- Email: [email protected]
phospholipid ratio on the physicochemical and interfacial properties of biomimetic milk fat globules
Food Hydrocolloids 97 (2019) 105179
Contents lists available at ScienceDirect
Food Hydrocolloids journal homepage: www.elsevier.com/locate/foodhyd
The influence of protein/phospholipid ratio on the physicochemical and interfacial properties of biomimetic milk fat globules
T
Min Chena,∗, Leonard M.C. Sagisb a b
College of Food Science and Engineering, Ocean University of China, Yushan Road 5, Qingdao, China Physics and Physical Chemistry of Food, Wageningen University, Bornse Weilanden 9, 6708WG, Wageningen, the Netherlands
A R T I C LE I N FO
A B S T R A C T
Keywords: Emulsion Milk fat Protein Phospholipids Oil/water interface
The effect of protein/phospholipid ratio on the physicochemical and interfacial properties of biomimetic milk fat globules (BMFGs) was investigated. Butter milk phospholipid (BMP) vesicles prepared with ultrasonication were mixed with whey protein isolate (WPI) at different protein/phospholipid ratios (w/w). After emulsification, the Turbiscan stability index (TSI) was used to quantify the initial stability of BMFGs. At different pH values, visual observation and CLSM were performed to obtain the macro stability and microstructure of BMFGs. In addition, large amplitude oscillatory surface rheology was applied and analysed with Lissajous curves, to explore the oil/ water interfacial properties of BMFGs in the nonlinear viscoelastic regime. The TSI indicated a faster creaming rate of BMP stabilized droplets compared to those stabilized by WPI. The macro stability of WPI/BMP co-stabilized BMFGs at different pH was dominated by WPI and BMFGs prepared at the WPI/BMP ratio of 20:80 phase separated from pH 3.0 to 5.0. At acidic pH, the WPI/BMP ratios strongly affected the coalescence and flocculation of BMFGs at the micro level. The surface elasticity indicated that WPI dominated the rheological properties of WPI/BMP co-stabilized oil/water interfaces. Besides, the significant dependence of surface elasticity on deformation amplitudes within the nonlinear regime of WPI stabilized oil/water interfaces indicated a clearly different structure from BMP stabilized interfaces. Conclusively, results of this study showed that the physicochemical and interfacial properties of BMFGs were significantly influenced by the protein/phospholipid ratio in the bulk, which needs to be considered in the design and application of biomimetic milk fat globules.
1. Introduction Milk fat, in human breast milk present as milk fat globules (MFGs) with an average diameter of 4–5 μm, contains a core of triglycerides (TAG) covered with a milk fat globule membrane (MFGM). Due to the bio-synthesis process of human MFGs (Smoczyński, 2017), the MFGM exhibits a unique lamellar phospholipid trilayer structure and a multicomponent composition containing polar lipids and functional membrane proteins. These bio-functional components are vital for the brain development, cell metabolism and immune maturation of newborns (Claumarchirant et al., 2016; Hernell, Timby, Domellöf, & Lönnerdal, 2016). Therefore, biomimetic milk fat globules (BMFGs) prepared with MFGM enriched materials have received enormous attention recently (Gallier et al., 2015; Livney, Ruimy, Ye, Zhu, & Singh, 2017; Lopez et al., 2017), often with the aim of developing improved infant milk formulas. The gastrointestinal fate of emulsions can be modulated by the dairy emulsifier composition and the structure of oil droplets (Lecomte et al.,
∗
2015; Liang, Zhang, Wang, Jin, & McClements, 2018; Mathiassen et al., 2015). The composition of MFGM enriched ingredients (including butter serum, butter milk, MFGM polar lipid fraction and MFGM extracts etc.), with respect to the content of milk proteins, polar lipids and membrane proteins can be rather different due to variations in preparation methods (Gassi et al., 2016; Haddadian, Eyres, Bremer, & Everett, 2018; Holzmüller, Müller, Himbert, & Kulozik, 2016; Jukkola, Partanen, Rojas, & Heino, 2018; Le, Van Camp, Rombaut, van Leeckwyck, & Dewettinck, 2009). A comparison of emulsifying properties of MFGM materials isolated from different dairy by-products has demonstrated that not only the polar lipid content but also the presence of other components (e.g., whey proteins, caseins, MFGM-specific proteins, and minerals) determine their emulsifying properties Phan, Le, Van der Meeren, & Dewettinck, 2014). BMFGs have been found to be coated solely with milk proteins, solely with MFGM polar lipids and membrane proteins, or with their mixtures. This effect of the multicomponent composition of MFGM enriched ingredients on their emulsifying and interfacial properties has so far not been studied
Corresponding author. E-mail addresses: [email protected], [email protected] (M. Chen).
https://doi.org/10.1016/j.foodhyd.2019.105179 Received 12 April 2019; Received in revised form 18 June 2019; Accepted 21 June 2019 Available online 22 June 2019 0268-005X/ © 2019 Elsevier Ltd. All rights reserved.
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
(Bergfreund, Bertsch, Kuster, & Fischer, 2018). Therefore, to mimic the lipid composition and structure of breast milk, an oil formulation including five kinds of oils was used to prepare BMFGs in this research. Vesicles of milk polar lipids (BMP) were prepared with ultrasonication. Then WPI was mixed with vesicle dispersions to prepare WPI/BMP mixtures containing different fraction of BMP (0%, 20%, 50%, 80%, 100%, w/w). After the preparation of BMFGs, the emulsion stability was assessed with a turbiscan. At different pH, the stability and morphology of oil droplets was measured by visual appearance and confocal laser scanning microscopy (CLSM), respectively. In addition, the interfacial properties including adsorption kinetics and linear/nonlinear dilatational surface rheology were characterized with an automatic drop tensiometer. It was hypothesized that varying the protein/ phospholipid ratios would lead to a different oil/water interfacial composition and structure, which would affect the initial stability, electrostatic properties and morphology of biomimetic milk fat globules at various pH conditions. A deeper understanding of the relation between stabilizer composition and the physicochemical and interfacial properties of model milk fat emulsions can supply a theoretical basis for the development of better infant formulas.
systematically. Whey protein isolate (WPI) has been intensively applied in food and medical emulsions, for example, in beverages, products with encapsulated oils, and functional delivery systems for people with special medical conditions (Cornacchia & Roos, 2011; Jiménez-Colmenero, 2013; Melnikov et al., 2017; Nejadmansouri, Hosseini, Niakosari, Yousefi, & Golmakani, 2016). Most of the stabilizing ingredients for oil emulsification in infant formula production also contain more than 50% of WPI. Phospholipids are the second abundant ingredient. Due to their amphiphilic property, phospholipids tend to form micelles, vesicles and lamellar assemblies in aqueous solutions. The form in which they are dispersed, the composition of phospholipid species, and their fatty acid composition will significantly influence the emulsifying properties of phospholipids (Magnusson, Nilsson, & Bergenståhl, 2016). In protein/ phospholipid mixed systems, the composition of phospholipid species and their fatty acid composition can influence the interaction between proteins and phospholipids, thereby modulating their emulsifying properties (García-Moreno, Horn, & Jacobsen, 2014). It was reported that MFGM protein concentrate exhibits better emulsifying/stabilising properties compared to phospholipids (Phan, Le, Van de Walle, Van der Meeren, & Dewettinck, 2016). However, the effect of bulk composition (protein/phospholipid ratios) of emulsifiers on the physicochemical properties of BMFGs (for instance, the interaction and microstructure of droplets at various electrostatic conditions) is not clear yet. In addition, how the properties of oil/water interfaces stabilized by MFGM ingredients are affected by their bulk composition is also not fully understood. Recently, the oil/water interfacial properties of three fractions of MFGM enriched materials were studied with a Langmuir trough and results indicated that the surface adsorption kinetics vary significantly with different protein and polar lipid content in the fraction (Priyanka, Sabine, Aman, Koen, & Christophe, 2017). The studied samples were a mixture of milk proteins, polar lipids and membrane proteins. The exact role of each component on the oil/water interfacial structure was not clearly elucidated. At air/water interfaces stabilized by protein/phospholipid mixtures, it was observed that the interaction between proteins and phospholipids at the interface was significantly influenced by pH, which led to different interfacial structures (Caro, Niño, & Patino, 2009). Besides planar interfaces, pendant/rising drop tensiometry has been widely applied in the characterization of interfacial properties of curved oil/water interfaces in correlation to emulsifying properties. For pure phospholipids, the dispersed form of phospholipid molecules, including micelles, vesicles or lamellar assemblies can influence the adsorption kinetics and surface load of phospholipids at the oil/water interface (Magnusson et al., 2016). The rearrangement of phospholipid molecules between the oil/water interface and bulk phase during the evolution of the emulsions strongly dominates the emulsion stability against coalescence and Ostwald ripening (Sommerling et al., 2018). In mixed systems of polar lipids and proteins, a “skin-like” folded film consisting of a phospholipid/protein complex layer can be formed at the water/chloroform interface of a pendant drop, and its formation is accelerated by the co-adsorption of proteins and phospholipids (He et al., 2008). Until now, studies on dilatational surface rheology of interfaces stabilized by proteins and phospholipids were mostly conducted in a linear regime, although in reality the destabilization of emulsions, in particular coalescence, involves large deformations of the interface which occurs in the nonlinear viscoelastic regime. A characterization of the nonlinear surface rheology of oil/water interfaces stabilized by phospholipid/protein mixed systems has so far not been performed. The aim of this study is to determine the effect of protein/phospholipid bulk composition of mixtures of WPI and BMP on the stability and physicochemical properties of biomimetic milk fat globules prepared with these mixtures, and link this effect to properties of the oil/ water interfaces. The adsorption and rheological properties of WPI at the oil/water interface are greatly affected by oil hydrophobicity
2. Materials and methods 2.1. Materials Buttermilk powder was purchased from Dairygold Co., Ltd. (Cork, Ireland). Whey protein isolate (WPI) BiPRO (Protein > 90%) was obtained from Davisco Foods International, Inc. (Le Sueur, Minnesota, USA). Medium chain triacylglycerol MIGLYOL812N was obtained from IOI Oleo Chemicals (Shanghai, China). Nile Red (5H-Benzo α-phenoxazine-5-one, 9-diethylamino) and FITC were purchased from SigmaAldrich (St. Louis, USA). OPO structured lipid was supplied by Yihai Kerry (Beijing, China). DHA algal oil was obtained from Xiamen Huison biotech Co., Ltd. (Xiamen, China) and ARA enriched oil from mortierella alpine were supplied by Xiamen Kingdomway Group Company (Xiamen, China). Coconut oil and sunflower oil were purchased from the local supermarket. Phospholipid standards including phosphatidylcholine (PC), phosphotidyl ethanolamine (PE), phosphotidylserine (PS), sphingomyelin (SM) and phosphatidylinositol (PI) were purchased from Sigma-Aldrich (St. Louis, USA). The other analytic chemicals were bought from Sinopharm Chemical Reagent Co., Ltd. (Beijing, China). 2.2. Methods 2.2.1. Extraction of milk phospholipids from reconstituted buttermilk Buttermilk powder of 2.0 kg was reconstituted in 6.0 L MilliQ water and prewarmed to 40 °C before their pH was adjusted to 4.6 with 2.0 M HCl. Subsequently, the solutions were centrifuged twice at 4745 g for 20 min each. After centrifugation, the obtained supernatant (around 2.6 L) concentrated to 650 mL with a vacuum rotatory evaporator at 60 °C. This was mixed with 2.1 L of chloroform/methanol 2:1 and centrifuged at 4745 g for 5 min. The chloroform layer, filtered through membranes with a pore size of 0.45 μm, was further centrifuged at 7104 g for 5 min. The purified lipid fraction was dried under evaporation at 60 °C and dissolved in cold acetone afterwards. The precipitate in cold acetone was collected, dissolved in chloroform, dried under vacuum rotatory evaporation, and designated as BMP. Their phospholipid composition was further characterized with HPLC-ELSD. 2.2.2. Emulsion preparation BMP and whey protein isolate (WPI) were reconstituted in MilliQ at 10.0 mg/mL respectively and stirred for a whole night. Sodium azide was added at 0.02% (w/w) as preservative. To prepare phospholipid vesicles, the BMP solution was blended with an Ultra-Turrax T25 (IKA, Staufen, Germany) at 13200 rpm for 5 min, and subsequently processed with an ultrasound sonicater (Xinzhi biotechnology Co. Ltd, Ningbo, 2
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
et al., 2017). The staining of samples was briefly as follows: FITC was used to stain proteins. FITC solution of 20 μL (1.0 mg/mL in DMSO) was add to 1.0 mL of sample. Nile Red was used to stain the triglycerides, prepared at a concentration of 42 μg/mL in acetone. Nile Red solution of 10 μL was added to 1.0 mL of sample. Emulsions labelled with FITC and Nile red were kept in the dark for at least 1 h at room temperature prior to observation. The observation was performed using a 60 × (numerical aperture NA 1.4) oil immersion objective with an argon laser (Excitation wavelength of 488 nm, with emission detected between 500 nm and 545 nm) and a diode laser (Excitation wavelength of 559 nm, and emission detected between 570 nm and 625 nm) in a sequential scanning mode at room temperature. Two dimensional images were captured at a resolution of 512 × 512 pixels by further zooming in 1.5 × .
China) at 400 W for 15 min. WPI and polar lipid vesicle mixtures containing various fraction of BMP (0%, 20%, 50%, 80% and 100%, w/w) at a fixed total concentration of 2.0 mg/mL were prepared and prewarmed to 50 °C in a water bath. The oil fraction prewarmed to 50 °C was mixed with emulsifiers at a ratio of 1:9 (v/v), and was pre-homogenized with an Ultra-Turrax T25 at 13200 rpm for 2 min. Subsequently, the coarse emulsion was sonicated at 400 W for 5 min. The composition of the oil was chosen to closely mimic the TAG core of breast milk fat globules: OPO structured lipid (44.0%, w/w), coconut oil (13.3%, w/w), sunflower oil (38.5%, w/w), DHA algal oil (1.6%, w/ w) and ARA enriched oil from mortierella alpine (2.6%, w/w). Based on this resemblance, we consider these model fat globules as biomimetic milk fat globules. 2.2.3. Transmission electronic microscopy (TEM) The morphology of BMP vesicles was determined with a JEM 2100 (JEOL, Japan) equipped with a CCD camera ORIUS SC1000 and an accelerating voltage of 200 kV, according to the method described elsewhere (Magnusson et al., 2016), with some modifications. Shortly, samples of 10 μL was deposited on a lacey carbon film supported by a copper grid for 60 s, and excess liquid was removed with filter papers. Subsequently, the above procedure was repeated with 10 μL of phosphotungstic acid (2.0%) solution. The grid was kept at room temperature and air dried before observation. Images were captured with a magnification of 12000 × .
2.2.7. Oil/water interfacial properties of WPI/BMP mixtures The oil/water interfaces stabilized with WPI/BMP mixed systems were investigated with a Tracker automatic drop tensiometer (TECLIS Instruments, Civrieux d’Azergues, France). A rising droplet of caprylic/ capric triglyceride (MIGLYOL812N, IOI Oleo Chemicals, Hamburg, Germany) was created at the tip of a motorized syringe, in the aqueous phase. Surface pressure was monitored without oscillation for the first 100 s at room temperature (15–18 °C). Subsequently, a sinusoidal dilatational deformation was applied to the oil/water interface with an amplitude of 10% for 120 min, after which the interface was considered to be close to equilibrium. After this phase a strain sweep was performed, with three amplitudes of deformation (10%, 20% and 30%) at a frequency of 0.01 Hz. The area and volume of the non-deformed oil droplet were kept constant at 21.6 mm2 and 10.0 μL, respectively. Three cycles were performed at each amplitude and each cycle included five periods separated by one blank period (100 s). At least two measurements were performed for each sample. The dilatational elastic modulus (E’) and viscous modulus (E″) were calculated with the software WDROP 10.10.9.0 (TECLIS Instruments, France). These moduli are calculated based on the intensity and phase of the first harmonic of the Fourier transform of the oscillating surface pressure signal. Any contributions from higher harmonics, resulting from nonlinearity of the response, are neglected, so results of these calculations are accurate only in the linear response regime. To characterize the nonlinear response at large deformation amplitude, Lissajous curves of Π(t) versus ΔA(t)/A0 were plotted, where Π(t) was the oscillating surface pressure and ΔA(t)/A0 was the deformation of the droplet area.
2.2.4. Particle size and zeta potential The particle size of vesicles and oil droplets was measured with a Nanozs90 (Malvern Instruments, Malvern, U.K.) as described elsewhere (Lopez et al., 2017; Moran-Valero, Ruiz-Henestrosa, & Pilosof, 2017). The refractive index was set to 1.333 for water, 1.46 for BMP vesicles and 1.458 (633 nm) for oil droplets, respectively. The volume mean diameter (De Broucker diameter D43, D43 =
∑1n Di4 vi ∑1n Di3 vi
) and z-average were
used to characterize the particle size distribution of vesicles and fat droplets. The zeta potential (ζ-potential) was also determined with Nanozs90. Samples were diluted 20 times with continuous phases for vesicles and 100 times for oil droplets. The initial pH value was 6.11 for BMP-stabilized emulsion, 7.39 for WPI-stabilized emulsion and in-between 6.11 and 7.39 for emulsions stabilized by WPI/BMP mixtures at different ratios. Besides the initial pH values, the pH of oil droplets was adjusted to 2.0, 3.0, 4.0 and 5.0 before measurements. All experiments were performed at 25 °C, and at least in triplicate for each sample.
3. Results and discussion 2.2.5. Emulsion stability assessment 2.2.5.1. Turbiscan stability index (TSI). The destabilization of BMFGs during aging was monitored with a Turbiscan LAB expert (Formulaction, Toulouse, France). The turbiscan stability index (TSI), calculated based on the variation in the backscattering light intensity at different locations of the test tube (Top, middle and bottom), was used to interpret different destabilizing process like creaming, flocculation and coalescence. In this study, freshly prepared emulsions (20.0 mL) were cooled down at room temperature for 30 min, before a 2 hmeasurement at 25 °C. At least 2 tests were conducted for each sample. The TSI of the bottom part (16.7 mm from the vial bottom) represents the stability of studied emulsions.
3.1. Characterization of butter milk phospholipid (BMP) vesicles BMP vesicles were characterized for their phospholipid composition, particle morphology, particle size distribution and zeta-potential. Results from HPLC-ELSD showed that BMP contained 8.1 ± 0.9% of PI (phosphatidylinositol), 23.7 ± 0.3% of PE (phosphatidylethanolamine), 12.2 ± 0.7% of PS (phosphatidylserine), 38.4 ± 2.1% of PC (phosphatidylcholine) and 17.6 ± 0.8% of SM (sphingomyelin), which was consistent with the phospholipid composition of buttermilk in previous research (Morin, Britten, Jiménez-Flores, & Pouliot, 2007; Morin, Jiménez-Flores, & Pouliot, 2004, 2007). The phospholipid composition and salt concentration can influence the dispersed form of phospholipids in aqueous solution. With more than 80% of PE present, phospholipids mainly form microtubules, however, with less PE present, they mainly form spherical vesicles, which are unilamellar in water but multilamellar in salt or protein buffer (Waninge, Nylander, Paulsson, & Bergenståhl, 2003). According to the TEM image in Fig. 1, polydisperse spherical particles were observed after sonication with a similar morphology as found by (Thompson, Haisman, & Singh, 2006; Thompson, Hindmarsh, Haisman, Rades, & Singh, 2006). A clear delineation in terms of particle size between small unilamellar vesicles
2.2.5.2. Emulsion stability at different pH. The pH of studied emulsions was first measured with a pH meter, subsequently adjusted to 2.0, 3.0, 4.0 and 5.0 respectively, and left undisturbed for a whole night. The stability of emulsions at different pH values was compared by their visual appearance in the images taken with a common camera. 2.2.6. Confocal laser scanning microscopy (CLSM) The microstructure of oil droplets was investigated with a Fluo View TM FV1000 (Olympus, Japan) according to previous research (Lopez 3
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
Fig. 1. Particle size distribution in volume fraction (%) and the TEM image of butter milk phospholipid (BMP) vesicle dispersion with a scale bar at 200 nm.
Fig. 2. The turbiscan stability index (TSI) of WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP.
(SUVs) and large unilamellar vesicles (LUVs) is still debated (Akbarzadeh et al., 2013). The BMP vesicles shown in Fig. 1 were roughly in a size range of 20–1000 nm. Therefore, the BMP dispersion prepared in this study contained mainly SUVs and a small fraction of LUVs. The volume mean diameter (D43) of BMP vesicles was 122.7 ± 16.8 nm and the zeta potential was −46.4 ± 1.2 mV. The strong negative charge was closely related to the phospholipid composition of BMP (Morini et al., 2015).
3.2. Formation and stability of biomimetic milk fat globules (BMFGs) The BMFGs were prepared with a special oil formulation, and emulsified by WPI/BMP mixed systems at different protein/phospholipid ratios. The turbiscan stability index (TSI) was determined to evaluate emulsion stability (Fig. 2). The data indicated that creaming was the main destabilizing factor of emulsions prepared with WPI, BMP and their mixtures. For the WPI/BMP mixed systems, there was no significant increase in TSI with increasing BMP vesicle concentration in the bulk, up to 50% (w/w). However, when the mixture was composed of 80% of BMP vesicles and 20% of WPI, the emulsifying capability seemed to be less effective than pure BMP dispersions. This was probably caused by the competitive adsorption between WPI and BMP vesicles at the oil/water interface (Phan et al., 2016). The TSI of emulsion stabilized with pure BMP vesicle dispersion was about twice of that for WPI, which indicated that the creaming rate of the former was faster than the latter. As shown in Fig. 3, the volume mean diameter and zaverage particle size of the emulsions became significantly larger when the BMP vesicles fraction increased to 80% of the mixture. To a first approximation, the stability of a food emulsion to creaming can be estimated using Stokes law, which predicts a creaming velocity of the droplets that scales with the square of the radius of the droplets (McClements, 2015). The mean radius of WPI-coated droplets and BMPcoated droplets was approximately equal to 250 nm and 350 nm respectively. Then according to Stokes law the creaming rate of droplets in the BMP stabilized emulsion would be a factor of 1.96 compared to droplets in the WPI stabilized emulsion, which was well consistent with the TSI results from the turbiscan. This implies the emulsions were unstable to creaming, but no significant flocculation or coalescence occurred shortly after the emulsion preparation.
Fig. 3. The volume mean diameter (D43) and Z-average particle size of biomimetic milk fat globules (BMFGs) prepared with WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP.
3.3. Effect of WPI/BMP ratio on the physicochemical properties of BMFGs at different pH The visual appearance of BMFGs before and after pH adjustment was shown in Fig. 4(a-e). With different WPI/BMP ratios in the bulk, emulsions phase separated at different pH values. At pH 5.0, close to the isoelectric point of WPI (PI = 4.8), all WPI/BMP co-stabilized BMFGs phase separated, but BMFGs stabilized by pure BMP did not. At 4
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
Fig. 4. Visual appearance of BMFGs prepared with WPI/BMP mixed systems containing different amount of BMP at different pH values: (a) initial pH, (b) 5.0, (c) 4.0, (d) 3.0 and (e) 2.0. For each image, test tubes from left to right represented 0%, 20%, 50%, 80% and 100% of BMP present in bulk.
Fig. 5. CLSM images of BMFGs prepared with WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP at pH 3.0 (a–e) and 4.0 (f–j). The protein was stained with FITC and the triglyceride was stained with Nile Red. In the images, green referred to proteins, red referred to oil and yellowish colour referred to the area where proteins and oil were merged together. The scale bar was 10 μm.
pH 2.0, close to the isoelectric point of the phospholipids (PI = 2.0), all WPI/BMP co-stabilized emulsions remained stable, but BMFGs coated with only BMP phase separated. This indicated that the stability of BMFGs at acidic conditions was dominated by WPI for WPI/BMP mixed systems. Visual phase separation occurred to BMFGs in the pH range of 3.0–5.0 when they were prepared with 20% of WPI and 80% of BMP. The specific pH where phase separation occurs was dependent on the protein/phospholipid ratio in the bulk, which needs to be taken into account in food processing and design. Observation by CLSM was further conducted to determine the microstructure of fat droplets at pH 3.0 and 4.0 (Fig. 5 a-j). At pH 4.0, more coalescence occurred and droplets were generally larger compared to pH 3.0. So the oil/water interface was apparently less stable against coalescence at this pH. The organization of droplets was different in visually phase-separated emulsions. Micron-sized spherical droplets without flocculation were observed for pure BMP stabilized BMFGs. When there was WPI present, large protein aggregates with attached droplets were formed (Fig. 5 d and i) and irregular shaped oil lumps were observed (Fig. 5 h), probably due to bridging flocculation. BMP and WPI have opposite charges at pH 3.0 and 4.0 which can give attractive electrostatic interactions between negative phospholipid patches on the droplet interface, and positively charged WPI patches on the interface, or between WPI and BMP in the bulk. For those emulsions that did not phase separate at pH 3.0 and 4.0 visually, a clear phase separation occurred at the micro level (Fig. 5 b, c and g). These differences in the organization of droplets at the micro and macro level were closely related to the interactions between droplets at different pH values. As shown in Fig. 6, the zeta potential of freshly prepared BMFGs was between −46.7 and −58.2 mV, so they were highly negatively charged around neutral pH. This high negative charge induced a strong repulsion between droplets which kept them from aggregating. At pH 3.0, BMFGs appeared to be highly positively charged when they were prepared with 50% BMP or less in the bulk solution. The samples containing 20% and 50% BMP showed a limited degree of flocculation, most likely due to bridging flocculation mediated by negatively-charged BMP. For droplets coated with 80% and
Fig. 6. Zeta potential (ζ) of BMFGs prepared with WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP at different pH values (from initial pH to 2.0).
100% BMP the zeta potentials are around +20 and −20 mV respectively, so these systems were only marginally stable at pH 3.0, which may have led to the formation of the larger aggregates at 80% BMP. At pH 4.0 the 100% BMP stabilized droplets have a zeta potential below −40 mV, and were stabilized by strong electrostatic repulsion. The 5
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
50% and 80% BMP samples have a very low zeta potential and were extensively flocculated. As a result of a weaker interface, these systems also showed significant coalescence. The 20% BMP and pure WPI systems have a highly positive zeta potential, which for the latter resulted in a stable emulsion, and the former into a network structure most likely formed by bridging interactions mediated by negative BMPs. From these observations it is clear that the stability of BMFGs against aggregation and coalescence at different pH values is strongly affected by the WPI/BMP ratios. Most likely this dependence is coming from the effect of this ratio on the initial oil/water interfacial composition and structure. Therefore, the effect of bulk composition on the oil/water interfacial composition and microstructure was investigated. 3.4. Oil/water interfaces stabilized with WPI/BMP mixed system The interfacial properties of WPI, BMP and WPI/BMP mixed systems were determined with the rising drop method using an automatic drop tensiometer. The interfacial tension of the clean oil/water interface was 20.69 ± 0.20 mN/m. The surface pressure was in a range of 10.87 ± 0.05–12.19 ± 0.05 mN/m after 2 h-monitoring for all studied samples. With 80% of BMP present, the surface pressure was slightly lower than the other samples, which was in line with the larger droplet size in Fig. 3. After this initial phase, the dependence of the surface elastic modulus E’ and viscous modulus E” of the oil/water interface on deformation amplitude was measured, by applying a sinusoidal oscillatory deformation, and the results based on a first harmonic analysis were presented in Fig. 7. The E” of the oil/water interfaces was below 5.0 mN/m for all samples, and not significantly different at various WPI/BMP ratios. For all samples E’ was significantly higher than E”, so at this particular frequency (0.01 Hz), the interfaces appeared to have a predominantly elastic response. The pure WPI-stabilized oil/water interface was significantly more elastic than the pure BMP-stabilized one. In addition, the elastic modulus of the pure WPIstabilized interface decreased significantly when the deformation amplitude increased from 10% to 30%, whereas the moduli of the pure BMP stabilized interface appeared to be independent of strain amplitude, and appeared to be measured in the linear response regime. For the WPI/BMP co-stabilized oil/water interfaces, E’ also showed a significant dependence on deformation amplitude, which could be an indication that WPI was the dominant ingredient in the rheological response of the WPI/BMP co-stabilized interfaces. With increasing
Fig. 8. Lissajous curves (surface pressure Π versus strain ΔA/A) of the oil/water interfaces stabilized with WPI (A) and BMP (B) at different deformation amplitudes (10%, 20% and 30%). The red curve with solid circle represented 10% deformation, the blue curve with solid circle represented 20% deformation and black curve with solid circle represented 30% deformation in area. The slope of the dashed lines indicates E’ at maximum extension/compression, sometimes also referred to as the secant moduli.
amount of BMP vesicles present in the bulk, the surface elasticity first decreased till 50% BMP but increased slightly with 80% of BMP present, but this increase was small and significant only at 10% deformation. To explore the dynamic responses within one sinusoidal cycle, Lissajous curves of surface pressure versus deformation for both WPI and BMP stabilized interfaces were plotted in Fig. 8. The symmetric elliptic shape of the Lissajous curves at 10% deformation and asymmetric shape at 20% deformation indicated that the maximum strain in the linear viscoelastic region of the oil/water interface was between 10% and 20%. In the nonlinear regime, the response of the interfaces became relatively more viscous for both WPI and BMP, the oil/water interface exhibited strain softening in extension and strain hardening in compression (Sagis & Fischer, 2014). Consistent with the findings in Fig. 7, the elastic modulus of the WPI stabilized oil/water interface at maximum extension and compression (indicated by the dashed lines in Fig. 8, and sometimes referred to as the secant modulus) notably decreased with increasing deformation, which was not observed for BMP. This indicated a significantly different surface organization of adsorbed protein and phospholipids at the oil/water interface. Previous research indicated that BSA and lysozyme could form a heterogeneous protein layer at the oil/water interface by molecular clustering/crowding (McUmber, Larson, Randolph, & Schwartz, 2015). Sagis et al. (2019) observed a similar heterogeneity for WPI (and other proteins) at the air/water interface. They argue that the behaviour of WPI is similar to that of a disordered 2d solid (Sagis et al., 2019), and the behaviour seen in Fig. 8(A) is consistent with that behaviour. When the oil/water interface was expanded with a large amplitude, the network formed in the adsorbed protein layer was broken down, and the resulting decrease in
Fig. 7. The surface elasticity (E’) and surface viscosity (E”) of oil/water interfaces stabilized with WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP at different deformation amplitudes (10%, 20% and 30%). 6
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
surface pressure could not be compensated by newly adsorbed proteins. This leads to a softening of the structure, which is visible as a decrease of the slope of the upper part of the Lissajous curve. The oil/water interface stabilized by the phospholipid vesicle dispersion seemed to be more flexible and mobile, and hence the softening effect is nearly negligible there. Phospholipid vesicles can fuse and reorganize into phospholipid monolayers at the oil/water interface (Brake, Daschner, & Abbott, 2005). The dynamic responses of WPI/BMP co-stabilized oil/ water interfaces to surface dilatation were in between those of WPI and BMP stabilized interfaces (data not shown), closely related to the WPI/ BMP ratio in bulk composition. No clear relationship between initial emulsion stability and surface elasticity could be seen when WPI was compared to BMP, or to any of the different WPI/BMP ratios, since a higher surface elasticity did not necessarily result in a better emulsion stability. However, the rheological data do show that the protein/ phospholipid ratios significantly influenced the initial oil/water interfacial composition and structure, which led to large differences in the stability against flocculation and coalescence of oil droplets when the pH was changed to acidic conditions.
Journal, 61, 228–238. Cornacchia, L., & Roos, Y. H. (2011). Stability of β-carotene in protein-stabilized oil-inwater delivery systems. Journal of Agricultural and Food Chemistry, 59(13), 7013–7020. Gallier, S., Vocking, K., Post, J. A., Van De Heijning, B., Acton, D., Van Der Beek, E. M., et al. (2015). A novel infant milk formula concept: Mimicking the human milk fat globule structure. Colloids and Surfaces B: Biointerfaces, 136, 329–339. García-Moreno, P. J., Horn, A. F., & Jacobsen, C. (2014). Influence of casein–phospholipid combinations as emulsifier on the physical and oxidative stability of fish oil-in-water emulsions. Journal of Agricultural and Food Chemistry, 62(5), 1142–1152. Gassi, J. Y., Blot, M., Beaucher, E., Robert, B., Leconte, N., Camier, B., et al. (2016). Preparation and characterisation of a milk polar lipids enriched ingredient from fresh industrial liquid butter serum: Combination of physico-chemical modifications and technological treatments. International Dairy Journal, 52, 26–34. Haddadian, Z., Eyres, G. T., Bremer, P., & Everett, D. W. (2018). Polar lipid composition of the milk fat globule membrane in buttermilk made using various cream churning conditions or isolated from commercial samples. International Dairy Journal, 81, 138–142. Hernell, O., Timby, N., Domellöf, M., & Lönnerdal, B. (2016). Clinical benefits of milk fat globule membranes for infants and children. The Journal of Pediatrics, 173, S60–S65. He, Q., Zhang, Y., Lu, G., Miller, R., Möhwald, H., & Li, J. (2008). Dynamic adsorption and characterization of phospholipid and mixed phospholipid/protein layers at liquid/liquid interfaces. Advances in Colloid and Interface Science, 140(2), 67–76. Holzmüller, W., Müller, M., Himbert, D., & Kulozik, U. (2016). Impact of cream washing on fat globules and milk fat globule membrane proteins. International Dairy Journal, 59, 52–61. Jiménez-Colmenero, F. (2013). Potential applications of multiple emulsions in the development of healthy and functional foods. Food Research International, 52(1), 64–74. Jukkola, A., Partanen, R., Rojas, O. J., & Heino, A. (2018). Effect of heat treatment and pH on the efficiency of micro-diafiltration for the separation of native fat globules from cream in butter production. Journal of Membrane Science, 548, 99–107. Lecomte, M., Bourlieu, C., Meugnier, E., Penhoat, A., Cheillan, D., Pineau, G., et al. (2015). Milk polar lipids affect in vitro digestive lipolysis and postprandial lipid metabolism in mice. Journal of Nutrition, 145(8), 1770–1777. Le, T. T., Van Camp, J., Rombaut, R., van Leeckwyck, F., & Dewettinck, K. (2009). Effect of washing conditions on the recovery of milk fat globule membrane proteins during the isolation of milk fat globule membrane from milk. Journal of Dairy Science, 92(8), 3592–3603. Liang, L., Zhang, X., Wang, X., Jin, Q., & McClements, D. J. (2018). Influence of dairy emulsifier type and lipid droplet size on gastrointestinal fate of model emulsions: In vitro digestion study. Journal of Agricultural and Food Chemistry, 66(37), 9761–9769. Livney, Y. D., Ruimy, E., Ye, A. M., Zhu, X., & Singh, H. (2017). A milkfat globule membrane-inspired approach for encapsulation of emulsion oil droplets. Food Hydrocolloids, 65, 121–129. Lopez, C., Cauty, C., Rousseau, F., Blot, M., Margolis, A., & Famelart, M.-H. (2017). Lipid droplets coated with milk fat globule membrane fragments: Microstructure and functional properties as a function of pH. Food Research International, 91, 26–37. Magnusson, E., Nilsson, L., & Bergenståhl, B. (2016). Effect of the dispersed state of phospholipids on emulsification Part 1. Phosphatidylcholine. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 506, 794–803. Mathiassen, J. H., Nejrup, R. G., Frøkiær, H., Nilsson, Å., Ohlsson, L., & Hellgren, L. I. (2015). Emulsifying triglycerides with dairy phospholipids instead of soy lecithin modulates gut lipase activity. European Journal of Lipid Science and Technology, 117(10), 1522–1539. McClements, D. J. (2015). Food emulsions principles, practices, and techniques (3rd ed.). Boca Raton: CRC Press 3rd Edition. McUmber, A. C., Larson, N. R., Randolph, T. W., & Schwartz, D. K. (2015). Molecular trajectories provide signatures of protein clustering and crowding at the oil/water interface. Langmuir, 31(21), 5882–5890. Melnikov, S. M., Popp, A. K., Miao, S., Patel, A. R., Flendrig, L. M., & Velikov, K. P. (2017). Colloidal emulsion based delivery systems for steroid glycosides. Journal of Functional Foods, 28, 90–95. Moran-Valero, M. I., Ruiz-Henestrosa, V. M. P., & Pilosof, A. M. R. (2017). Synergistic performance of lecithin and glycerol monostearate in oil/water emulsions. Colloids and Surfaces B: Biointerfaces, 151, 68–75. Morin, P., Britten, M., Jiménez-Flores, R., & Pouliot, Y. (2007). Microfiltration of buttermilk and washed cream buttermilk for concentration of milk fat globule membrane components. Journal of Dairy Science, 90(5), 2132–2140. Morini, M. A., Sierra, M. B., Pedroni, V. I., Alarcon, L. M., Appignanesi, G. A., & Disalvo, E. A. (2015). Influence of temperature, anions and size distribution on the zeta potential of DMPC, DPPC and DMPE lipid vesicles. Colloids and Surfaces B: Biointerfaces, 131, 54–58. Morin, P., Jiménez-Flores, R., & Pouliot, Y. (2004). Effect of temperature and pore size on the fractionation of fresh and reconstituted buttermilk by microfiltration. Journal of Dairy Science, 87(2), 267–273. Morin, P., Jiménez-Flores, R., & Pouliot, Y. (2007). Effect of processing on the composition and microstructure of buttermilk and its milk fat globule membranes. International Dairy Journal, 17(10), 1179–1187. Nejadmansouri, M., Hosseini, S. M. H., Niakosari, M., Yousefi, G. H., & Golmakani, M. T. (2016). Physicochemical properties and storage stability of ultrasound-mediated WPI-stabilized fish oil nanoemulsions. Food Hydrocolloids, 61, 801–811. Phan, T. T. Q., Le, T. T., Van der Meeren, P., & Dewettinck, K. (2014). Comparison of emulsifying properties of milk fat globule membrane materials isolated from different dairy by-products. Journal of Dairy Science, 97(8), 4799–4810. Phan, T. T. Q., Le, T. T., Van de Walle, D., Van der Meeren, P., & Dewettinck, K. (2016). Combined effects of milk fat globule membrane polar lipids and protein concentrate
4. Conclusions In this study, biomimetic milk fat globules (BMFGs) prepared with WPI/BMP mixed systems were mainly subjected to creaming after preparation and the creaming rate of BMP-coated droplets was faster than that of WPI. The macro stability of WPI/BMP co-stabilized BMFGs at different pH values was dominated by WPI. At acidic pH, WPI/BMP ratios strongly affected the coalescence and flocculation of BMFGs. Emulsions that phase separated microscopically but not macroscopically were stabilized by a strong repulsion between flocculated droplets. Based on the large amplitude oscillatory surface rheology, WPI/BMP co-stabilized oil/water interfaces were dominated by WPI. The surface elasticity of WPI stabilized oil/water interfaces showed a significant dependence on deformation amplitudes in the nonlinear regime indicating a significantly different interfacial structure from BMP. A 2d solid-like adsorbed protein layer probably formed at the oil/ water interface by molecular clustering/crowding, instead, the interface stabilized with BMP vesicles seemed to be more flexible and mobile to reorganize under oscillation. No direct correlation was found between the surface elasticity and emulsion stability among different protein/phospholipid ratios, however, the protein/phospholipid ratios influenced the initial oil/water interfacial composition and structure, which led to large differences in the stability and microstructure of oil droplets at acidic pH conditions. Acknowledgement We acknowledge the China Postdoctoral Science Foundation (2018M632724) and National Natural Science Foundation of China (31801500) for funding. References Akbarzadeh, A., Rezaei-Sadabady, R., Davaran, S., Joo, S. W., Zarghami, N., Hanifehpour, Y., et al. (2013). Liposome: Classification, preparation, and applications. Nanoscale Research Letters, 8(1), 102–110. Bergfreund, J., Bertsch, P., Kuster, S., & Fischer, P. (2018). Effect of oil hydrophobicity on the adsorption and rheology of β-lactoglobulin at oil–water interfaces. Langmuir, 34(16), 4929–4936. Brake, J. M., Daschner, M. K., & Abbott, N. L. (2005). formation and characterization of phospholipid monolayers spontaneously assembled at interfaces between aqueous phases and thermotropic liquid crystals. Langmuir, 21(6), 2218–2228. Caro, A. L., Niño, M. R. R., & Patino, J. M. R. (2009). The effect of pH on structural, topographical, and rheological characteristics of β-casein–DPPC mixed monolayers spread at the air–water interface. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 332(2), 180–191. Claumarchirant, L., Cilla, A., Matencio, E., Sanchez-Siles, L. M., Castro-Gomez, P., Fontecha, J., et al. (2016). Addition of milk fat globule membrane as an ingredient of infant formulas for resembling the polar lipids of human milk. International Dairy
7
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
et al. (2018). Instability mechanisms of water-in-oil nanoemulsions with phospholipids: Temporal and morphological structures. Langmuir, 34(2), 572–584. Thompson, A. K., Haisman, D., & Singh, H. (2006). Physical stability of liposomes prepared from milk fat globule membrane and soya phospholipids. Journal of Agricultural and Food Chemistry, 54(17), 6390–6397. Thompson, A. K., Hindmarsh, J. P., Haisman, D., Rades, T., & Singh, H. (2006). Comparison of the structure and properties of liposomes prepared from milk fat globule membrane and soy phospholipids. Journal of Agricultural and Food Chemistry, 54(10), 3704–3711. Waninge, R., Nylander, T., Paulsson, M., & Bergenståhl, B. (2003). Milk membrane lipid vesicle structures studied with Cryo-TEM. Colloids and Surfaces B: Biointerfaces, 31(1), 257–264.
on the stability of oil-in-water emulsions. International Dairy Journal, 52, 42–49. Priyanka, M., Sabine, D., Aman, P., Koen, D., & Christophe, B. (2017). The interfacial properties of various milk fat globule membrane components using Langmuir isotherms. Food Bioscience, 20, 96–103. Sagis, L. M. C., & Fischer, P. (2014). Nonlinear rheology of complex fluid–fluid interfaces. Current Opinion in Colloid & Interface Science, 19(6), 520–529. Sagis, L. M. C., Liu, B., Li, Y., Essers, J., Yang, J., Moghimikheirabadi, A., et al. (2019). Dynamic heterogeneity in complex interfaces of soft interface-dominated materials. Scientific Reports, 9(1), 2938–2949. Smoczyński, M. (2017). Role of phospholipid flux during milk secretion in the mammary gland. Journal of Mammary Gland Biology and Neoplasia, 22(2), 117–129. Sommerling, J.-H., de Matos, M. B. C., Hildebrandt, E., Dessy, A., Kok, R. J., Nirschl, H.,
8
Contents lists available at ScienceDirect
Food Hydrocolloids journal homepage: www.elsevier.com/locate/foodhyd
The influence of protein/phospholipid ratio on the physicochemical and interfacial properties of biomimetic milk fat globules
T
Min Chena,∗, Leonard M.C. Sagisb a b
College of Food Science and Engineering, Ocean University of China, Yushan Road 5, Qingdao, China Physics and Physical Chemistry of Food, Wageningen University, Bornse Weilanden 9, 6708WG, Wageningen, the Netherlands
A R T I C LE I N FO
A B S T R A C T
Keywords: Emulsion Milk fat Protein Phospholipids Oil/water interface
The effect of protein/phospholipid ratio on the physicochemical and interfacial properties of biomimetic milk fat globules (BMFGs) was investigated. Butter milk phospholipid (BMP) vesicles prepared with ultrasonication were mixed with whey protein isolate (WPI) at different protein/phospholipid ratios (w/w). After emulsification, the Turbiscan stability index (TSI) was used to quantify the initial stability of BMFGs. At different pH values, visual observation and CLSM were performed to obtain the macro stability and microstructure of BMFGs. In addition, large amplitude oscillatory surface rheology was applied and analysed with Lissajous curves, to explore the oil/ water interfacial properties of BMFGs in the nonlinear viscoelastic regime. The TSI indicated a faster creaming rate of BMP stabilized droplets compared to those stabilized by WPI. The macro stability of WPI/BMP co-stabilized BMFGs at different pH was dominated by WPI and BMFGs prepared at the WPI/BMP ratio of 20:80 phase separated from pH 3.0 to 5.0. At acidic pH, the WPI/BMP ratios strongly affected the coalescence and flocculation of BMFGs at the micro level. The surface elasticity indicated that WPI dominated the rheological properties of WPI/BMP co-stabilized oil/water interfaces. Besides, the significant dependence of surface elasticity on deformation amplitudes within the nonlinear regime of WPI stabilized oil/water interfaces indicated a clearly different structure from BMP stabilized interfaces. Conclusively, results of this study showed that the physicochemical and interfacial properties of BMFGs were significantly influenced by the protein/phospholipid ratio in the bulk, which needs to be considered in the design and application of biomimetic milk fat globules.
1. Introduction Milk fat, in human breast milk present as milk fat globules (MFGs) with an average diameter of 4–5 μm, contains a core of triglycerides (TAG) covered with a milk fat globule membrane (MFGM). Due to the bio-synthesis process of human MFGs (Smoczyński, 2017), the MFGM exhibits a unique lamellar phospholipid trilayer structure and a multicomponent composition containing polar lipids and functional membrane proteins. These bio-functional components are vital for the brain development, cell metabolism and immune maturation of newborns (Claumarchirant et al., 2016; Hernell, Timby, Domellöf, & Lönnerdal, 2016). Therefore, biomimetic milk fat globules (BMFGs) prepared with MFGM enriched materials have received enormous attention recently (Gallier et al., 2015; Livney, Ruimy, Ye, Zhu, & Singh, 2017; Lopez et al., 2017), often with the aim of developing improved infant milk formulas. The gastrointestinal fate of emulsions can be modulated by the dairy emulsifier composition and the structure of oil droplets (Lecomte et al.,
∗
2015; Liang, Zhang, Wang, Jin, & McClements, 2018; Mathiassen et al., 2015). The composition of MFGM enriched ingredients (including butter serum, butter milk, MFGM polar lipid fraction and MFGM extracts etc.), with respect to the content of milk proteins, polar lipids and membrane proteins can be rather different due to variations in preparation methods (Gassi et al., 2016; Haddadian, Eyres, Bremer, & Everett, 2018; Holzmüller, Müller, Himbert, & Kulozik, 2016; Jukkola, Partanen, Rojas, & Heino, 2018; Le, Van Camp, Rombaut, van Leeckwyck, & Dewettinck, 2009). A comparison of emulsifying properties of MFGM materials isolated from different dairy by-products has demonstrated that not only the polar lipid content but also the presence of other components (e.g., whey proteins, caseins, MFGM-specific proteins, and minerals) determine their emulsifying properties Phan, Le, Van der Meeren, & Dewettinck, 2014). BMFGs have been found to be coated solely with milk proteins, solely with MFGM polar lipids and membrane proteins, or with their mixtures. This effect of the multicomponent composition of MFGM enriched ingredients on their emulsifying and interfacial properties has so far not been studied
Corresponding author. E-mail addresses: [email protected], [email protected] (M. Chen).
https://doi.org/10.1016/j.foodhyd.2019.105179 Received 12 April 2019; Received in revised form 18 June 2019; Accepted 21 June 2019 Available online 22 June 2019 0268-005X/ © 2019 Elsevier Ltd. All rights reserved.
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
(Bergfreund, Bertsch, Kuster, & Fischer, 2018). Therefore, to mimic the lipid composition and structure of breast milk, an oil formulation including five kinds of oils was used to prepare BMFGs in this research. Vesicles of milk polar lipids (BMP) were prepared with ultrasonication. Then WPI was mixed with vesicle dispersions to prepare WPI/BMP mixtures containing different fraction of BMP (0%, 20%, 50%, 80%, 100%, w/w). After the preparation of BMFGs, the emulsion stability was assessed with a turbiscan. At different pH, the stability and morphology of oil droplets was measured by visual appearance and confocal laser scanning microscopy (CLSM), respectively. In addition, the interfacial properties including adsorption kinetics and linear/nonlinear dilatational surface rheology were characterized with an automatic drop tensiometer. It was hypothesized that varying the protein/ phospholipid ratios would lead to a different oil/water interfacial composition and structure, which would affect the initial stability, electrostatic properties and morphology of biomimetic milk fat globules at various pH conditions. A deeper understanding of the relation between stabilizer composition and the physicochemical and interfacial properties of model milk fat emulsions can supply a theoretical basis for the development of better infant formulas.
systematically. Whey protein isolate (WPI) has been intensively applied in food and medical emulsions, for example, in beverages, products with encapsulated oils, and functional delivery systems for people with special medical conditions (Cornacchia & Roos, 2011; Jiménez-Colmenero, 2013; Melnikov et al., 2017; Nejadmansouri, Hosseini, Niakosari, Yousefi, & Golmakani, 2016). Most of the stabilizing ingredients for oil emulsification in infant formula production also contain more than 50% of WPI. Phospholipids are the second abundant ingredient. Due to their amphiphilic property, phospholipids tend to form micelles, vesicles and lamellar assemblies in aqueous solutions. The form in which they are dispersed, the composition of phospholipid species, and their fatty acid composition will significantly influence the emulsifying properties of phospholipids (Magnusson, Nilsson, & Bergenståhl, 2016). In protein/ phospholipid mixed systems, the composition of phospholipid species and their fatty acid composition can influence the interaction between proteins and phospholipids, thereby modulating their emulsifying properties (García-Moreno, Horn, & Jacobsen, 2014). It was reported that MFGM protein concentrate exhibits better emulsifying/stabilising properties compared to phospholipids (Phan, Le, Van de Walle, Van der Meeren, & Dewettinck, 2016). However, the effect of bulk composition (protein/phospholipid ratios) of emulsifiers on the physicochemical properties of BMFGs (for instance, the interaction and microstructure of droplets at various electrostatic conditions) is not clear yet. In addition, how the properties of oil/water interfaces stabilized by MFGM ingredients are affected by their bulk composition is also not fully understood. Recently, the oil/water interfacial properties of three fractions of MFGM enriched materials were studied with a Langmuir trough and results indicated that the surface adsorption kinetics vary significantly with different protein and polar lipid content in the fraction (Priyanka, Sabine, Aman, Koen, & Christophe, 2017). The studied samples were a mixture of milk proteins, polar lipids and membrane proteins. The exact role of each component on the oil/water interfacial structure was not clearly elucidated. At air/water interfaces stabilized by protein/phospholipid mixtures, it was observed that the interaction between proteins and phospholipids at the interface was significantly influenced by pH, which led to different interfacial structures (Caro, Niño, & Patino, 2009). Besides planar interfaces, pendant/rising drop tensiometry has been widely applied in the characterization of interfacial properties of curved oil/water interfaces in correlation to emulsifying properties. For pure phospholipids, the dispersed form of phospholipid molecules, including micelles, vesicles or lamellar assemblies can influence the adsorption kinetics and surface load of phospholipids at the oil/water interface (Magnusson et al., 2016). The rearrangement of phospholipid molecules between the oil/water interface and bulk phase during the evolution of the emulsions strongly dominates the emulsion stability against coalescence and Ostwald ripening (Sommerling et al., 2018). In mixed systems of polar lipids and proteins, a “skin-like” folded film consisting of a phospholipid/protein complex layer can be formed at the water/chloroform interface of a pendant drop, and its formation is accelerated by the co-adsorption of proteins and phospholipids (He et al., 2008). Until now, studies on dilatational surface rheology of interfaces stabilized by proteins and phospholipids were mostly conducted in a linear regime, although in reality the destabilization of emulsions, in particular coalescence, involves large deformations of the interface which occurs in the nonlinear viscoelastic regime. A characterization of the nonlinear surface rheology of oil/water interfaces stabilized by phospholipid/protein mixed systems has so far not been performed. The aim of this study is to determine the effect of protein/phospholipid bulk composition of mixtures of WPI and BMP on the stability and physicochemical properties of biomimetic milk fat globules prepared with these mixtures, and link this effect to properties of the oil/ water interfaces. The adsorption and rheological properties of WPI at the oil/water interface are greatly affected by oil hydrophobicity
2. Materials and methods 2.1. Materials Buttermilk powder was purchased from Dairygold Co., Ltd. (Cork, Ireland). Whey protein isolate (WPI) BiPRO (Protein > 90%) was obtained from Davisco Foods International, Inc. (Le Sueur, Minnesota, USA). Medium chain triacylglycerol MIGLYOL812N was obtained from IOI Oleo Chemicals (Shanghai, China). Nile Red (5H-Benzo α-phenoxazine-5-one, 9-diethylamino) and FITC were purchased from SigmaAldrich (St. Louis, USA). OPO structured lipid was supplied by Yihai Kerry (Beijing, China). DHA algal oil was obtained from Xiamen Huison biotech Co., Ltd. (Xiamen, China) and ARA enriched oil from mortierella alpine were supplied by Xiamen Kingdomway Group Company (Xiamen, China). Coconut oil and sunflower oil were purchased from the local supermarket. Phospholipid standards including phosphatidylcholine (PC), phosphotidyl ethanolamine (PE), phosphotidylserine (PS), sphingomyelin (SM) and phosphatidylinositol (PI) were purchased from Sigma-Aldrich (St. Louis, USA). The other analytic chemicals were bought from Sinopharm Chemical Reagent Co., Ltd. (Beijing, China). 2.2. Methods 2.2.1. Extraction of milk phospholipids from reconstituted buttermilk Buttermilk powder of 2.0 kg was reconstituted in 6.0 L MilliQ water and prewarmed to 40 °C before their pH was adjusted to 4.6 with 2.0 M HCl. Subsequently, the solutions were centrifuged twice at 4745 g for 20 min each. After centrifugation, the obtained supernatant (around 2.6 L) concentrated to 650 mL with a vacuum rotatory evaporator at 60 °C. This was mixed with 2.1 L of chloroform/methanol 2:1 and centrifuged at 4745 g for 5 min. The chloroform layer, filtered through membranes with a pore size of 0.45 μm, was further centrifuged at 7104 g for 5 min. The purified lipid fraction was dried under evaporation at 60 °C and dissolved in cold acetone afterwards. The precipitate in cold acetone was collected, dissolved in chloroform, dried under vacuum rotatory evaporation, and designated as BMP. Their phospholipid composition was further characterized with HPLC-ELSD. 2.2.2. Emulsion preparation BMP and whey protein isolate (WPI) were reconstituted in MilliQ at 10.0 mg/mL respectively and stirred for a whole night. Sodium azide was added at 0.02% (w/w) as preservative. To prepare phospholipid vesicles, the BMP solution was blended with an Ultra-Turrax T25 (IKA, Staufen, Germany) at 13200 rpm for 5 min, and subsequently processed with an ultrasound sonicater (Xinzhi biotechnology Co. Ltd, Ningbo, 2
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
et al., 2017). The staining of samples was briefly as follows: FITC was used to stain proteins. FITC solution of 20 μL (1.0 mg/mL in DMSO) was add to 1.0 mL of sample. Nile Red was used to stain the triglycerides, prepared at a concentration of 42 μg/mL in acetone. Nile Red solution of 10 μL was added to 1.0 mL of sample. Emulsions labelled with FITC and Nile red were kept in the dark for at least 1 h at room temperature prior to observation. The observation was performed using a 60 × (numerical aperture NA 1.4) oil immersion objective with an argon laser (Excitation wavelength of 488 nm, with emission detected between 500 nm and 545 nm) and a diode laser (Excitation wavelength of 559 nm, and emission detected between 570 nm and 625 nm) in a sequential scanning mode at room temperature. Two dimensional images were captured at a resolution of 512 × 512 pixels by further zooming in 1.5 × .
China) at 400 W for 15 min. WPI and polar lipid vesicle mixtures containing various fraction of BMP (0%, 20%, 50%, 80% and 100%, w/w) at a fixed total concentration of 2.0 mg/mL were prepared and prewarmed to 50 °C in a water bath. The oil fraction prewarmed to 50 °C was mixed with emulsifiers at a ratio of 1:9 (v/v), and was pre-homogenized with an Ultra-Turrax T25 at 13200 rpm for 2 min. Subsequently, the coarse emulsion was sonicated at 400 W for 5 min. The composition of the oil was chosen to closely mimic the TAG core of breast milk fat globules: OPO structured lipid (44.0%, w/w), coconut oil (13.3%, w/w), sunflower oil (38.5%, w/w), DHA algal oil (1.6%, w/ w) and ARA enriched oil from mortierella alpine (2.6%, w/w). Based on this resemblance, we consider these model fat globules as biomimetic milk fat globules. 2.2.3. Transmission electronic microscopy (TEM) The morphology of BMP vesicles was determined with a JEM 2100 (JEOL, Japan) equipped with a CCD camera ORIUS SC1000 and an accelerating voltage of 200 kV, according to the method described elsewhere (Magnusson et al., 2016), with some modifications. Shortly, samples of 10 μL was deposited on a lacey carbon film supported by a copper grid for 60 s, and excess liquid was removed with filter papers. Subsequently, the above procedure was repeated with 10 μL of phosphotungstic acid (2.0%) solution. The grid was kept at room temperature and air dried before observation. Images were captured with a magnification of 12000 × .
2.2.7. Oil/water interfacial properties of WPI/BMP mixtures The oil/water interfaces stabilized with WPI/BMP mixed systems were investigated with a Tracker automatic drop tensiometer (TECLIS Instruments, Civrieux d’Azergues, France). A rising droplet of caprylic/ capric triglyceride (MIGLYOL812N, IOI Oleo Chemicals, Hamburg, Germany) was created at the tip of a motorized syringe, in the aqueous phase. Surface pressure was monitored without oscillation for the first 100 s at room temperature (15–18 °C). Subsequently, a sinusoidal dilatational deformation was applied to the oil/water interface with an amplitude of 10% for 120 min, after which the interface was considered to be close to equilibrium. After this phase a strain sweep was performed, with three amplitudes of deformation (10%, 20% and 30%) at a frequency of 0.01 Hz. The area and volume of the non-deformed oil droplet were kept constant at 21.6 mm2 and 10.0 μL, respectively. Three cycles were performed at each amplitude and each cycle included five periods separated by one blank period (100 s). At least two measurements were performed for each sample. The dilatational elastic modulus (E’) and viscous modulus (E″) were calculated with the software WDROP 10.10.9.0 (TECLIS Instruments, France). These moduli are calculated based on the intensity and phase of the first harmonic of the Fourier transform of the oscillating surface pressure signal. Any contributions from higher harmonics, resulting from nonlinearity of the response, are neglected, so results of these calculations are accurate only in the linear response regime. To characterize the nonlinear response at large deformation amplitude, Lissajous curves of Π(t) versus ΔA(t)/A0 were plotted, where Π(t) was the oscillating surface pressure and ΔA(t)/A0 was the deformation of the droplet area.
2.2.4. Particle size and zeta potential The particle size of vesicles and oil droplets was measured with a Nanozs90 (Malvern Instruments, Malvern, U.K.) as described elsewhere (Lopez et al., 2017; Moran-Valero, Ruiz-Henestrosa, & Pilosof, 2017). The refractive index was set to 1.333 for water, 1.46 for BMP vesicles and 1.458 (633 nm) for oil droplets, respectively. The volume mean diameter (De Broucker diameter D43, D43 =
∑1n Di4 vi ∑1n Di3 vi
) and z-average were
used to characterize the particle size distribution of vesicles and fat droplets. The zeta potential (ζ-potential) was also determined with Nanozs90. Samples were diluted 20 times with continuous phases for vesicles and 100 times for oil droplets. The initial pH value was 6.11 for BMP-stabilized emulsion, 7.39 for WPI-stabilized emulsion and in-between 6.11 and 7.39 for emulsions stabilized by WPI/BMP mixtures at different ratios. Besides the initial pH values, the pH of oil droplets was adjusted to 2.0, 3.0, 4.0 and 5.0 before measurements. All experiments were performed at 25 °C, and at least in triplicate for each sample.
3. Results and discussion 2.2.5. Emulsion stability assessment 2.2.5.1. Turbiscan stability index (TSI). The destabilization of BMFGs during aging was monitored with a Turbiscan LAB expert (Formulaction, Toulouse, France). The turbiscan stability index (TSI), calculated based on the variation in the backscattering light intensity at different locations of the test tube (Top, middle and bottom), was used to interpret different destabilizing process like creaming, flocculation and coalescence. In this study, freshly prepared emulsions (20.0 mL) were cooled down at room temperature for 30 min, before a 2 hmeasurement at 25 °C. At least 2 tests were conducted for each sample. The TSI of the bottom part (16.7 mm from the vial bottom) represents the stability of studied emulsions.
3.1. Characterization of butter milk phospholipid (BMP) vesicles BMP vesicles were characterized for their phospholipid composition, particle morphology, particle size distribution and zeta-potential. Results from HPLC-ELSD showed that BMP contained 8.1 ± 0.9% of PI (phosphatidylinositol), 23.7 ± 0.3% of PE (phosphatidylethanolamine), 12.2 ± 0.7% of PS (phosphatidylserine), 38.4 ± 2.1% of PC (phosphatidylcholine) and 17.6 ± 0.8% of SM (sphingomyelin), which was consistent with the phospholipid composition of buttermilk in previous research (Morin, Britten, Jiménez-Flores, & Pouliot, 2007; Morin, Jiménez-Flores, & Pouliot, 2004, 2007). The phospholipid composition and salt concentration can influence the dispersed form of phospholipids in aqueous solution. With more than 80% of PE present, phospholipids mainly form microtubules, however, with less PE present, they mainly form spherical vesicles, which are unilamellar in water but multilamellar in salt or protein buffer (Waninge, Nylander, Paulsson, & Bergenståhl, 2003). According to the TEM image in Fig. 1, polydisperse spherical particles were observed after sonication with a similar morphology as found by (Thompson, Haisman, & Singh, 2006; Thompson, Hindmarsh, Haisman, Rades, & Singh, 2006). A clear delineation in terms of particle size between small unilamellar vesicles
2.2.5.2. Emulsion stability at different pH. The pH of studied emulsions was first measured with a pH meter, subsequently adjusted to 2.0, 3.0, 4.0 and 5.0 respectively, and left undisturbed for a whole night. The stability of emulsions at different pH values was compared by their visual appearance in the images taken with a common camera. 2.2.6. Confocal laser scanning microscopy (CLSM) The microstructure of oil droplets was investigated with a Fluo View TM FV1000 (Olympus, Japan) according to previous research (Lopez 3
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
Fig. 1. Particle size distribution in volume fraction (%) and the TEM image of butter milk phospholipid (BMP) vesicle dispersion with a scale bar at 200 nm.
Fig. 2. The turbiscan stability index (TSI) of WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP.
(SUVs) and large unilamellar vesicles (LUVs) is still debated (Akbarzadeh et al., 2013). The BMP vesicles shown in Fig. 1 were roughly in a size range of 20–1000 nm. Therefore, the BMP dispersion prepared in this study contained mainly SUVs and a small fraction of LUVs. The volume mean diameter (D43) of BMP vesicles was 122.7 ± 16.8 nm and the zeta potential was −46.4 ± 1.2 mV. The strong negative charge was closely related to the phospholipid composition of BMP (Morini et al., 2015).
3.2. Formation and stability of biomimetic milk fat globules (BMFGs) The BMFGs were prepared with a special oil formulation, and emulsified by WPI/BMP mixed systems at different protein/phospholipid ratios. The turbiscan stability index (TSI) was determined to evaluate emulsion stability (Fig. 2). The data indicated that creaming was the main destabilizing factor of emulsions prepared with WPI, BMP and their mixtures. For the WPI/BMP mixed systems, there was no significant increase in TSI with increasing BMP vesicle concentration in the bulk, up to 50% (w/w). However, when the mixture was composed of 80% of BMP vesicles and 20% of WPI, the emulsifying capability seemed to be less effective than pure BMP dispersions. This was probably caused by the competitive adsorption between WPI and BMP vesicles at the oil/water interface (Phan et al., 2016). The TSI of emulsion stabilized with pure BMP vesicle dispersion was about twice of that for WPI, which indicated that the creaming rate of the former was faster than the latter. As shown in Fig. 3, the volume mean diameter and zaverage particle size of the emulsions became significantly larger when the BMP vesicles fraction increased to 80% of the mixture. To a first approximation, the stability of a food emulsion to creaming can be estimated using Stokes law, which predicts a creaming velocity of the droplets that scales with the square of the radius of the droplets (McClements, 2015). The mean radius of WPI-coated droplets and BMPcoated droplets was approximately equal to 250 nm and 350 nm respectively. Then according to Stokes law the creaming rate of droplets in the BMP stabilized emulsion would be a factor of 1.96 compared to droplets in the WPI stabilized emulsion, which was well consistent with the TSI results from the turbiscan. This implies the emulsions were unstable to creaming, but no significant flocculation or coalescence occurred shortly after the emulsion preparation.
Fig. 3. The volume mean diameter (D43) and Z-average particle size of biomimetic milk fat globules (BMFGs) prepared with WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP.
3.3. Effect of WPI/BMP ratio on the physicochemical properties of BMFGs at different pH The visual appearance of BMFGs before and after pH adjustment was shown in Fig. 4(a-e). With different WPI/BMP ratios in the bulk, emulsions phase separated at different pH values. At pH 5.0, close to the isoelectric point of WPI (PI = 4.8), all WPI/BMP co-stabilized BMFGs phase separated, but BMFGs stabilized by pure BMP did not. At 4
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
Fig. 4. Visual appearance of BMFGs prepared with WPI/BMP mixed systems containing different amount of BMP at different pH values: (a) initial pH, (b) 5.0, (c) 4.0, (d) 3.0 and (e) 2.0. For each image, test tubes from left to right represented 0%, 20%, 50%, 80% and 100% of BMP present in bulk.
Fig. 5. CLSM images of BMFGs prepared with WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP at pH 3.0 (a–e) and 4.0 (f–j). The protein was stained with FITC and the triglyceride was stained with Nile Red. In the images, green referred to proteins, red referred to oil and yellowish colour referred to the area where proteins and oil were merged together. The scale bar was 10 μm.
pH 2.0, close to the isoelectric point of the phospholipids (PI = 2.0), all WPI/BMP co-stabilized emulsions remained stable, but BMFGs coated with only BMP phase separated. This indicated that the stability of BMFGs at acidic conditions was dominated by WPI for WPI/BMP mixed systems. Visual phase separation occurred to BMFGs in the pH range of 3.0–5.0 when they were prepared with 20% of WPI and 80% of BMP. The specific pH where phase separation occurs was dependent on the protein/phospholipid ratio in the bulk, which needs to be taken into account in food processing and design. Observation by CLSM was further conducted to determine the microstructure of fat droplets at pH 3.0 and 4.0 (Fig. 5 a-j). At pH 4.0, more coalescence occurred and droplets were generally larger compared to pH 3.0. So the oil/water interface was apparently less stable against coalescence at this pH. The organization of droplets was different in visually phase-separated emulsions. Micron-sized spherical droplets without flocculation were observed for pure BMP stabilized BMFGs. When there was WPI present, large protein aggregates with attached droplets were formed (Fig. 5 d and i) and irregular shaped oil lumps were observed (Fig. 5 h), probably due to bridging flocculation. BMP and WPI have opposite charges at pH 3.0 and 4.0 which can give attractive electrostatic interactions between negative phospholipid patches on the droplet interface, and positively charged WPI patches on the interface, or between WPI and BMP in the bulk. For those emulsions that did not phase separate at pH 3.0 and 4.0 visually, a clear phase separation occurred at the micro level (Fig. 5 b, c and g). These differences in the organization of droplets at the micro and macro level were closely related to the interactions between droplets at different pH values. As shown in Fig. 6, the zeta potential of freshly prepared BMFGs was between −46.7 and −58.2 mV, so they were highly negatively charged around neutral pH. This high negative charge induced a strong repulsion between droplets which kept them from aggregating. At pH 3.0, BMFGs appeared to be highly positively charged when they were prepared with 50% BMP or less in the bulk solution. The samples containing 20% and 50% BMP showed a limited degree of flocculation, most likely due to bridging flocculation mediated by negatively-charged BMP. For droplets coated with 80% and
Fig. 6. Zeta potential (ζ) of BMFGs prepared with WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP at different pH values (from initial pH to 2.0).
100% BMP the zeta potentials are around +20 and −20 mV respectively, so these systems were only marginally stable at pH 3.0, which may have led to the formation of the larger aggregates at 80% BMP. At pH 4.0 the 100% BMP stabilized droplets have a zeta potential below −40 mV, and were stabilized by strong electrostatic repulsion. The 5
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
50% and 80% BMP samples have a very low zeta potential and were extensively flocculated. As a result of a weaker interface, these systems also showed significant coalescence. The 20% BMP and pure WPI systems have a highly positive zeta potential, which for the latter resulted in a stable emulsion, and the former into a network structure most likely formed by bridging interactions mediated by negative BMPs. From these observations it is clear that the stability of BMFGs against aggregation and coalescence at different pH values is strongly affected by the WPI/BMP ratios. Most likely this dependence is coming from the effect of this ratio on the initial oil/water interfacial composition and structure. Therefore, the effect of bulk composition on the oil/water interfacial composition and microstructure was investigated. 3.4. Oil/water interfaces stabilized with WPI/BMP mixed system The interfacial properties of WPI, BMP and WPI/BMP mixed systems were determined with the rising drop method using an automatic drop tensiometer. The interfacial tension of the clean oil/water interface was 20.69 ± 0.20 mN/m. The surface pressure was in a range of 10.87 ± 0.05–12.19 ± 0.05 mN/m after 2 h-monitoring for all studied samples. With 80% of BMP present, the surface pressure was slightly lower than the other samples, which was in line with the larger droplet size in Fig. 3. After this initial phase, the dependence of the surface elastic modulus E’ and viscous modulus E” of the oil/water interface on deformation amplitude was measured, by applying a sinusoidal oscillatory deformation, and the results based on a first harmonic analysis were presented in Fig. 7. The E” of the oil/water interfaces was below 5.0 mN/m for all samples, and not significantly different at various WPI/BMP ratios. For all samples E’ was significantly higher than E”, so at this particular frequency (0.01 Hz), the interfaces appeared to have a predominantly elastic response. The pure WPI-stabilized oil/water interface was significantly more elastic than the pure BMP-stabilized one. In addition, the elastic modulus of the pure WPIstabilized interface decreased significantly when the deformation amplitude increased from 10% to 30%, whereas the moduli of the pure BMP stabilized interface appeared to be independent of strain amplitude, and appeared to be measured in the linear response regime. For the WPI/BMP co-stabilized oil/water interfaces, E’ also showed a significant dependence on deformation amplitude, which could be an indication that WPI was the dominant ingredient in the rheological response of the WPI/BMP co-stabilized interfaces. With increasing
Fig. 8. Lissajous curves (surface pressure Π versus strain ΔA/A) of the oil/water interfaces stabilized with WPI (A) and BMP (B) at different deformation amplitudes (10%, 20% and 30%). The red curve with solid circle represented 10% deformation, the blue curve with solid circle represented 20% deformation and black curve with solid circle represented 30% deformation in area. The slope of the dashed lines indicates E’ at maximum extension/compression, sometimes also referred to as the secant moduli.
amount of BMP vesicles present in the bulk, the surface elasticity first decreased till 50% BMP but increased slightly with 80% of BMP present, but this increase was small and significant only at 10% deformation. To explore the dynamic responses within one sinusoidal cycle, Lissajous curves of surface pressure versus deformation for both WPI and BMP stabilized interfaces were plotted in Fig. 8. The symmetric elliptic shape of the Lissajous curves at 10% deformation and asymmetric shape at 20% deformation indicated that the maximum strain in the linear viscoelastic region of the oil/water interface was between 10% and 20%. In the nonlinear regime, the response of the interfaces became relatively more viscous for both WPI and BMP, the oil/water interface exhibited strain softening in extension and strain hardening in compression (Sagis & Fischer, 2014). Consistent with the findings in Fig. 7, the elastic modulus of the WPI stabilized oil/water interface at maximum extension and compression (indicated by the dashed lines in Fig. 8, and sometimes referred to as the secant modulus) notably decreased with increasing deformation, which was not observed for BMP. This indicated a significantly different surface organization of adsorbed protein and phospholipids at the oil/water interface. Previous research indicated that BSA and lysozyme could form a heterogeneous protein layer at the oil/water interface by molecular clustering/crowding (McUmber, Larson, Randolph, & Schwartz, 2015). Sagis et al. (2019) observed a similar heterogeneity for WPI (and other proteins) at the air/water interface. They argue that the behaviour of WPI is similar to that of a disordered 2d solid (Sagis et al., 2019), and the behaviour seen in Fig. 8(A) is consistent with that behaviour. When the oil/water interface was expanded with a large amplitude, the network formed in the adsorbed protein layer was broken down, and the resulting decrease in
Fig. 7. The surface elasticity (E’) and surface viscosity (E”) of oil/water interfaces stabilized with WPI/BMP mixed systems containing 0%, 20%, 50%, 80% and 100% of BMP at different deformation amplitudes (10%, 20% and 30%). 6
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
surface pressure could not be compensated by newly adsorbed proteins. This leads to a softening of the structure, which is visible as a decrease of the slope of the upper part of the Lissajous curve. The oil/water interface stabilized by the phospholipid vesicle dispersion seemed to be more flexible and mobile, and hence the softening effect is nearly negligible there. Phospholipid vesicles can fuse and reorganize into phospholipid monolayers at the oil/water interface (Brake, Daschner, & Abbott, 2005). The dynamic responses of WPI/BMP co-stabilized oil/ water interfaces to surface dilatation were in between those of WPI and BMP stabilized interfaces (data not shown), closely related to the WPI/ BMP ratio in bulk composition. No clear relationship between initial emulsion stability and surface elasticity could be seen when WPI was compared to BMP, or to any of the different WPI/BMP ratios, since a higher surface elasticity did not necessarily result in a better emulsion stability. However, the rheological data do show that the protein/ phospholipid ratios significantly influenced the initial oil/water interfacial composition and structure, which led to large differences in the stability against flocculation and coalescence of oil droplets when the pH was changed to acidic conditions.
Journal, 61, 228–238. Cornacchia, L., & Roos, Y. H. (2011). Stability of β-carotene in protein-stabilized oil-inwater delivery systems. Journal of Agricultural and Food Chemistry, 59(13), 7013–7020. Gallier, S., Vocking, K., Post, J. A., Van De Heijning, B., Acton, D., Van Der Beek, E. M., et al. (2015). A novel infant milk formula concept: Mimicking the human milk fat globule structure. Colloids and Surfaces B: Biointerfaces, 136, 329–339. García-Moreno, P. J., Horn, A. F., & Jacobsen, C. (2014). Influence of casein–phospholipid combinations as emulsifier on the physical and oxidative stability of fish oil-in-water emulsions. Journal of Agricultural and Food Chemistry, 62(5), 1142–1152. Gassi, J. Y., Blot, M., Beaucher, E., Robert, B., Leconte, N., Camier, B., et al. (2016). Preparation and characterisation of a milk polar lipids enriched ingredient from fresh industrial liquid butter serum: Combination of physico-chemical modifications and technological treatments. International Dairy Journal, 52, 26–34. Haddadian, Z., Eyres, G. T., Bremer, P., & Everett, D. W. (2018). Polar lipid composition of the milk fat globule membrane in buttermilk made using various cream churning conditions or isolated from commercial samples. International Dairy Journal, 81, 138–142. Hernell, O., Timby, N., Domellöf, M., & Lönnerdal, B. (2016). Clinical benefits of milk fat globule membranes for infants and children. The Journal of Pediatrics, 173, S60–S65. He, Q., Zhang, Y., Lu, G., Miller, R., Möhwald, H., & Li, J. (2008). Dynamic adsorption and characterization of phospholipid and mixed phospholipid/protein layers at liquid/liquid interfaces. Advances in Colloid and Interface Science, 140(2), 67–76. Holzmüller, W., Müller, M., Himbert, D., & Kulozik, U. (2016). Impact of cream washing on fat globules and milk fat globule membrane proteins. International Dairy Journal, 59, 52–61. Jiménez-Colmenero, F. (2013). Potential applications of multiple emulsions in the development of healthy and functional foods. Food Research International, 52(1), 64–74. Jukkola, A., Partanen, R., Rojas, O. J., & Heino, A. (2018). Effect of heat treatment and pH on the efficiency of micro-diafiltration for the separation of native fat globules from cream in butter production. Journal of Membrane Science, 548, 99–107. Lecomte, M., Bourlieu, C., Meugnier, E., Penhoat, A., Cheillan, D., Pineau, G., et al. (2015). Milk polar lipids affect in vitro digestive lipolysis and postprandial lipid metabolism in mice. Journal of Nutrition, 145(8), 1770–1777. Le, T. T., Van Camp, J., Rombaut, R., van Leeckwyck, F., & Dewettinck, K. (2009). Effect of washing conditions on the recovery of milk fat globule membrane proteins during the isolation of milk fat globule membrane from milk. Journal of Dairy Science, 92(8), 3592–3603. Liang, L., Zhang, X., Wang, X., Jin, Q., & McClements, D. J. (2018). Influence of dairy emulsifier type and lipid droplet size on gastrointestinal fate of model emulsions: In vitro digestion study. Journal of Agricultural and Food Chemistry, 66(37), 9761–9769. Livney, Y. D., Ruimy, E., Ye, A. M., Zhu, X., & Singh, H. (2017). A milkfat globule membrane-inspired approach for encapsulation of emulsion oil droplets. Food Hydrocolloids, 65, 121–129. Lopez, C., Cauty, C., Rousseau, F., Blot, M., Margolis, A., & Famelart, M.-H. (2017). Lipid droplets coated with milk fat globule membrane fragments: Microstructure and functional properties as a function of pH. Food Research International, 91, 26–37. Magnusson, E., Nilsson, L., & Bergenståhl, B. (2016). Effect of the dispersed state of phospholipids on emulsification Part 1. Phosphatidylcholine. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 506, 794–803. Mathiassen, J. H., Nejrup, R. G., Frøkiær, H., Nilsson, Å., Ohlsson, L., & Hellgren, L. I. (2015). Emulsifying triglycerides with dairy phospholipids instead of soy lecithin modulates gut lipase activity. European Journal of Lipid Science and Technology, 117(10), 1522–1539. McClements, D. J. (2015). Food emulsions principles, practices, and techniques (3rd ed.). Boca Raton: CRC Press 3rd Edition. McUmber, A. C., Larson, N. R., Randolph, T. W., & Schwartz, D. K. (2015). Molecular trajectories provide signatures of protein clustering and crowding at the oil/water interface. Langmuir, 31(21), 5882–5890. Melnikov, S. M., Popp, A. K., Miao, S., Patel, A. R., Flendrig, L. M., & Velikov, K. P. (2017). Colloidal emulsion based delivery systems for steroid glycosides. Journal of Functional Foods, 28, 90–95. Moran-Valero, M. I., Ruiz-Henestrosa, V. M. P., & Pilosof, A. M. R. (2017). Synergistic performance of lecithin and glycerol monostearate in oil/water emulsions. Colloids and Surfaces B: Biointerfaces, 151, 68–75. Morin, P., Britten, M., Jiménez-Flores, R., & Pouliot, Y. (2007). Microfiltration of buttermilk and washed cream buttermilk for concentration of milk fat globule membrane components. Journal of Dairy Science, 90(5), 2132–2140. Morini, M. A., Sierra, M. B., Pedroni, V. I., Alarcon, L. M., Appignanesi, G. A., & Disalvo, E. A. (2015). Influence of temperature, anions and size distribution on the zeta potential of DMPC, DPPC and DMPE lipid vesicles. Colloids and Surfaces B: Biointerfaces, 131, 54–58. Morin, P., Jiménez-Flores, R., & Pouliot, Y. (2004). Effect of temperature and pore size on the fractionation of fresh and reconstituted buttermilk by microfiltration. Journal of Dairy Science, 87(2), 267–273. Morin, P., Jiménez-Flores, R., & Pouliot, Y. (2007). Effect of processing on the composition and microstructure of buttermilk and its milk fat globule membranes. International Dairy Journal, 17(10), 1179–1187. Nejadmansouri, M., Hosseini, S. M. H., Niakosari, M., Yousefi, G. H., & Golmakani, M. T. (2016). Physicochemical properties and storage stability of ultrasound-mediated WPI-stabilized fish oil nanoemulsions. Food Hydrocolloids, 61, 801–811. Phan, T. T. Q., Le, T. T., Van der Meeren, P., & Dewettinck, K. (2014). Comparison of emulsifying properties of milk fat globule membrane materials isolated from different dairy by-products. Journal of Dairy Science, 97(8), 4799–4810. Phan, T. T. Q., Le, T. T., Van de Walle, D., Van der Meeren, P., & Dewettinck, K. (2016). Combined effects of milk fat globule membrane polar lipids and protein concentrate
4. Conclusions In this study, biomimetic milk fat globules (BMFGs) prepared with WPI/BMP mixed systems were mainly subjected to creaming after preparation and the creaming rate of BMP-coated droplets was faster than that of WPI. The macro stability of WPI/BMP co-stabilized BMFGs at different pH values was dominated by WPI. At acidic pH, WPI/BMP ratios strongly affected the coalescence and flocculation of BMFGs. Emulsions that phase separated microscopically but not macroscopically were stabilized by a strong repulsion between flocculated droplets. Based on the large amplitude oscillatory surface rheology, WPI/BMP co-stabilized oil/water interfaces were dominated by WPI. The surface elasticity of WPI stabilized oil/water interfaces showed a significant dependence on deformation amplitudes in the nonlinear regime indicating a significantly different interfacial structure from BMP. A 2d solid-like adsorbed protein layer probably formed at the oil/ water interface by molecular clustering/crowding, instead, the interface stabilized with BMP vesicles seemed to be more flexible and mobile to reorganize under oscillation. No direct correlation was found between the surface elasticity and emulsion stability among different protein/phospholipid ratios, however, the protein/phospholipid ratios influenced the initial oil/water interfacial composition and structure, which led to large differences in the stability and microstructure of oil droplets at acidic pH conditions. Acknowledgement We acknowledge the China Postdoctoral Science Foundation (2018M632724) and National Natural Science Foundation of China (31801500) for funding. References Akbarzadeh, A., Rezaei-Sadabady, R., Davaran, S., Joo, S. W., Zarghami, N., Hanifehpour, Y., et al. (2013). Liposome: Classification, preparation, and applications. Nanoscale Research Letters, 8(1), 102–110. Bergfreund, J., Bertsch, P., Kuster, S., & Fischer, P. (2018). Effect of oil hydrophobicity on the adsorption and rheology of β-lactoglobulin at oil–water interfaces. Langmuir, 34(16), 4929–4936. Brake, J. M., Daschner, M. K., & Abbott, N. L. (2005). formation and characterization of phospholipid monolayers spontaneously assembled at interfaces between aqueous phases and thermotropic liquid crystals. Langmuir, 21(6), 2218–2228. Caro, A. L., Niño, M. R. R., & Patino, J. M. R. (2009). The effect of pH on structural, topographical, and rheological characteristics of β-casein–DPPC mixed monolayers spread at the air–water interface. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 332(2), 180–191. Claumarchirant, L., Cilla, A., Matencio, E., Sanchez-Siles, L. M., Castro-Gomez, P., Fontecha, J., et al. (2016). Addition of milk fat globule membrane as an ingredient of infant formulas for resembling the polar lipids of human milk. International Dairy
7
Food Hydrocolloids 97 (2019) 105179
M. Chen and L.M.C. Sagis
et al. (2018). Instability mechanisms of water-in-oil nanoemulsions with phospholipids: Temporal and morphological structures. Langmuir, 34(2), 572–584. Thompson, A. K., Haisman, D., & Singh, H. (2006). Physical stability of liposomes prepared from milk fat globule membrane and soya phospholipids. Journal of Agricultural and Food Chemistry, 54(17), 6390–6397. Thompson, A. K., Hindmarsh, J. P., Haisman, D., Rades, T., & Singh, H. (2006). Comparison of the structure and properties of liposomes prepared from milk fat globule membrane and soy phospholipids. Journal of Agricultural and Food Chemistry, 54(10), 3704–3711. Waninge, R., Nylander, T., Paulsson, M., & Bergenståhl, B. (2003). Milk membrane lipid vesicle structures studied with Cryo-TEM. Colloids and Surfaces B: Biointerfaces, 31(1), 257–264.
on the stability of oil-in-water emulsions. International Dairy Journal, 52, 42–49. Priyanka, M., Sabine, D., Aman, P., Koen, D., & Christophe, B. (2017). The interfacial properties of various milk fat globule membrane components using Langmuir isotherms. Food Bioscience, 20, 96–103. Sagis, L. M. C., & Fischer, P. (2014). Nonlinear rheology of complex fluid–fluid interfaces. Current Opinion in Colloid & Interface Science, 19(6), 520–529. Sagis, L. M. C., Liu, B., Li, Y., Essers, J., Yang, J., Moghimikheirabadi, A., et al. (2019). Dynamic heterogeneity in complex interfaces of soft interface-dominated materials. Scientific Reports, 9(1), 2938–2949. Smoczyński, M. (2017). Role of phospholipid flux during milk secretion in the mammary gland. Journal of Mammary Gland Biology and Neoplasia, 22(2), 117–129. Sommerling, J.-H., de Matos, M. B. C., Hildebrandt, E., Dessy, A., Kok, R. J., Nirschl, H.,
8