Interfacial and foaming properties of bovine β-lactoglobulin: Galactose Maillard conjugates

1 Interfacial and foaming properties of bovine -lactoglobulin:galactose 1 Maillard conjugates 2 1Marta Corzo-Martínez, 2Cecilio Carrera Sánchez, 1F. Javier Moreno, 2Juan M. 3 Rodríguez Patino and 1Mar Villamiel* 4 5 6 1Instituto de Investigación en Ciencias de la Alimentación, CIAL (CSIC-UAM), 7 Nicolás Cabrera 9, Campus de la Universidad Autónoma de Madrid, 28049, Madrid, 8 Spain. 9 2Departamento de Ingeniería Química, Facultad de Química, Universidad de Sevilla, 10 Prof. García González 1, 41012, Sevilla, Spain. 11 12 13 14 15 16 17 18 *Author to who correspondence should be addressed 19 Tel +34 910017951; Fax +34 910017905 20 E-mail: [email protected]s 21 Current address: Instituto de Investigación en Ciencias de la Alimentación, CIAL 22 (CSIC-UAM), Nicolás Cabrera 9, 28049, Madrid, Spain. 23 24 25 *Manuscript Click here to view linked References 2 Abstract 26 In this paper, the effect of the initial and advanced steps of glycosylation by 27 Maillard reaction (MR) (glycation) of -lactoglobulin (-Lg) with galactose on the 28 interfacial and foaming (foamability and foam stability) properties of this protein has 29 been studied at both pH 7 and pH 5. Hardly any effect of glycation was observed at pH 30 7. However, a pH 5, due to its increased solubility, -Lg glycated at 50°C during 48 h 31 (advanced steps of MR) presented the best dynamic of adsorption which lead to an 32 increase of the surface dilatational modulus of adsorbed film. This resulted in a better 33 foaming capacity, as well as higher stability of foams of -Lg glycoconjugates with 34 respect to native and control heated protein. These results could extend the applicability 35 of -Lg as a foaming agent, particularly in acid foods. 36 37 38 3 1. Introduction 39 Foaming characteristics of food dispersions are important in determining quality 40 attributes of many foods (milk, meat, mayonnaise, spreads, ice cream, frozen desserts, 41 cakes, breads, whipped toppings, etc.). The structure of many of these products depends 42 upon the formation and stability of foam which facilitates mixing, imparts structure and 43 contributes to sensory qualities. These dispersions are thermodynamically unstable, and 44 their relative stability depends on the properties of the surface-active components in the 45 system (Carrera & Rodríguez Patino, 2005; Rodríguez Patino, Carrera & Rodríguez 46 Niño, 2008). 47 In the food industry, foams are stabilized mainly by proteins (Rullier, Novales, & 48 Axelos, 2008), milk proteins being one of the most utilized. In particular -49 lactoglobulin (-Lg), which represents 50% of the total mass of the whey proteins, is 50 widely used due to its high capacity to be adsorbed at the air/water interface, to decrease 51 surface tension and to build interfacial elastic networks after unfolding (Kinsella, 1984; 52 Phillips, Whitehead, & Kinsella, 1994; Murray, 1998). This protein is known to form 53 thick interfacial layers close to its isoelectric point (pI 5.2) (Kinsella, 1984; Phillips et 54 al., 1994; Wilde & Clark, 1996) and, under heat treatment, a very strong aggregation at 55 pH close to pI can be produced. Thus, the formation of covalently bound protein 56 aggregates through disulphide bridges (Schmitt et al., 2005) might alter the foaming 57 properties of protein. However, at neutral pH it has been shown that partial unfolding of 58 -Lg through heat treatment improves its foaming properties (Bals & Kulozik, 2003; 59 Davis & Foegeding, 2004; Kim, Cornec, & Narsimham, 2005). In this context, the 60 search for processes that can efficiently improve the functional properties of proteins 61 and therefore increase their degree of applicability is of increasing interest. 62 4 Among the different physical, chemical, or enzymatic treatments, leading to the 63 modification of protein functionality, a great deal of attention has been focussed on the 64 covalent interaction protein/carbohydrate via the Maillard reaction (MR). During this 65 reaction, the conjugation of a reducing carbohydrate to the -amino group of lysine 66 occurs spontaneously under heating conditions without the utilization of toxic chemical 67 products (Chevalier, Chobert, Dalgalarrondo, & Haertlè, 2001a). Moreover, it is well-68 known that the Maillard reaction, carried out under dry state and well controlled 69 conditions (temperature, relative humidity and time), is an adequate method for 70 improving functionality of proteins without important structural changes (Morgan, 71 Leonil, Molle, & Bouhallab, 1997; Oliver, Melton, & Stanley, 2006a; Oliver, 2011). 72 Several studies have shown that glycation under controlled conditions, in addition to 73 improve the heat stability of food proteins, including whey proteins, favours the protein 74 diffusion at the air/water interface and its adsorption to the same, especially due to an 75 increase in exposed hydrophobicity and molecular unfolding, improving the protein 76 ability to form and stabilize foams (Schmitt, Bovay, & Frossard, 2005; Medrano, 77 Abirached, Panizzolo, Moyna, & Anon, 2009). In this sense, the study of glycosylation 78 via the MR (glycation) of -Lg as a tool to improve its foaming and stabilizing capacity, 79 particularly at pH values close to its pI, could be of interest. 80 Several authors have described a direct relationship between the foam formation 81 and stability and the interfacial properties of adsorbed protein films (Martin, Grolle, 82 Bos, Cohen-Stuart, & van Vliet, 2002; Murray, 2002; Rouimi, Schorsch, Valentini, & 83 Vaslin, 2005; Rodríguez Patino et al., 2008). Among them, the dynamic of adsorption 84 and the rheological properties of interfacial films have been shown to influence foam 85 properties, depending on the mechanisms causing foam destabilization (Baeza, Carrera, 86 Rodríguez Patino, & Pilosof, 2005; Rodríguez Patino et al., 2008; Martínez, Carrera, 87 5 Rodríguez Patino, & Pilosof, 2009). To the best of our knowledge, studies in the 88 literature about the impact of -Lg glycation on the interfacial properties and, 89 consequently, on the foaming properties of this protein are very scarce. Schmitt et al. 90 (2005) in -Lg:acacia gum conjugated by Maillard reaction at pH 4.2, 5.3 and 7.0 91 observed a higher capacity to form and stabilize foams of glycoconjugates than 92 unglycated -Lg, especially at pH 5.3. These authors needed 14 days at 60 ºC to obtain 93 the maximum level (15%) of NH2 loss. Because of the reaction with polysaccharides 94 needs strong conditions and long incubation periods, which would be more expensive 95 from the industrial standpoint, the use of monosaccharides such as galactose, might be 96 of interest, since it allows obtaining modified proteins with a high yield under milder 97 reaction conditions (Corzo-Martínez et al. 2008). 98 Thus, the aim of this work was i) to study the effect of glycation with galactose on 99 the adsorption of -Lg at the air/water interface and to characterize the rheological 100 properties of the interfacial films; and ii) to evaluate foaming properties (foamability 101 and foam stability) of -Lg glycoconjugates in relation to their interfacial behaviour. 102 103 2. Materials and methods 104 2.1. Materials 105 Galactose (Gal) and bovine β-lactoglobulin (β-Lg) (mixture of A and B variants) 106 were purchased from Sigma-Aldrich (St. Louis, MO, USA). All other reagents were of 107 analytical grade. 108 109 2.2. Preparation and purification of β-Lactoglobulin-galactose conjugates 110 Gal and -Lg in a weight ratio of 1:1 were dissolved in 0.1 M sodium phosphate 111 buffer, pH 7 (Merck, Darmstadt, Germany), and lyophilized. The -Lg-Gal powders 112 6 were kept at 40 and 50 °C for 24 and 48 h, respectively (Corzo-Martínez, Moreno, 113 Olano, & Villamiel, 2008), under a vacuum in a desiccator equilibrated at an aw of 0.44, 114 achieved with a saturated K2CO3 solution (Merck). In addition, control experiments 115 were performed with -Lg stored at 40 and 50 °C without galactose during the same 116 periods (control heated -Lg). 117 After incubation, the products were reconstituted in distilled water to a protein 118 concentration of 1 mg/mL. To remove free carbohydrate, 2 mL portions were 119 ultrafiltered through hydrophilic 3 kDa cut-off membranes (Centricon YM-3, Millipore 120 Corp., Bedford, MA) by centrifugation at 1,548 x g for 2 h. After removal of free Gal, 121 samples were lyophilized and stored at -20 ºC for further analysis. 122 Incubations were performed in duplicate, and all analytical determinations were 123 performed at least in duplicate. 124 125 2.3. Solubility of  -lactoglobulin conjugates 126 For solubility evaluation, solutions of native, control and glycated -Lg in distilled 127 water (1 mg/mL) were adjusted to pH 5 and 7 using HCl or NaOH 1 N. After 30 min of 128 stirring at room temperature, the samples were centrifuged for 15 min at 4 ºC and 129 15,000 x g. The protein content in the supernatants was determined by measuring the 130 absorbance at 280 nm (A280) in a Beckman DU 70 spectrophotometer (Beckman 131 Instruments Inc., Fullerton, CA) and the solubility was expressed as the percentage of 132 the total protein content, considering as 100% the A280 of native -Lg. 133 134 2.4. Interfacial properties measurement 135 Interfacial properties (dynamic of surface pressure and surface dilatational 136 properties) of native, control heated and glycated -Lg were determined at pH 7 and 5. 137 7 For this, samples were dissolved in Trizma-HCl buffer (0.05 M, pH 7.0) or acetic 138 acid/acetate buffer (0.05 M, pH 5) (Sigma-Aldrich, St. Louis, MO), the final protein 139 concentration being 5 mg/mL. 140 Time-dependent surface pressure and surface dilatational measurements of native, 141 control heated and glycated -Lg adsorbed films at the air/water interface were 142 performed with an automatic pendant drop tensiometer (TRACKER, IT Concept, 143 Longessaine, France) as previously described (Rodríguez Patino, Rodríguez Niño, & 144 Carrera, 1999; Rodríguez Niño & Rodríguez Patino, 2002). The method involved a 145 periodic automated-controlled, sinusoidal interfacial compression and expansion 146 performed by decreasing and increasing the drop volume at a given desired amplitude 147 (ΔA/A) and angular frequency (ω), and the response of the surface pressure (mN·m-1) 148 is monitored throughout the experiment, being: 149  -1) 150 where 0 is the surface tension of aqueous solution, in the absence of protein (0 = 151 72.5 mN·m-1), and (mN·m-1) is the surface tension in the presence of protein. 152 Since rate of increase of  is initially controlled by the protein diffusion from the 153 bulk phase to the interface, in this work, dynamic of protein adsorption was evaluated 154 considering the first stage of the protein diffusion, by determinating the apparent 155 diffusion constant (Kdif). This was calculated as the slope of the line between the origin 156 (point 0.0) and first point on the plot  vs. square root of time (θ). 157 Regarding surface rheological parameter, the surface dilatational modulus (E) 158 derived from the change in interfacial tension (dilatational stress),  (Eq. (2)), resulting 159 from a small change in surface area (dilatational strain), A (Eq. (3)), may be described 160 by Eq. (4) (Lucassen and van den Tempel, 1972): 161 8 = 0sin (· θ +  ) (2) 162 A = A0sin ( · θ) (3) 163 E = = - = |E|ei  = Ed + iEv (4) 164 165 where 0 and A0 are the stress and strain amplitudes, respectively, θ is the time,  166 is the phase angle between stress and strain, and |E|, the absolute modulus, a measure of 167 the total unit material dilatational resistance to deformation (elastic + viscous), is the 168 ratio (0/ A0). 169 Surface dilatational modulus (E) is a complex quantity and it is composed of real 170 and imaginary parts. The real part of the dilatational modulus (or storage component) is 171 the dilatational elasticity, Ed = |E|·cos  . The imaginary part of the dilatational modulus 172 (or loss component) is the surface dilatational viscosity, Ev = |E|·sin  . The phase angle 173 (  ) between stress and strain is a measure of the relative film elasticity. For a perfectly 174 elastic material stress and strain are in phase (  = 0) and the imaginary term is zero. In 175 the case of a perfectly viscous material,  = 90º and the real part is zero. 176 Interfacial experiments were carried out at 20 ± 0.3 ºC. The temperature was 177 maintained constant by circulating water from a thermostat. Sample solutions were 178 placed in the syringe and subsequently in a compartment, and they were allowed to 179 stand for 30 min to reach the desired constant temperature. Then a drop was delivered 180 and allowed to stand for 10,800 s to achieve protein adsorption at the air–water 181 interface. Surface rheological parameters (E, Ed, Ev and  ) were measured as a function 182 of adsorption time (  ), at 10% of deformation amplitude (ΔA/A) and at 0.1 Hz of 183 angular frequency (). Sinusoidal oscillation for surface dilatational measurement was 184 d  dA/A d  d ln A 9 made with five oscillation cycles followed by a time of 50 cycles without any 185 oscillation up to the time required to complete adsorption. Measurements were made at 186 least twice. The average standard accuracy of the surface pressure was roughly 0.1 187 mN/m. The reproducibility of the results was better than 0.5% and 5.0% for surface 188 pressure and surface dilatational properties, respectively. 189 190 2.5. Foaming properties 191 The foaming properties of native, control heated and glycated -Lg solutions were 192 characterized through their foam formation and stability measured in a commercial 193 instrument (Foamscan IT Concept, Longessaigne, France), based on the ideas by 194 Popineau and co-authors (Guillerme, Loisel, Bertrand, & Popineau, 1993; Loisel, 195 Guégan, & Popineau, 1993). With this instrument the foam formation, the foam stability 196 and the drainage of liquid from the foam can be determined by conductimetric and 197 optical measurements. The foam is generated by blowing gas (nitrogen) at a flow of 45 198 mL/min through a porous glass filter (pore diameter 0.2 mm) at the bottom of a glass 199 tube where 20 mL of sample solution under investigation is placed. The foam volume is 200 determined by use of a CCD camera. The drainage of water from the foam is followed 201 via conductivity measurements at different heights of the foam column. A pair of 202 electrodes at the bottom of the column was used for measuring the quantity of liquid 203 that was not in the foam, while the volume of liquid in the foam was measured by 204 conductimetry in three pairs of electrodes located along the glass column. In all 205 experiments, the foam was allowed to reach a volume of 120 mL. The bubbling was 206 then stopped and the evolution of the foam was analyzed. Foaming properties were 207 measured at 20 ºC from protein aqueous solutions (5 mg/mL) at pH 5 and 7 and at an 208 ionic strength of 0.05 M. 209 16 interface, allowing the intermolecular interaction via thiol-disulfide exchange, that 358 increase the rigidity and cohesion of the interfacial film. 359 Control -Lg heated at both 40 and 50 ºC gave rise to the formation of a film with 360 higher E values than that of native -Lg (Figure 3 (A)), probably due to its higher 361 efficiency of adsorption at the interface (higher Kdif) (Figure 2 (C)) (Bos & van Vliet, 362 2001; Rodríguez Patino et al., 2008). 363 Likewise, whereas glycation at 40 ºC hardly altered rheological characteristics of 364 adsorbed film of -Lg (Figure 3 (A)), being only observed a slight decrease in the 365 dilatational modulus at long term adsorption (E at 10800 s, E10800) with respect to native 366 -Lg, protein glycated at 50 ºC led to the formation of a film with the lowest E values 367 for a given time as compared to films of native, control heated and glycated (24, 40 ºC) 368 protein. Wooster & Augustin (2007) obtained similar results in a study on the 369 rheological properties of the adsorbed films formed by WPI glycated with dextrans of 370 different molecular weights. In agreement with these authors and taking into account the 371 results of intrinsic fluorescence obtained in a previous work (Corzo-Martínez et al., 372 2008), the decrease observed in the dilatational modulus (E) of the -Lg:Gal [48 h, 50 373 ºC] adsorbed film might be due to structural changes undergone by protein during the 374 advanced stages of the MR, since alteration of the conformational state of protein is 375 responsible for the loss of its structural rigidity and, consequently, the loss of firmness 376 of the adsorbed film. 377 Moreover, as observed in Figure 3 (B), -Lg:Gal [48 h, 50 ºC] conjugate showed 378 the lowest and closest values to the ideal behaviour E- values, indicating the existence 379 of weak interactions between components of the adsorbed film. 380 On the other hand, the phase angle (  ) can be considered as a measure of the 381 relative elasticity of the adsorbed protein films. So the more pronounced the decline of 382 17 the phase angle values with the adsorption time () or the surface pressure (), the 383 greater the elasticity of the adsorbed protein film, and vice versa. 384 In general, for all the studied systems, including native, control heated and 385 glycated -Lg, the phase angle (  ) decreased with increasing adsorption time () 386 (Figure 3 (C)) and surface pressure () (Figure 3 (D)), indicating the formation of 387 elastic films. However, for a given time and pressure, the highest  values were 388 observed with control heated -Lg, indicating the formation of a film with a fluid 389 character. This result suggests that the higher E values observed with this system could 390 be due to its molecular packing as a result of the rapid protein adsorption at the 391 interface, and not due to the increase in the interaction degree between the adsorbed 392 molecules (Rodríguez Patino et al., 1999, 2003). Likewise, according to its low 393 dilatational modulus (E) (Figure 3 (A)), the film formed by -Lg glycated at 50 ºC 394 showed a phase angle (  ) for a given time () (Figure 3 (C)) and pressure () (Figure 3 395 (D)) higher than that of films of native and glycated (at 40 ºC) -Lg, indicative of a 396 lower interaction degree between the film components and, hence, of a more fluid 397 character of this film (Horne & Rodríguez Patino, 2003; Rodríguez Patino et al., 2008). 398 At pH 5, the variation of the dilatational modulus (E) over time () for native -399 Lg was little changed with respect to pH 7 (Figure 4 (A)). In other structural studies 400 carried out with -Lg films, other authors have demonstrated that the pH effect on the 401 dilatational modulus and structure of -Lg films is negligible as compared to that 402 observed for other proteins such as -casein (Rodríguez Patino et al., 1999; Rodríguez 403 Patino, Carrera, Rodríguez Niño, & Cejudo, 2001; Rawel, Rohn, Kruse, & Kroll, 2002; 404 Zhang, Foegeding, & Hardin, 2004; Medrano et al., 2009). These authors related the 405 results obtained to the globular nature of -Lg, since globular proteins generally retain 406 their native structure when they are initially adsorbed at the interface. 407 18 The film formed by control heated -Lg showed, for a given time () (Figure 4 408 (A)) and pressure () (Figure 4 (B)), E values lower than that of native -Lg and those 409 reached at pH 7, which could be related to its lower adsorption efficiency at pH 5. 410 Likewise, at pH 5, -Lg glycated at 40 ºC led to the formation of a film with E 411 values similar to those of native -Lg film at short times of adsorption. Moreover, 412 unlike at pH 7, surface dilatational modulus (E) of this film notably increased with the 413 adsorption time, suggesting the formation of high intensity interactions between the film 414 components. 415 At pH 5, the most remarkable differences with respect to pH 7 were observed with 416 -Lg:Gal [48 h, 50 ºC] conjugate, which gave rise to the film with the highest E values 417 for a given time () (Figure 4 (A)) and pressure () (Figure 4 (B)), suggesting, 418 respectively, the formation of a highly elastic and cohesive film, with a great interaction 419 degree between its components. These results are related to the improvement observed 420 in the solubility and, subsequently, in the dynamic of adsorption of this conjugate at pH 421 5, so that this leads to an increase of the surface dilatational modulus of adsorbed film. 422 In addition, -Lg glycated at 50 ºC displayed the lowest  values over the time (Figure 4 423 (C)) and pressure (Figure 4 (D)), which is indicative of the formation of a more elastic 424 and resistant film than that of native, control heated and glycated (at 40 ºC) protein, in 425 agreement with the high E values observed for this system. 426 427 428 429 430 431 432 19 3.3. Foaming properties 433 3.3.1 Foaming capacity 434 The values of the overall foaming capacity (OFC, mL/s), the foam capacity (FC), 435 the foam maximum density (MD), and the relative foam conductivity (Cf, %) obtained 436 with each of the systems assayed at pH 7 and 5 are shown in Figure 5. 437 At pH 7, native, control heated (40 and 50 ºC) and glycated (40 ºC) -Lg showed 438 the same foaming properties (no significant differences between values of OFC, FC and 439 MD), only differing in the value of Cf. These results indicate that the increase produced 440 in the protein diffusion rate (Kdif) as a result of the heat treatment or glycation at 40 ºC 441 (Figure 2 (C)) has no significant effect on its foaming capacity, probably due to that the 442 protein diffusion rate is already good enough for the system foams. This same 443 behaviour can best be seen in Figure 6 (A), where a higher Kdif value did no result in 444 significant increase (P<0.05) in the OFC value. 445 Glycation at 50 ºC, however, had a negative effect on -Lg foaming capacity at 446 pH 7, observing values for the formation parameters OFC and FC significantly (P<0.05) 447 lower with -Lg:Gal [48 h, 50 ºC] conjugate than with native, control heated (40 and 50 448 ºC) and glycated (40 ºC) protein. These results are related to the low Kdif and E values at 449 short times previously observed for this conjugate (Figures 2 (C) and 3 (A)). This fact 450 indicates that the low foaming capacity of -Lg:Gal [48 h, 50 ºC] conjugate at pH 7 is 451 likely due to that its rate of diffusion at the interface and dilatational characteristics of 452 adsorbed film are not good enough to stabilize the bubbles during its formation. 453 At pH 5 (Figure 5), similar to adsorption efficiency (Figure 2 (D)), the foaming 454 capacity of native -Lg did not undergo substantial changes with respect to pH 7. 455 Regarding the effect of the heat treatment in absence of Gal, foams formed with control 456 protein heated at 40 and 50 ºC showed OFC and FC values significantly (P<0.05) lower 457 20 than that formed with native -Lg, probably because its lower solubility and, 458 consequently, worse adsorption efficiency at the air/water interface at this pH (Figure 459 5). 460 -Lg glycated at 40 and 50 ºC displayed a foaming capacity significantly (P<0.05) 461 higher than that of control heated protein and similar to that of native protein, observing 462 no significant differences between OFC and FC values. These results are in good 463 agreement with the dynamic of adsorption previously observed at pH 5 for these 464 systems, which, regardless of being glycated or unglycated, showed a diffusion rate 465 (KDif) and a surface activity (-θ) very similar (Figures 2 (B) and (D)). 466 Moreover, by comparing the results obtained at pH 5 and at pH 7, we observed no 467 important differences between the OFC and FC values of native and glycated (at 40 ºC) 468 -Lg, but a significant increase (P<0.05) in these parameters was found in the case of -469 Lg glycated at 50 °C. This increase was probably due to the higher diffusion rate (Kdif) 470 to the air/water interface displayed by this conjugate at pH 5 (Figure 2 (D)) with respect 471 to that showed at pH 7. This behaviour can best be seen in Figure 6 (B), where it can be 472 observed how systems with a higher Kdif also showed a higher OFC. 473 474 3.3.2 Foam stability 475 To evaluate the capacity to stabilize foams of -Lg glycoconjugates, the half-life 476 time (1/2, s) of foams formed with all the systems assayed was determined (Figure 7). 477 As observed in Figure 7 (A), at pH 7, stability of foam formed with native -Lg 478 was higher than that of foams with control heated and glycated protein, particularly at 479 50 ºC. This is consistent with the worse surface dilatational properties of adsorbed films 480 formed by these systems (Figure 3). 481 21 At pH 5 (Figure 7 (B)), the half-life time of foam with native -Lg (569 ± 26.87 s) 482 did not substantially changed with respect to that obtained at pH 7 (575 ± 0.00 s), a fact 483 that is related to the stability of surface dilatational modulus (E) of film of this protein 484 against changes in pH. Likewise, the worse interfacial characteristics (dynamic of 485 adsorption and surface dilatational properties) observed for the films formed by control 486 heated -Lg at pH 5 as compared to those of native and glycated protein resulted in a 487 lower stability of foams containing control heated protein as foaming agent. 488 On the other hand, unlike at pH 7, glycoconjugates were found to be the best 489 stabilizing agents at pH 5. Thus, the half-life time (1/2, s) of foam with -Lg glycated, 490 particularly at 50 ºC, was notably (P<0.05) higher than that of foams with native and 491 control heated protein. This could be attributed to the increase observed in surface 492 dilatational modulus (E) with increasing time (Figure 4 (A)) and pressure (Figure 4 (B)) 493 for this system, suggesting the formation of an elastic film with a high degree of 494 interaction between its components and, hence, with a high stability against mechanisms 495 of foam destabilization such as drainage of fluid, diffusion or collapse. 496 497 4. Conclusions 498 Although at pH 7 glycation hardly changed the interfacial and foaming 499 characteristics of -Lg, at pH 5, both -Lg:Gal glycoconjugates showed a better 500 dynamic of adsorption to the air/water interface as compared to their corresponding 501 controls of protein heated in absence of Gal. This resulted in a better foaming capacity 502 of -Lg glycoconjugates with respect to native and control heated protein. Likewise, the 503 higher rigidity, cohesion (interaction degree in the interface) and elasticity of adsorbed 504 films formed by -Lg glycated at 40 and, particularly, 50 ºC led to a higher stability of 505 22 foams containing these complexes as stabilizing agents as compared to those foams 506 with native and control heated -Lg. 507 Therefore, from the findings described in this work we can infer that conjugation 508 of -Lg with galactose via the Maillard reaction could be a good alternative to consider 509 when using this protein as a foaming agent. This reaction may extend the applicability 510 range of -Lg allowing its use as a foaming agent in acidic foods such as carbonated 511 beverages, protein-fortified beverages (fruit juices, sports drinks and varieties of these 512 beverages with long shelf-life), manufactured meats, reformed fish products, and a 513 variety of formulated foods. In this way, a future work will be the study of the stability 514 as foam agents of these potential ingredients during the processing and storage of acidic 515 foods. 516 517 Acknowledgments 518 519 This work has been funded by projects ALIBIRD S2009/AGR-1469 (CAM), 520 CICYT through Grant AGL2007-60045, and IBEROFUN 110AC0386 (CYTED). M. 521 Corzo-Martínez thanks Danone Institute for a grant. 522 523 524 525 526 527 528 529 530 23 References 531 532 Baeza, R., Carrera, C., Pilosof, A. M. R., & Rodríguez Patino, J. M. (2005). Interactions 533 of polysaccharides with -lactoglobulin adsorbed films at the air-water interface. 534 Food Hydrocolloids, 19, 239-248. 535 Bals, A., & Kulozik, U. (2003). Effect of preheating on the foaming properties of whey 536 protein isolate using a membrane foaming apparatus. International Dairy Journal, 537 13, 903-908. 538 Bos, M. A., & van Vliet, T. (2001). Interfacial rheological properties of adsorbed 539 protein layers and surfactants: a review. Advances in Colloid and Interface 540 Science, 91, 437-471. 541 Carrera, C., & Rodríguez Patino, J. M. (2005). Interfacial, foaming and emulsifying 542 characteristics of sodium caseinate as influenced by protein concentration in 543 solution. Food Hydrocolloids, 19, 407–416. 544 Chevalier, F., Chobert, J. -M., Dalgalarrondo, M., & Haertlé, T. (2001b). 545 Characterization of the Maillard reactions products of -lactoglobulin 546 glucosylated in mild conditions. Journal of Food Biochemistry, 25, 33–55. 547 Chevalier, F., Chobert, J. M., Popineau, Y., Nicolas, M. G., & Haertlé, T. (2001a). 548 Improvement of functional properties of -lactoglobulin glycated through the 549 Maillard reaction is related to the nature of the sugar. International Dairy Journal, 550 11 (3), 145-152. 551 Corzo-Martinez, M., Moreno, F. J., Olano, A., & Villamiel, M. (2008). Structural 552 characterization of bovine -lactoglobulin-galactose/tagatose Maillard complexes 553 24 by electrophoretic, chromatographic and spectroscopic methods. Journal of 554 Agricultural and Food Chemistry, 56, 4244-4252. 555 Corzo-Martinez, M., Moreno, F. J., Villamiel, M., & Harte, F. M. (2010). 556 Characterization and improvement of rheological properties of sodium caseinate 557 glycated with galactose, lactose and dextran. Food Hydrocolloids, 24, 88-97. 558 Damodaran, S. (1997). Food proteins: an overview. In S. Damodaran, & A. Paraf 559 (Eds.), Food proteins and their applications. Part I: physicochemical bases of 560 protein functionality (pp. 1-25). New York: Marcel Dekker Inc. 561 Davis, J. P., & Foegeding, E. A. (2004). Foaming and interfacial properties of 562 polymerized whey protein isolate. Journal of Food Science, 69, 404-410. 563 Guillerme, C., Loisel, W., Bertrand, D., & Popineau, Y. (1993). Study of foam stability 564 by video image analysis: relationship with the quantity of liquid in foams. Journal 565 of Texture Studies, 24, 287-303. 566 Horne, D. S., & Rodríguez Patino, J. M. (2003). Adsorbed biopolymers: behaviour in 567 food applications. In M. Malmsten (Ed.), Biopolymers at interfaces (pp. 857-900). 568 New York: Marcel Dekker, Inc. 569 Jimenez-Castaño, L., López-Fandiño, R., Olano, A., & Villamiel, M. (2005). Study on 570 β-lactoglobulin glycosylation with dextran: effect on solubility and heat stability. 571 Food Chemistry, 93, 689–695. 572 Jimenez-Castaño, L., Villamiel, M., & López-Fandiño, R. (2007). Glycosylation of 573 individual whey proteins by Maillard reaction using dextran of different molecular 574 mass. Food Hydrocolloids, 21, 433-443. 575 25 Kim, D. A., Cornec, M., & Narsimham, G. (2005). Effect of thermal treatment on 576 interfacial properties of -lactoglobulin. Journal of Colloid Interface Science, 285, 577 100-109. 578 Kinsella, J. E. (1984). Milk proteins: physicochemical and functional properties. CRC 579 Critical Reviews of Food Science and Nutrition, 21, 197-262. 580 Loisel, W., Guéguen, J., & Popineau, Y. (1993). A new apparatus for analyzing 581 foaming properties of proteins. In K. D. Schwenke, & R. Mothes (Eds.), Food 582 proteins: Structure and functionality (pp. 320–323). Weinheim, Germany: VCH. 583 Lucassen, J., & van den Tempel, M. (1972). Dynamic measurements of dilational 584 properties of a liquid interface. Chemical Engineering Science, 27, 1283-1291.J. 585 Lucassen-Reynders, E. H., Lucassen, J., Garrett, P. R., & Hollway, F. (1975). Dynamic 586 surface measurements as a tool to obtain equation-of-state data for soluble 587 monolayers. Advances in Chemical Series, 144, 272-285. 588 Martin, A. H., Grolle, K., Bos, M. A., Cohen-Stuart, M. A., & van Vliet, T. (2002). 589 Network forming properties of various proteins adsorbed at the air/water interface 590 in relation to foam stability. Journal of Colloid and Interface Science, 254, 173-591 183. 592 Martínez, M. J., Carrera, C., Rodríguez Patino, J. M., & Pilosof, A. M. R. (2009). 593 Interactions in the aqueous phase and adsorption at the air–water interface of 594 caseinoglycomacropeptide (GMP) and -lactoglobulin mixed systems. Colloids 595 and Surfaces B: Biointerfaces, 68, 39-47. 596 32 Figure 1. 0 10 20 30 40 50 60 70 80 90 100 110 pH 7 pH 5 % Solubility c b a a a a d e a a Native -Lg Control heated -Lg 24 h 40ºC -Lg:Gal 24 h 40ºC Control heated -Lg 48 h 50ºC -Lg:Gal 48 h 50ºC 33 Figure 2. pH 7 (C) c a b b b 0 10 20 30 40 50 60 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC KDif (mN·m-1·s-1) Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC pH 5 (D) a ab a b a 0 10 20 30 40 50 60 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC KDif (mN·m-1·s-1) Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC 0 5 10 15 20 25 30 35 40 02000 4000 6000 8000 10000 12000 0 5 10 15 20 25 30 35 40 02000 4000 6000 8000 10000 12000 0 5 10 15 20 25 30 35 40 tiempo (s)tiempo (s)  (mN/m) θ (s) pH 7 (A) 02000 4000 6000 8000 10000 12000 0 5 10 15 20 25 30 35 40 (mN·m-1) tiempo (s) θ (s) pH 5 (B) -Lg:Gal 24 h 40ºC Control heated -Lg 24 h 40ºC Native -Lg Control heated -Lg 48 h 50ºC -Lg:Gal 48 h 50ºC 34 Figure 3. Control heated -Lg 48 h 50ºC -Lg:Gal 48 h 50ºC -Lg:Gal 24 h 40ºC Control heated -Lg 24 h 40ºC Native -Lg 02000 4000 6000 8000 10000 12000 0 10 20 30 40 50 60 70 80 90 100 E (mN·m-1) tiempo (s) A θ (s) 02000 4000 6000 8000 10000 12000 0 5 10 15 20 25 30 35 40 45 50  tiempo (s) θ (s) C B 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 0 10 20 30 40 50 60 70 80 90 100 E (mN·m-1)  (mN·m-1) D 0246810 12 14 16 18 20 22 24 26 28 0 5 10 15 20 25 30 35 40 45 50   (mN·m-1) 35 Figure 4. A 02000 4000 6000 8000 10000 12000 0 10 20 30 40 50 60 70 80 90 100 E (mN·m-1) tiempo (s) θ (s) Control heated -Lg 48 h 50ºC -Lg:Gal 48 h 50ºC -Lg:Gal 24 h 40ºC Control heated -Lg 24 h 40ºC Native -Lg 02000 4000 6000 8000 10000 12000 0 5 10 15 20 25 30 35 40 45 50  tiempo (s) θ (s) C B 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 0 10 20 30 40 50 60 70 80 90 100 E (mN·m-1)  (mN·m-1) D 0246810 12 14 16 18 20 22 24 26 28 0 5 10 15 20 25 30 35 40 45 50   (mN·m-1) 36 Figure 5. a a a a b a b b a a 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC OFC Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC a a a a a a b a b b 0 0.2 0.4 0.6 0.8 1 1.2 1.4 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC FC Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC a a a b b a ab b a ab 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC MD Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC 0 5 10 15 20 25 30 35 40 45 50 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC Cf a a b c c c c c c a Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC 37 Figure 6. Control heated -Lg 48 h 50ºC -Lg:Gal 48 h 50ºC -Lg:Gal 24 h 40ºC Control heated -Lg 24 h 40ºC Native -Lg pH 5 (B) pH 7 (A) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 30 35 40 45 50 55 60 65 70 KDif (mN·m-1·s-1) OFC 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 30 35 40 45 50 55 60 65 70 KDif (mN·m-1·s-1) OFC 38 Figure 7. pH 5 (B) a b c d e 0 200 400 600 800 1000 1200 1400 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC 1/2 (S) Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC pH 7 (A) a b c d e 0 100 200 300 400 500 600 700 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC 1/2 (S) Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC pH 5 (B) a b c d e 0 200 400 600 800 1000 1200 1400 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC q1/2 (S) Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC pH 7 (A) a b c d e 0 100 200 300 400 500 600 700 b-Lg nativa b-Lg 24h40ºC b-Lg:Gal 24h40ºC b-Lg 48h50ºC b-Lg:Gal 48h50ºC q1/2 (S) Native -Lg Control heated -Lg 24 h, 40 ºC -Lg:Gal 24 h, 40 ºC Control heated -Lg 48 h, 50 ºC -Lg:Gal 48 h, 50 ºC *Graphical Abstract