This is the accepted version of the journal article: Soukoulis, Christos; Cambier, Sébastien; Serchi, Tommaso; [et al.]. «Rheological and structural characterisation of whey protein acid gels co-structured with chia (Salvia hispanica L.) or flax seed (Linum usitatissimum L.) mucilage». Food Hy- drocolloids, Vol. 89 (April 2022), p. 542-553. DOI 10.1016/j.foodhyd.2018.11.002 This version is available at https://ddd.uab.cat/record/280989 under the terms of the license 1 RHEOLOGICAL AND STRUCTURAL CHARACTERISATION OF WHEY PROTEIN 2 ACID GELS CO-STRUCTURED WITH CHIA (Salvia hispanica L.) OR FLAX SEED (Linum 3 usitatissimum L.) MUCILAGE 4 Christos Soukoulis1,*, Sébastien Cambier1, Tommaso Serchi1, Maria Tsevdou3, Claire 5 Gaiani2, Pau Ferrer1, Petros S. Taoukis3 and Lucien Hoffmann1 6 1Luxembourg Institute of Science and Technology, Environmental Research and Innovation, 5 7 avenue des Hauts-Fourneaux, L4362, Esch-sur-Alzette, LUXEMBOURG 8 2Université de Lorraine, LIBio, Laboratoire d’Ingénierie des Biomolécules, 2 av de la Forêt de 9 Haye, BP 20163, F-54505, Vandoeuvre lès Nancy, FRANCE 10 3National Technical University of Athens, School of Chemical Engineering, Heroon 11 Polytechniou 5, 15780, Zografou Campus, Athens, GREECE 12 *Author to whom correspondence should be sent 13 Dr Christos Soukoulis 14 christos.soukoulis@list.lu 15 16 17 18 19 20 21 22 23 24 25 26 27 1 28 ABSTRACT 29 The effects of different plant seed mucilage (PSM) extracts, namely chia seed (CSM) and 30 flaxseed (FSM), on the kinetics of -glucono-lactone induced acidification and gelation 31 phenomena of whey proteins (5% w/w WPI) were investigated. The rheological and 32 microstructural properties of mixed whey protein-PSM (0.05 to 0.75% w/w) cold-set gels 33 produced at 30 or 37 °C were studied by means of oscillatory rheology and confocal 34 microscopy. On exceeding 0.125% of PSM, a significant reduction of the gelation time due to 35 the formation of loosely entangled whey protein soluble aggregates was observed. The impact 36 of PSM on the gelation rates was closely related to the PSM type and concentration. CSM 37 addition induced a gradual reduction of maximal gelation rate over the entire concentration 38 range tested. On the other hand, FSM conferred a steep impedance of the gelation when 39 exceeded 0.375%, which was associated with the occurrence of segregative phase separation. 40 Fitting the elastic modulus – gelation time data to a model adapted to the Flory-Stockmayer 41 theory, it was demonstrated that the presence of PSM inhibits the whey protein crosslinking 42 capacity under both tested acidification regimes, leading to the formation of shorter protein 43 crosslinks and therefore, to lower gel stiffness. However, the formation rate of elastically active 44 chain networks was found to be increasing for CSM and FSM contents up to 0.5 and 0.25% 45 respectively, suggesting a synergistic acid gel structuring effect of PSM under these conditions. 46 47 48 49 50 Keywords: cold-set gelation; seed coat mucilage; biopolymers; viscoelasticity; protein gel 51 microstructure 2 52 1. INTRODUCTION 53 Owing to their particular texturising, structuring, stabilising and nutritional aspects, milk 54 proteins and polysaccharides constitute hydrocolloids of paramount importance in food product 55 innovation. The addition of polysaccharides in dairy food matrices aims primarily to the 56 increase in the macroviscosity of the serum phase and the hampering of protein instability such 57 as sedimentation (Goh, Sarkar, & Singh, 2008). Polysaccharide – milk protein interactions are 58 governed by the ability of the former to adsorb onto proteins, their surface charge (anionic, 59 cationic or non-ionic) and concentration as well as by extrinsic factors such as pH, temperature, 60 ionic strength and presence of cationic species (Syrbe, Bauer, & Klostermeyer, 1998). Non- 61 adsorbing polysaccharide – milk protein interactions are mainly repulsive inducing exclusion 62 of the individual biopolymer phases known as depletion flocculation (e.g. galactomannans); 63 however, phase separation can be obviated when the non-adsorbing polysaccharide exceeds a 64 certain (critical) concentration (Corredig, Sharafbafi, & Kristo, 2011; Syrbe et al., 1998). On 65 the other hand, adsorbing polysaccharides may lead to milk protein aggregative phenomena 66 due to bridging flocculation (Everett & McLeod, 2005; Girard & Schaffer-Lequart, 2008; Pang, 67 Deeth, & Bansal, 2015). 68 Whey protein acid gel development is considered a two-stage physicochemical process. First, 69 heat treatment of whey proteins, under low ionic strength and protein concentration (CpIWPI) via the electrostatic interaction between negatively charged chain 11 275 segments of the polysaccharides and positive charge bearing patches in the whey protein 276 molecules such as amino, imidazole and guanidine groups (Girard, Turgeon, & Gauthier, 277 2002). It has been shown that the increase in the WPI to polysaccharide ratio (>10:1) and total 278 surface charge density of the anionic polysaccharides may substantially influence the 279 complexation of the latter to whey proteins (Girard & Schaffer-Lequart, 2008). According to 280 the -potential data, the increase in the CSM and FSM to WPI ratio (from 1:100 to 3:20) 281 resulted in a significant increase in the absolute charge density of the binary blends, although 282 no significant differences between CSM and FSM were detected, regardless their differing total 283 hexuronic acid contents (14.67±0.06 and 22.44±0.19 g/100g respectively). This may be 284 attributed to higher buffering capacity of FSM stemming from its higher protein content 285 compared to CSM. The -potential data were positively correlated (Suppl. Fig. 1) with k2 values 286 and negatively with k1 values (only for 37 °C). According to the branching exponential model 287 kinetic parameters (Table 1), the c values ranged from 0.142 to 0.181 h corresponding to pH 288 values ranging from 5.87 to 6.03 and 6.01 to 6.31 h for systems incubated at 30 and 37 °C, 289 respectively. Girard & Schaffer-Lequart (2008) and Girard et al. (2002) studying the 290 complexation of anionic exopolysaccharides (EPS) and low methylated pectins to whey 291 proteins observed an initiation of the electrostatic interaction at values pH 6 with the amount 292 of the non-interacting whey proteins to be decreasing on temperature increase and pH decline, 293 which generally corroborates our findings. Finally, the tacid values (i.e. the time required to 294 achieve pH = 4.5) were well correlated with -potential and flow consistency coefficient (logK) 295 of the WPI/PSM binary systems as shown in Suppl. Fig. 2. Although it is believed that the logK 296 dependence of tacid is primarily associated with the extent of the PSM whey proteins 297 interactions, it is hypothesised that the PSM induced increase in the macroviscosity might have 298 affected indirectly the overall acidification kinetics by sterically hindering the diffusion of 299 gluconic acid at the liquid-liquid interface. 12 300 3.2 Impact of mucilage on gel development kinetics 301 The development of complex modulus G* of the WPI/PSM binary systems as a function of 302 incubation time and pH is illustrated in Fig. 1. In most of cases, gelation time denotes the 303 crossover point of the viscoelastic moduli i.e. tan = G /G = 1. In the present work, the micro- 304 gelled character of CSM endowed a well-defined viscoelastic behaviour in the initial WPI- 305 CSM systems (G G ) and therefore, the calculated complex module (G* = (G')2 + (G")2) 306 values were fitted to a four parametric modified Gompertz model which has been previously 307 used to describe the viscosimetric response of milk protein systems throughout lactic acid 308 fermentation (Soukoulis, Panagiotidis, Koureli, & Tzia, 2007), as follows: G* e 309 logG * = (logG * logG * 0) exp exp ( G* t) + 1 (6) (logG * logG * 0) 310 where G* , G*0 denote the final and initial (pseudo-equilibrium) values of complex modulus, 311 respectively, e is the Euler number, G* denotes the maximal gelation rate (logPa s-1), G* is 312 the gelation lag phase duration (s) and t is the incubation time (s). 313 As shown in Fig. 1, the gelation curves pattern was remarkably affected by the type and 314 concentration of PSM. The gelation lag phase duration G* was significantly (p<0.001) 315 influenced by the incubation temperature and PSM amount (Fig. 2). As expected, the increase 316 in the incubation temperature led to the shortening of gelation lag phase duration by ca. 40% 317 independently to the PSM type, but being significantly tuned by the PSM content. In the latter 318 case, a minimum PSM to WPI ratio of 1:40 was required to detect a significant reduction of 319 G*, whereas no lag phase was detected when PSM reached to 0.75% w/w. In agreement to the 320 G* findings, the increase in the incubation temperature accelerated (p<0.001) the gelation 321 process by ca. 15%. Interestingly, the responsiveness of the maximal gelation rate pattern to 322 the PSM type and content was fairly diversified. Thus, the gelation rate remained constant for 323 up to 0.375% of FSM followed by a steep reduction when exceeded the aforementioned level. 13 324 On the other hand, CSM affected adversely the values for the entire concentration range 325 tested in this work. Parameters such as the acidification rate, the incubation temperature the 326 protein content, the ionic strength and the presence of polysaccharides are known as impacting 327 gelation rate (Cavallieri & da Cunha, 2008; de Jong et al., 2009; Liu et al., 2018; Pang, Deeth, 328 & Bansal, 2015; Pang, Deeth, Sharma, et al., 2015). The G* values denote the time where 329 gelation is initiated and therefore, both incubation temperature and mucilage presence were 330 influencing parameters. According to the complex modulus – pH profiles (data not shown) 331 when CSM or FSM was added up to the level of 0.125%, the gelation process was faster 332 initiated compared to the WPI systems, which could be ascribed to the increase in the WPI 333 soluble aggregates size in the presence of polysaccharides. Kharlamova et al. (2018) 334 demonstrated that although the whey protein aggregates do not possess a determinant role on 335 the acid gels viscoelastic properties, the gelation event may proceed rather faster when larger 336 aggregates are formed after the heat treatment step. When CSM or FSM concentration 337 exceeded 0.125%, a sharp shift of the pHgel towards lower values was observed for both 338 incubation temperature regimes. Thus, it is postulated that the increase in PSM to WPI ratio 339 inhibits sterically the complexation of whey proteins to the anionic polysaccharide backbone 340 and favours their intermolecular repulsion due to the excess of negative charge bearing 341 carboxylic groups (Ye, 2008). Indeed, CLSM analysis of the microstructural aspects of FSM- 342 WPI blends (Fig. 3B) gave support to the previous hypothesis; in general, a good 343 thermodynamic compatibility with WPI was achieved for an FSM content up to 0.125%, whilst 344 spontaneous segregative phase separation was occurred when FSM exceeded 0.125% w/w in 345 accordance to previous studies (Corredig et al., 2011; Khalloufi, Corredig, Goff, & Alexander, 346 2009). On the contrary, CSM based systems did not exhibit any evident biopolymer depleted 347 (WPI or mucilage) microdomains (Fig. 3A); instead, the presence of a rather continuous WPI 348 microstructure with intruding microgelled CSM-rich patches was detected in agreement to the 14 349 observations of Goh et al. (2016). It is therefore assumed that as the CSM concentration 350 increases the associative interactions between the whey protein aggregates is sterically 351 inhibited, a mechanism that has been previously reported in anionic polysaccharide stabilised 352 milk protein gels such as pectins and carrageenans (Everett & McLeod, 2005; Pang et al., 353 2015). 354 The determination of the fractal dimension of the whey protein aggregates by means of 355 rheological, microstructural, light scattering or gel permeability data offer significant 356 information about the interplay between the macroscopic properties of the protein gel and 357 colloidal aspects of the protein aggregates (Ould Eleya, Ko, & Gunasekaran, 2004). On the 358 other hand, models based on percolation or branching (cascade) theory can provide a 359 satisfactory estimation of the mechanical properties of protein gels, especially close to the 360 percolation threshold (Kavanagh, Clark, & Ross-Murphy, 2000). 361 In order to gain insight to the ability of whey protein monomers to undergo non-covalent 362 crosslinking via sulfhydryl ( S H) and disulphide bonds ( S S ) throughout acidification, the 363 storage modulus G data were fitted to an empirical low dimensional model previously used 364 for investigating in situ sol-gel transition kinetics of biopolymers (Adibnia & Hill, 2016; 365 Calvet, Wong, & Giasson, 2004) as follows: t 366 G'(t) = G ' (7) t + 367 where G ' (in Pa s) denote the elastic modulus at the pseudo-equilibrium (tacid), t is the gelation 368 time, is the gelation half-time (s), and is a constant relative to the asymptomatic slope, P ( G' 369 ), at the gelation half-time. According the cascade theory (an extension of Flory- 4 370 Stockmayer concept of gelation), the elastic modulus at the pseudo-equilibrium can be 371 described as follows: 15 f 372 G ' = N 2 a (1 - v)2 (1 - ) P k T (8) f 373 where Ne = N 2 a (1 - v)2 (1 - ) denotes the number of the elastically active network chains 374 per primary volume, f is the number of sites among each molecule’s length, a is the fraction of 375 the sites that have reacted, whereas v and are parameters derived from cascade theory for gel 376 network formation (Lopes da Silva, Rao & Fu 1998). G' 377 At gelation half-time point, P = , and hence, Eq. 8 can be written as follows: 4 4 k T ne 378 G' = (9) 379 where: k is the Boltzmann constant and T (in K) is the incubation temperature and e (mol s-1) 380 denotes the rate of elastically active network chains formed per whey protein aggregate 381 monomers volume (Lopes da Silva, Rao, & Fu, 1998). 382 As illustrated in Fig. 4, the gelation half-time values did not exert a clear response pattern to 383 PSM concentration (<0.5% w/w) with the gelation half-time values to oscillate around 384 1217±106 and 777±115 s at 30 and 37 °C, respectively. Corroborating the G* values, the 385 increase in incubation temperature shortened significantly (p<0.001) the values. Suchlike to 386 the observations of Adibnia & Hill (2016), it is assumed that can be used as a measure of the 387 whey protein crosslinking capacity. As clearly depicted in Fig. 4, the presence of PSM led to a 388 significant (p<0.001) reduction of values, most likely due to the reduction of the sites number 389 allowing whey protein interaction via S S and S H. Interestingly, neither the PSM type nor 390 the incubation temperature were influential on values. 391 At gelation half-time, the e values were varied according to the PSM type and concentration 392 as well as incubation temperature (Fig. 4). In CSM based systems, an increase in the e values 393 was observed when the WPI to mucilage ratio did not exceed 40:3 and 10:1 at 37 and 30 °C, 394 respectively. In the case of FSM, the gel elasticity synergism boundary was shifted to 16 395 remarkably higher WPI to mucilage ratios i.e. 40:1 to 40:3, as a result of the occurring 396 segregative phase separation. It should be noted that for WPI to FSM ratios exceeding 40:1, 397 the elastically active network formation rate was almost negligible for both tested incubation 398 temperatures, exerting a minor recovery of e values at the highest FSM content. On the other 399 hand, CSM addition did not compromise the formation of the elastically effective network 400 chains regardless the reduction of whey protein monomers crosslinking capacity. It is therefore 401 presumed that the microgelled structure conformation of CSM favours the overall gel elasticity 402 via an active filler related mechanism. The latter appears to be supported not only by the elastic 403 modulus values but also by the CLSM micrographs of the acid gels illustrating a satisfactory 404 protein adsorbing ability of CSM (Fig. 5 and Suppl. Fig. 3). 405 3.3 Impact of mucilage on the mechanical and structural aspects of the acid gels 406 The whey protein acid gels were pre-conditioned at 25 °C for 2 h prior to frequency sweep 407 rheological testing (Figs. 6 and 7). The individual WPI acid gels exerted significantly (p<0.05) 408 higher stiffness and elasticity when produced at lower incubation temperatures corroborating 409 previous findings (Lucey, 2002). As illustrated in Fig. 6, CSM addition imparted sufficient 410 structuring as indicated by the ca. 4-fold increase in the gels stiffness. Nevertheless, the 411 addition of CSM did not modify substantially the domineering weak gel viscoelastic behaviour 412 (tan >0.1) observed in WPI individual acid gels. CLSM analysis of the CSM based acid gels 413 (Fig. 5A), revealed a sufficient adsorption of mucilage components onto whey proteins as 414 evidenced by the purple coloured electrostatically complexed protein-polysaccharide 415 aggregates. Due to the selectivity of Calcofluor stain to the cellulosic fraction, we were not 416 able to justify the topological presence of pectinaceous RG-I matter e.g. found in the interspace 417 of the protein network or being stranded to the gel network. However, it is presumed that RG- 418 I fraction of mucilage would be also preferably adsorbed to whey proteins at pH