared to other natural/synthetic colorants. J. Food Sci. 2008;73:241–248
  38. Wierenga, P. A. 2005. Basics of Macroscopic Properties of Adsorbed Protein Layers, Formed at Air-Water Interfaces, Based on Molecular Parameters. PhD dissertation, Wageningen University. Available http://library.wur.nl ISBN 90-8504-327-1.

PII: S0022-0302(09)71303-7

doi:10.3168/jds.2009-2261

Journal of Dairy Science
Volume 92, Issue 12 , Pages 5843-5853, December 2009

Article Tools

* Image Usage & Resolution