Behaviour of protein-stabilised emulsions under various physiological conditions

Advances in Colloid and Interface Science 165 (2011) 47–57

Contents lists available at ScienceDirect

Advances in Colloid and Interface Science j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c i s

Behaviour of protein-stabilised emulsions under various physiological conditions Harjinder Singh ⁎, Anwesha Sarkar 1 Riddet Institute, Massey University, Private Bag 11 222, Palmerston North 4442, New Zealand

a r t i c l e

i n f o

Available online 15 February 2011 Keywords: Emulsion Protein Digestion Enzyme Gastro-intestinal tract Lipid digestion

a b s t r a c t Emulsion forms a major part of many processed food formulations. During the past few decades, the physicochemical properties of oil-in-water emulsions under various food processing conditions have been extensively studied. However, over the recent years, interest has turned to understanding the behaviour of emulsions during consumption, i.e. physiological processing. In general, on ingestion, an emulsion is exposed to a relatively narrow range of physical (e.g. shear and temperature) and biochemical (e.g. dilution, pH, pepsin, pancreatin, mucins and bile salts) environments as it passes through the mouth into the stomach and then the intestines. There is currently limited knowledge of the physico-chemical and structural changes, which an emulsion may undergo when it passes through the physiologically active regime. A better understanding of the gastro-intestinal processing of emulsions would allow manipulation of physicochemical and interfacial properties to modulate lipid ingestion, improve bioavailability of lipid soluble nutrients and reduce absorption of saturated fats, cholesterol and trans fats. Food emulsions are commonly stabilised by proteins, as they are not only excellent emulsifiers but also provide nutritional benefits to the product. The effects of digestion conditions on interfacial protein structures are complicated because of potential breakdown of these structures by proteolytic enzymes of the gastrointestinal tract. Studies dealing directly with the behaviour of protein-based emulsions under digestion conditions are very limited. This paper provides an overview of the behaviour of oil-in-water emulsions stabilised with globular proteins, namely lactoferrin and β-lactoglobulin. Recent advances in understanding the interactions between interfacial proteins on oil droplets and various physiological materials (e.g. enzymes and bile salts) in in vitro digestion systems are considered. Major emphasis is placed on the recent work carried out in our laboratory at Massey University on the behaviour of milk protein based emulsions (lactoferrin or β-lactoglobulin) during their passage through the gastro-intestinal tract. © 2011 Elsevier B.V. All rights reserved.

Contents 1. 2. 3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Protein-stabilised emulsions . . . . . . . . . . . . . . . . . . . . . . . . Interactions of protein-stabilised emulsions under physiological conditions . . 3.1. Oral conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Gastric conditions . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Intestinal conditions . . . . . . . . . . . . . . . . . . . . . . . . . 4. Sequential processing of protein-based emulsions in an oral-to-gastrointestinal 4.1. Droplet size and microstructure . . . . . . . . . . . . . . . . . . . 4.2. ζ-Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Overall mechanism . . . . . . . . . . . . . . . . . . . . . . . . . 5. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . model . . . . . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

47 48 49 49 50 50 51 51 52 53 55 56

1. Introduction ⁎ Corresponding author. Tel.: + 64 6 356 4401; fax: + 64 6 350 5655. E-mail address: [email protected] (H. Singh). 1 Current address: Nestlé Research Center, Vers-Chez-Les Blanc, CH-1000 Lausanne 26, Switzerland. 0001-8686/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.cis.2011.02.001

Dietary lipids, fats (if they are solid) and oils (if they are liquid), derived from plant and animal sources, perform many important

48

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57

functions in the body, such as provide essential fatty acids, act as carriers of fat-soluble vitamins (vitamins A, D, E and K) and provide a concentrated source of energy [1,2]. Although lipids are essential components of daily foods, their overconsumption can lead to serious medical conditions. Recent concerns about obesity, atherosclerosis and other food-linked diseases have raised considerable challenges to designing healthier foods with a focus on reducing fat, cholesterol, salt and carbohydrates without sacrificing the organoleptic properties of the food [2–4]. The consumption of low fat foods has become a common response to public health concerns. As a result, over the past few decades, reduced-fat foods and fat substitutes have increasingly become a part of food business because of consumer demand. Unfortunately, many, if not most, of the low fat foods do not meet the consumer's sensorial expectations, as reduction in fat content adversely affects flavour release, creaminess, mouthfeel and texture perception [5–8]. In recent years, there has been an upsurge in efforts to understand how food structure influences the rates of digestion of macronutrients, with a major focus on lipid digestion [2,3,9,10]. This research is being undertaken with a view to developing novel foods that regulate calorie intake (lipid is the most energy dense), provide increased satiety responses, provide controlled lipid digestion and/or deliver bioactive molecules. Hence, knowledge concerning the behaviour of food emulsions during gastrointestinal transit is a subject of paramount importance for developing specific strategies to modulate lipid digestion. This paper provides an overview of the behaviour of oil-in-water emulsions stabilised with protein during their passage through the gastrointestinal tract. Recent advances in understanding the interactions between interfacial proteins on oil droplets and various physiological materials (e.g. enzymes and bile salts) in in vitro digestion systems are considered. Major emphasis is placed on the work carried out in our laboratory at Massey University. 2. Protein-stabilised emulsions Generally, food emulsions are prepared using high shear equipment, such as colloid mills, high speed blenders and high pressure valve homogenisers, that mixes an oil phase and an aqueous phase together in the presence of a surface-active agent [11]. The basic procedure is to force a coarse mixture of oil phase and aqueous phase through a narrow slit under the action of high pressure, resulting in cavitation, intense laminar shear flow and turbulence. At the same time, the surface-active agents, such as emulsifiers, because they are structurally amphiphilic molecules (having both hydrophobic and hydrophilic moieties), are adsorbed at the oil−water interface [12], creating a stabilising interfacial layer at the oil droplet surface and leading to the generation of fine uniformly dispersed droplets. In food systems, phospholipids, monoacylglycerols and diacylglyerols often feature as emulsifiers, but emulsions are more commonly stabilised by proteins. Examples of protein-stabilised food emulsions include milk, cream, salad dressings, mayonnaise, dairy desserts and ice cream. Milk proteins, in both soluble form and dispersed form, have been known to be excellent emulsifiers for many decades because of their amphiphilic nature [13,14]. They exhibit good surface-active properties by reducing the tension at the oil−water interface and form interfacial films with different rheological properties. The most commonly used forms of milk proteins in food emulsions are sodium caseinate and whey protein (whey protein isolates or whey protein concentrates), and it is possible to make stable emulsions at a relatively low protein to oil ratio (about 1:60). In these emulsions, the surface protein coverage increases with an increase in protein concentration until it reaches a plateau value of about 2.0–3.0 mg/m2 [15,16]. Caseins (as in sodium caseinate or pure proteins) have rather flexible structures, i.e. they do not contain much rigid α-helix and β-pleated sheet

structure [17]. As a result, they adsorb rapidly at the interface, forming extended adsorbed layers up to about 10 nm thick [18,19]. In contrast, globular whey proteins (such as β-lactoglobulin) unfold partially, somewhere intermediate between the native state and the fully denatured conformation, resulting in compact adsorbed layers that are only about 2 nm thick [20–25]. Moreover, this partial unfolding of the whey protein structure following adsorption exposes reactive sulphydryl groups, thus resulting in slow polymerisation of the adsorbed protein in aged interfacial layers via sulphydryl−disulphide interchange reactions [26,27]. The role of caseins and whey proteins in stabilising oil-inwater emulsions has been thoroughly investigated and published [see reviews 23,28–33]. Bovine milk contains low levels of lactoferrin, an iron-binding glycoprotein with about 700 amino acid residues and a molecular weight of about 80,000 Da [34]. The surface of the lactoferrin molecule has several regions with high concentrations of positive charge, giving it a high isoelectric point (pI ≈ 9). This positive charge is one of the features that distinguishes lactoferrin from other milk proteins that have isoelectric points in the range 4.5–5.5. Lactoferrin adsorbs on to the interface of oil-in-water emulsion droplets and forms positivelycharged emulsion droplets that are stable over a wide pH range (from 7.0 to 3.0) [35]. One of the interesting features of lactoferrin-stabilised emulsions is that the positively droplets can interact via electrostatic interactions with other milk proteins that are mostly negatively charged around neutral pH. For instance, in aqueous solutions, lactoferrin tends to form a complex with β-lactoglobulin via electrostatic interactions, and these complexes have the ability to adsorb to the droplet surface, forming thick interfacial layers [35]. Multilayered emulsions can be produced by interactions of oppositely charged milk proteins, i.e. lactoferrin and β-lactoglobulin or caseinate at neutral pH [36]. Aggregated forms of milk proteins (i.e. calcium caseinate, casein micelles in milk and milk protein concentrates, and micellar casein) also adsorb on to the emulsion droplet surface, although not as effectively as the molecular forms of caseins or whey proteins. Generally, much higher concentrations of protein are required to make stable emulsions and larger droplets are formed under similar homogenisation conditions. These aggregated materials generate thicker adsorbed layers, resulting in higher surface coverage; values in the range 5–20 mg/m2, depending on the protein concentration, have been reported [15]. This is related to limited spreading of protein at the interface, because the aggregates are held together by calcium bonds and/or colloidal calcium phosphate and these bonds are unlikely to be affected during the emulsification process. The higher conformational stability of these aggregates also contributes to their reduced emulsifying ability [15,37]. By forming charged and hydrodynamically dense adsorbed layers, these aggregated caseins contribute to the long term stability of emulsions against coalescence by both electrostatic mechanisms and steric stabilisation mechanisms [38]. As well as milk proteins, plant proteins are used as an alternative source of food emulsifiers. Among the available plant protein sources, fractions extracted from cereals (α-gliadin from wheat) [39,40] and legumes (pea albumin, globulins from pea and soy) [41–43] are preferred for food applications as they are soluble in water without the need for cosolvents (such as alcohol) [44]. Pea proteins have been shown to reduce the interfacial tension between the water and oil surfaces by forming a rigid membrane at the oil−water interface [44]. This is due to the surface properties of their constitutive protein fractions, generally classified according to their sedimentation coefficient (S): vicilin (a trimeric 7S globulin) and legumin (a hexameric 11S globulin) [41,45,46]. Soy protein is widely used as an ingredient in food products because of its foaming and emulsifying properties [42,47]. It aids in emulsion formation by reducing the interfacial tension between the water and oil, and also by helping to stabilise the emulsion by forming a physical barrier at the oil−water interface. Similar to the storage proteins in pea, two major multi-subunit protein fractions, accounting for 70% of total soy protein, are storage

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57

globulins: 7S (β-conglycinin) and 11S (glycinin), which are expected to dominate the interfacial behaviour of soy protein isolate. In an emulsion system containing both protein and polysaccharide, protein generally forms the primary interfacial layer by directly adsorbing to the oil surface. The hydrophilic polysaccharide possibly forms a thick secondary steric-stabilising layer on the outside of protein-adsorbed emulsion droplets provided the protein−polysaccharide interaction is satisfactorily attractive [48]. Generally, strong electrostatic interaction between the mutually oppositely charged adsorbed protein and the added polysaccharide leads to the formation of multi-layered interfacial membranes stabilising emulsion droplets [49–55]. Covalent conjugates formed via Maillard reactions between proteins and polysaccharides have also attracted a lot of interest because of their higher emulsification abilities and better stabilities over wide ranges of temperature, pH and ionic strength, compared with the biopolymers alone [56–65]. A protein-based emulsion may become unstable as a result of various types of physical and chemical processes. Physical instability refers to modifications to the spatial arrangement or size distribution of emulsion droplets, such as creaming, flocculation and coalescence, whereas chemical instability includes changes in the composition of the emulsion droplets themselves, such as oxidation and hydrolysis [11,66]. As long as sufficient protein is present during homogenisation to cover the oil droplets, emulsions stabilised by proteins are generally very stable to coalescence over prolonged storage. However, at low protein to oil ratios, there is insufficient protein to fully cover the oil–water interface during homogenisation and, as a result, protein molecules/particles are shared by two or more droplets, causing bridging flocculation. Bridging flocculation is commonly observed in emulsions formed with aggregated protein products, such as calcium caseinate or micellar casein, in which the droplets are bridged by casein aggregates or micelles. Optimum stability can generally be attained at protein concentrations high enough to allow full saturation coverage at the oil–water interface. However, at very high protein to oil ratios, the presence of excess, unadsorbed protein may lead to depletion flocculation in some emulsions. In emulsions (35 or 45 vol% n-tetradecane) formed with sodium caseinate, it was shown that, at a protein content of up to 2.0 wt.%, the emulsion droplets were protected from flocculation by a stericstabilising layer of casein molecules [67]. However, when the protein content was increased to above 3.0 wt.%, the presence of unadsorbed caseinate in the continuous phase gave rise to depletion flocculation, resulting in serum separation. This was considered to be due to the presence of small casein aggregates in the aqueous phase, above a certain critical concentration. In contrast, emulsions formed with whey proteins do not show depletion flocculation, probably because the molecular size of whey proteins is less than the optimum to induce depletion flocculation [15]. This is due to the fact that, for depletion flocculation to occur, the biopolymer has to have a fairly high molecular weight so that the radius of gyration (Rg) is relatively large. It is possible to design interfacial layers of different structures, thicknesses, compositions and charges using specific protein emulsifiers individually or by complex formation to meet the structural demands, environmental challenges and stabilities of food emulsions. 3. Interactions of protein-stabilised emulsions under physiological conditions Although the effects of processing conditions (heat treatments, high pressure treatment, pH, and ionic strength) on the stability of emulsions (prior to consumption) have been extensively investigated during the past two decades, there is very little understanding of the behaviour of emulsions during and after consumption [2]. It is critical to understand the oral processing of emulsions to successfully manipulate the physical and sensorial attributes of colloidal food systems, such as creaminess, smoothness and rate of flavour release.

49

In recent years, a number of studies on the in vitro digestion of emulsions have been published, based on the hypothesis that emulsion design is potentially a key regulator in lipid digestion. These studies have focused mainly on the lipolysis of emulsified lipid by pancreatic lipase preparations, although in many cases proteins have been used to stabilise the emulsions [53,68]. It would be expected that, for such emulsions, the proteolysis of interfacial protein by pepsin (under gastric conditions) and by trypsin/ chymotrypsin would play a critical role in emulsion stability, which would in turn affect lipid digestibility. 3.1. Oral conditions Upon ingestion, emulsions are exposed to a range of conditions, such as dilution effects due to mixing with saliva, access to salivary enzymes such as amylases, the presence of various biopolymers such as mucins, different electrolytes in the saliva together with a moderate change in pH, temperature (around 37 °C) and friction between the tongue and the oral mucosa [6,7,69–71]. Most of the published work dealing with the oral processing of emulsions has been largely directed towards understanding the sensory properties [72–75] and flavour release [76] in food emulsions, using in-mouth models [5,77–79] and in vivo experiments using an electronic nose [80]. However, limited information on physicochemical changes and interactions of emulsions with different components of saliva is available. Interestingly, there is some evidence to show that the behaviour of protein emulsions in the mouth is largely driven by the interactions of saliva with the adsorbed layer on emulsion droplets [81,82]. Emulsions formed with whey protein isolate, sodium caseinate or lysozyme showed flocculation of droplets when mixed with unstimulated human saliva. This flocculation was stated to be of predominantly non-covalent origin and was claimed to be driven by the highly glycosylated negatively charged mucin present in human saliva [69,83–85]. The emulsion flocculation in the presence of saliva was considered to be regulated by depletion, van der Waals' forces and/or electrostatic interactions between emulsion droplets and salivary proteins, and was largely dependent on the initial charge of the emulsion droplets [69,83–85]. Recent work from our laboratory investigated the behaviour of oilin-water emulsions stabilised with lactoferrin or β-lactoglobulin [83]. The interfacial layers formed by these proteins allowed the formation of cationic and anionic droplets at neutral pH. Negatively charged protein-stabilised emulsions, i.e. β-lactoglobulin emulsions, did not interact with the artificial saliva because of strong repulsive forces between anionic mucin and the anionic β-lactoglobulin interfacial layer at neutral pH, but underwent depletion flocculation on the addition of higher concentrations of mucin (≥1.0 wt.%). In contrast, positively charged lactoferrin-stabilised emulsions showed saltinduced aggregation on the addition of saliva containing only salts (even without the addition of mucin). Additionally, lactoferrinstabilised emulsion droplets interacted with mucin via electrostatic interactions. When there was insufficient mucin to form a complete secondary layer around the lactoferrin-stabilised droplets, some bridging type flocculation occurred. A mucin concentration that provides a complete secondary coverage around the droplets of approximately 1 mg/m2 gives rise to a stable emulsion. Excessive mucin concentration in the continuous phase gives rise to depletion type flocculation as well more complex aggregations involving the self-association of mucin molecules. These kinds of emulsion–saliva electrostatic interactions might occur upon consumption of emulsions in real situations and could result in different sensorial and textural perceptions in vivo. Recent studies by Vingerhoeds et al. [82] showed that positively charged lysozyme-stabilised emulsions, which underwent irreversible flocculation with saliva, were perceived to be dry and astringent in the

50

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57

mouth. This astringent perception was related to the loss of the lubricating effect of saliva, probably because of the precipitation of salivary proteins by lysozyme molecules. In addition to flocculation by the mechanisms discussed above, coalescence of emulsion droplets, which has been attributed to shear-, surface-, saliva- or air-induced interactions in the oral cavity, has also been reported [86,87]. Although mucin is considered to be mainly responsible for inducing flocculation, some irreversible aggregation has been seen in emulsions even when mixed with parotid saliva [69], in which the mucin concentration is almost negligible. This indicates that other salivary components may also possibly drive emulsion flocculation by an unknown mechanism. Understanding emulsion–oral interactions in the presence of other salivary peptides, enzymes (amylases) and proteins would provide valuable information. For example, the use of prolinerich proteins, lysozyme and other positive charged fractions present in saliva may result in bridging flocculation of even negatively charged emulsion droplets at the neutral pH of the mouth environment. 3.2. Gastric conditions After residing for a relatively short period (from a few seconds to a few minutes) in the mouth, emulsions are subjected to a highly acidic pH (typically between 1 and 3) in the human stomach and mechanical agitation due to peristaltic movements of the stomach. Moreover, the emulsion is mixed with digestive juices, containing proteolytic (pepsin) and lipolytic (gastric lipase) enzymes, mucins and salts. Protein-stabilised emulsions would be expected to undergo major changes in the stomach because of the possible action of pepsin on the interfacial layers, the effects of low pH and ionic strength on the droplet charge and interactions of mucin with interfacial protein. There is limited understanding of the effects of gastric environments on emulsions stabilised by different kinds of food proteins. Work undertaken in our laboratory has shown that hydrolysis of the adsorbed protein layer by pepsin is the most important factor in determining emulsion stability under gastric conditions [88]. Both lactoferrin- and β-lactoglobulin-stabilised emulsions underwent flocculation followed by some degree of coalescence on exposure to simulated gastric conditions. This was caused by hydrolysis of the adsorbed layers by pepsin, resulting in a loss of positive charge on the droplet surface as well as a reduction in the thickness of the adsorbed layer. The peptides that remain at the interface are unable to provide sufficient electrostatic repulsions and/or steric effects. As a result, these emulsions are highly susceptible to flocculation and coalescence. Furthermore, the destabilisation initiated by pepsin is accelerated in the presence of high concentrations of salt [89]. Although β-lactoglobulin is usually resistant to pepsin attack in its native state in aqueous solution, it became susceptible to proteolysis when present as the interfacial layer in an emulsion [88]. This could be explained on the basis of a change in the conformation of the β-lactoglobulin molecules upon adsorption at the oil–water interface, which exposed the peptic cleavage sites for proteolysis. However, some portions of β-lactoglobulin appeared to remain inaccessible to pepsin during the early stages of digestion. Similar behaviour of β-lactoglobulin was also observed by Macierzanka et al. [90], who suggested that the protein adopted a range of different states of accessibility/folding because the population of adsorbed β-lactoglobulin that was susceptible to pepsinolysis was digested at a range of different rates. They also compared the gastric digestibility of β-lactoglobulin with that of a more flexible protein, i.e. β-casein, in an olive oil emulsion. The rate of gastric digestion of the adsorbed β-casein in the emulsion was twice as fast as that in solution and caused a substantial destabilisation of the emulsion. The presence of a low level of mucin appears to promote the flocculation of β-lactoglobulin-stabilised emulsions, possibly through a bridging mechanism [89]. Further work is required to better understand the interactions of gastric mucins with protein emulsions

at a range of different concentrations and how these interactions influence the pepsin accessibility of the adsorbed protein layers. 3.3. Intestinal conditions After passing through the stomach, food emulsions enter the small intestine and this stimulates contraction of the gall bladder, resulting in the delivery of digestive enzymes and bile into the duodenum. The presence of various lipolytic as well as proteolytic enzymes at different concentrations, inorganic salts and surface-active bile acids and the neutral–alkaline pH (6.0–7.5) of the upper intestinal fluid [2,9,91] together with the remnants of oral and gastric digestion make the overall intestinal processing highly complex and difficult to unravel. Bile salts are generally known to displace protein-stabilised interfaces by virtue of their higher surface activities. Mun et al. [68] studied the interface of emulsion droplets during storage. The dependence of bile-salt-induced displacement mechanisms on the nature of the adsorbed layer was further supported by another study, in which the digestibility of adsorbed milk proteins under simulated gastrointestinal conditions was investigated [90]. The in vitro digestibilities of β-lactoglobulin- and β-casein-adsorbed surfaces were compared, with and without the addition of surfactants (bile salts and phosphatidyl choline). The authors reported that the final structure of the β-lactoglobulin-stabilised interface was largely driven by competitive displacement by bile salts and/or phosphatidyl choline, in contrast to adsorbed β-casein, where gastric hydrolysis by pepsin was the major cause of destabilisation of the β-casein-stabilised interface. In another study, the competitive displacement of βlactoglobulin from air–water and oil–water interfaces by bile salts was investigated in an in vitro duodenal model [91]. These authors suggested that bile salts might completely displace the β-lactoglobulin from the interface, when passing through the duodenum in vivo, which in turn would affect the rate of lipid digestion of the β-lactoglobulinstabilised emulsified droplets. The interactions of bile salts with lactoferrin- or β-lactoglobulinstabilised emulsion droplets have been reported recently by Sarkar et al. [92]. On the addition of bile salts, release of peptides was observed, which could have arisen only as a result of proteolysis due to the presence of contaminating enzymes in the bile salts. Therefore, commercial bile salts must be checked for proteolytic activity and any residual enzymes could be eliminated by adequate heat treatments. In emulsions stabilised by β-lactoglobulin, displacement of protein was observed even at the lowest concentration of bile salts, but bile salts appeared to bind to the positively charged lactoferrin emulsion droplet with no protein displacement (Fig. 1). Hence, the initial charge of the protein adsorbed layer influences the bile salt displacement reactions. A number of studies have focused on understanding how the interface of an emulsion affects the rate of lipid digestion, using pancreatic lipases in in vitro intestinal models [53,68,93,94]. The extents of lipid hydrolysis, measured as the amount of fatty acid released during 2 h of hydrolysis, for caseinate- and whey-protein-stabilised emulsions were found to be similar, although the oil droplets in the whey protein emulsions were more unstable [68]. Hydrolysis of lipids is greatly enhanced in most systems in the presence of bile salts. As lipolysis progresses, surface-active components, such as monoglycerides and diglycerides are produced and these compounds competitively displace many surface-active molecules (bile salts, peptides etc.) from the droplet surface. Protonated free fatty acids released during gastric lipolysis limit the extent of lipid digestion at high concentrations. We showed that lactoferrin- and β-lactoglobulin-stabilised emulsions underwent a significant degree of coalescence on the addition of physiological concentrations of pancreatin and bile salts [95]. For both emulsions, this emulsion destabilisation in simulated intestinal fluid was largely attributed to the lipolysis of the lipid hydrophobic core by the lipase fractions of the pancreatin as well as the proteolysis of the

Total protein content of continuous phase (%)

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57

0.5

0.4

0.3

0.2

0.1

0.0 0.0

5.0

10.0

15.0

20.0

25.0

30.0

Bile salts concentration (mg/ml) Fig. 1. Changes in protein content of the continuous phase of lactoferrin (○) and β-lactoglobulin (●) emulsions as a function of bile salts concentration in preheated simulated intestinal fluid based on Kjeldahl method (reproduced with the permission of Elsevier Inc. [89]).

adsorbed protein layer by the trypsin or other proteolytic fractions present in pancreatin. It is important to consider the purity of commercially available lipases when trying to unravel the mechanisms of lipase action on protein-based emulsions. In our experience, even the purest commercially available lipase shows some proteolytic activity. Moreover, human gastric fluid also contains acid-stable gastric lipase that accounts for 10–30% of the overall lipid digestion of ingested triacylglycerols [96–99]. These acid-stable lipases could act on emulsions, resulting in the generation of surface-active fatty acids and monoglycerides from the hydrophobic lipid core. Some of these compounds could displace the initial emulsifier materials from the emulsion interface [97,99]. It is possible that the gastric lipase digestion products would displace intact or hydrolysed protein from the interface of protein-stabilised emulsions. This could lead to a change in droplet size and could affect the overall stability of the emulsion. However, there are no reported studies on how protein-stabilised emulsions are affected by the action of gastric lipase. This is largely due to the unavailability of commercial gastric lipases. Systematic studies on the behaviour of protein-stabilised emulsions on the addition of gastric lipase under simulated gastric conditions need to be carried out. 4. Sequential processing of protein-based emulsions in an oral-to-gastrointestinal model Most of the published studies on protein emulsions have used in vitro digestion systems, basically simulating the conditions in the mouth, the stomach or the intestine. Broadly, attempts to simulate the temperature, pH, ionic strength, enzymes, co-enzymes, bile salts, biopolymers etc. of the physiological fluids have been made. The composition, microstructure, lipid droplet size distribution, droplet charge, rheology, interfacial composition, binding interactions with the physiological biopolymers and competitive interfacial exchange reactions are determined by treating the emulsions with these simulated physiological fluids under appropriate mixing conditions. Most of the studies have attempted to unravel the interaction of protein-stabilised emulsions at a particular site (mouth, stomach or duodenum) to gain fundamental insights into the influence of individual physiological variables. However, significant physicochemical changes that might have occurred during pre-processing in the mouth, followed by sequential gastric digestion, have not been explored when studying lipid digestion in the intestine. Importantly,

51

in the absence of the proteolytic enzymes normally found in the gastric phase, it is difficult to know what the persistence of the initial protein layers might be. These issues were largely addressed by a recent paper involving an in vitro digestion model [100] that included oral, gastric and duodenal phases in real sequence. The authors showed that the initial emulsifier used to stabilise oil-in-water emulsions had only a limited effect on the microstructural changes that occurred during the digestion of lipid droplets. Recently, a complete in vitro digestion model for toxicological studies has been proposed by some researchers to simulate various physiological processes occurring in the mouth, stomach and small intestine at a temperature of 37 °C [101,102]. The emulsion–biological fluid mixtures also need to be carefully agitated depending on the shear and the motility in the mouth and the human stomach respectively. Such an in vitro digestion model could be applicable as a protocol for studying the digestion of emulsified lipids under fed state conditions. This kind of complete in vitro model could be used to elucidate the interactions of protein-stabilised emulsions during the entire physiological regime. We have recently explored the interactions of cationic (lactoferrin) and anionic (β-lactoglobulin) oil-in-water emulsions using an in vitro oral-to-gastrointestinal model, with the fluids added in order of their sequence during real digestion. The in vitro physiological model used in this study was a modified method in terms of composition and constituents of the artificial media [100,102]. Basically, it consisted of a conical flask (250 mL) containing the simulated fluids (mimicking conditions in the mouth, stomach and small intestine) maintained at 37 °C with continuous shaking at 95 rev/min in a temperaturecontrolled water bath (Lab-Line shaker bath, Model LZ33070, Barnstead International, Dubuque, IA, USA). The artificial saliva (AS), simulated gastric fluid (SGF) and simulated intestinal fluid (SIF) were prepared according to the compositions shown in Table 1. The pH and the temperature were continuously monitored. The in vitro physiological model was used as described below. a) Before ingestion: Emulsion samples (20 wt.% soy oil stabilised by 1 wt.% lactoferrin or β-lactoglobulin) at neutral pH before any treatment. b) Artificial oral conditions: 10 mL of emulsion sample was mixed for 5 min with 10 mL of AS containing 0.02 wt.% porcine mucin [80]. c) Simulated gastric conditions: Nearly 20 mL of SGF containing 0.32 wt.% pepsin was added (mixture pH ~ 1.5) and the mixture was agitated for 2 h [85]. d) Simulated small intestinal conditions: About 5 mL of 0.5 M HCO− 3 solution was added to change the pH of the gastric-digested emulsion samples to ~ 6.5. Nearly 20 mL of SIF at pH 7.5 with slight modification (39 mM K2HPO4, 150 mM NaCl and 30 mM CaCl2), containing 0.5 wt.% bile salts and 1.0 wt.% pancreatin, was subsequently added (mixture pH ~ 7.2) and the mixture was agitated for 2 h [92,95]. The soy oil concentration was 20 wt.% initially, and was diluted to 10.0 wt.% in the oral step, to 5.0 wt.% in the gastric step and to 1.8 wt.% in the duodenal step. Initially, both emulsions had a neutral pH (~7.0) and the pH values largely remained the same in the oral environment. The pH significantly declined to ~1.5 when the emulsions entered the gastric system, whereas it increased again to ~ 7.2 under the simulated duodenal conditions. No significant differences in the pHs of the two emulsions were found, indicating similar buffering capacities of β-lactoglobulin and lactoferrin. 4.1. Droplet size and microstructure The droplet size distributions of the emulsions, determined using static light scattering (MasterSizer 2000 Hydro), showed that both the β-lactoglobulin- and lactoferrin-stabilised emulsions were uniformly dispersed, with all the droplets being under 5 μm in size (Fig. 2A and B).

52

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57

Table 1 Chemical compositions of various simulated fluids used in the in vitro physiological model.

Components

Antimicrobial agent pH Ionic strength (Calculated values)

Artificial saliva (AS)

Simulated gastric fluid (SGF)

Bicarbonate solution (HCO− 3 )

Simulated intestinal fluid (SIF)

0.1594 wt.% NaCl 0.0328 wt.% NH4NO3 0.0636 wt.% KH2PO4 0.0202 wt.% KCl 0.0308 wt.% C6H5K3O7 0.002 wt.% Uric acid-sodium salt 0.0146 wt.% Lactic acid-sodium salt 0.02 wt.% NaN3 ~ 6.8 ~ 0.18 M

0.32 wt.% Pepsin 0.7 mL Conc. HCl 0.2 wt.% NaCl

0.5 M NaHCO3

0.68 wt.% K2HPO4 150 mM NaCl 30 mM CaCl2 1.0 wt.% Pancreatin 0.5 wt.% Bile salts 19 mL 0.2 N NaOH

0.02 wt.% NaN3 ~ 1.2 ~ 0.034 M

0.02 wt.% NaN3 ~ 6.5 ~ 0.5 M

0.02 wt.% NaN3 ~ 7.5 ~ 0.35 M

Confocal laser scanning microscopy showed that, before any treatment, the fresh emulsions made with β-lactoglobulin and lactoferrin (Fig. 3A and 4A) had fine and rather monodisperse droplets, confirming the light scattering results (Fig. 2). On mixing with AS, the β-lactoglobulin-stabilised emulsions showed no change in droplet size distribution (Fig. 2) and microstructure (Fig. 3B). However, the lactoferrin-stabilised emulsions showed a significant population of larger droplets (5–10 μm) and a bimodal size distribution under the same conditions (Fig. 2B). Pronounced droplet aggregation was also observed in the lactoferrin-stabilised emulsions in the presence of AS by confocal laser scanning microscopy (Fig. 4B). When the lactoferrin emulsion–saliva mixtures were mixed gently with 2.0% SDS to break up the aggregates,

A 10 (A)

Volume (%)

8

6

4

2

0 0.01

0.1

1

10

100

1000

100

1000

Droplet size (µm)

B 10 (B)

Volume (%)

8

6

4

2

0 0.01

0.1

1

10

Droplet size (µm) Fig. 2. Droplet size distribution of β-lactoglobulin (A) and lactoferrin (B) emulsions as they pass through the in vitro physiological model: before digestion (○), after oral treatment for 5 min (●), after gastric digestion for 2 h (▲) and after duodenal digestion for 2 h (■). Each distribution is the average of measurements on duplicate samples.

the bimodal distributions reverted to nearly monomodal (data not shown), similar to that of the untreated emulsion, indicating that the larger size droplets were due to flocculation rather than coalescence. This flocculation of lactoferrin emulsion droplets in a simulated oral environment can be attributed to a combination of salivary-saltinduced aggregation and mucin-mediated bridging of the positively charged emulsion droplets, as discussed in our previous studies [83]. On mixing the emulsion–saliva mixtures with SGF for 2 h, there was a significant increase in droplet diameter for both emulsions, and bimodal distributions, with a second peak in the region 5–10 μm and a corresponding decrease in the area of the first peak, were observed (Fig. 2A and B). When these samples (showing bimodal peaks) were dispersed in 2.0% SDS buffer, the population of larger droplets was minimised, although the size distribution remained bimodal (data not shown). This indicated that, in addition to droplet flocculation, some coalescence of the emulsion droplets in these systems had occurred, which was clearly shown by the appearance of some distinct large droplets (N10 μm) in the confocal micrographs of both emulsions (Figs. 3C and 4C). This significant change in the droplet characteristics and microstructures of both protein-stabilised emulsions in the simulated gastric system can be attributed to the pH change and pepsin-induced hydrolysis of the interfacial protein (β-lactoglobulin or lactoferrin) layer [88,89]. On passing the emulsion-digested samples through the last step of the in vitro model, i.e. mixing with SIF in the artificial duodenal system, the size distribution curve showed a third peak in the size range 10–150 μm (Fig. 2A and B). When these samples were dispersed in SDS, the size distribution remained multimodal with a slight reduction in the population of droplets of size ~ 10–100 μm, which suggested the presence of some coalesced emulsion droplets. Confocal microscopy showed that the extent of coalescence appeared to be slightly less in the presence of SIF than in the gastric step (Figs. 3C, D, 4C and D). Furthermore, the system also showed some finely dispersed droplets of 0.5–2.0 μm in size in both emulsions after treatment with SIF; these droplets seemed to be more spherical and very different in appearance from those in the initial emulsion (Figs. 3D and 4D). Coalescence for both kinds of emulsion droplets in the duodenal step together with the appearance of these smaller droplets (b5 μm) could be attributed to the displacement of and/or binding to the interfacial proteins by bile salts together with the interfacial proteolysis by the trypsin fractions of the pancreatin [92,95]. This displacement might have further promoted the accessibility of lipase to act on the hydrophobic lipid core, generating lipid digestion products (mono- and/or diglycerides, fatty acids etc.) at the droplet surface, thus showing some spherical droplets largely stabilised by the lipid digestion products. 4.2. ζ-Potential The dependence of the ζ-potential of emulsion droplets on the conditions in the in vitro physiological model is shown in Fig. 5. Freshly prepared β-lactoglobulin- and lactoferrin-stabilised

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57

53

Fig. 3. Confocal micrographs of β-lactoglobulin emulsions as they pass through the in vitro physiological model: before digestion (A), after oral treatment for 5 min (B), after gastric digestion for 2 h (C) and after duodenal digestion for 2 h (D). Scale bar corresponds to 10 μm. Emulsion–physiological fluid mixtures was stained with 1.0% (w/v) Nile Blue (fluorescent dye), covered with a cover slip and finally observed with a 100× magnification lens in a confocal microscope (Leica DM6000 B, Heidelberg, Germany) using an Ar/Kr laser with an excitation wavelength of 488 nm.

emulsions at pH 7.0 had ζ-potentials of ~−55 and ~+52 mV respectively. Addition of AS drastically reduced the ζ-potential (~+7 mV) of the lactoferrin-stabilised emulsion due mainly to the binding of negatively-charged mucin to the droplet surface, as described in previous sections. There were no significant changes in the ζ-potential of β-lactoglobulin-stabilised emulsion droplets in the oral environment (p N 0.05), as there were no interactions between anionic β-lactoglobulin-stabilised emulsion droplets and mucin in the saliva. Addition of SGF to the emulsions resulted in charge reversal for the β-lactoglobulin-stabilised emulsion and significantly affected the ζ-potential values of both emulsions from the oral environment. In the gastric system, both emulsion droplets had a significant amount of positive charge after 2 h of incubation in SGF although peptic hydrolysis had taken place. This indicates that the presence of some amphiphilic interfacial peptides provided some protection to the droplets against complete coalescence. The most prominent change in ζ-potential occurred when the emulsion droplets moved to the in vitro duodenal model, resulting in a highly negative droplet charge for both emulsions. The ζ-potential values in the presence of SIF were fairly similar for both types of emulsion droplets (p N 0.05), irrespective of the initial droplet charge, because the initial interfacial layers were possibly digested by the proteolytic fractions of the pancreatin and were displaced by bile salts and/or lipid digestion products. The impact of the initial protein type on the extent of lipolysis was determined by measuring the release of free fatty acids as a function of incubation time in the SIF after the gastric digestion step (Fig. 6).

The amount of free fatty acids released per unit volume of emulsion was similar for both emulsions (p N 0.05). The results from the sequential processing do not appear to differ significantly from what was generally found in the non-sequential studies, at least qualitatively, suggesting that not only the charge but the nature of the protein are not particularly important in the final digestion of lipids. Also, the behaviour of lactoferrin and β-lg emulsions with regard to physical stability appears to be similar in this sequential study, although some differences have been observed in non-sequential studies. 4.3. Overall mechanism Fig. 7 illustrates the possible changes in protein-stabilised emulsions after they are taken inside the mouth and subsequently as they traverse through the gastrointestinal tract using a simulated physiological model [2]. On ingestion, a protein-stabilised emulsion is mixed with saliva, undergoes a change in pH, ionic strength and temperature, becomes exposed to enzymes (amylases) and biopolymers (mucin and other proteins) and experiences a complex salivary flow and friction with the oral surfaces. The interactions of the emulsion with the saliva largely depend on the initial charge of the emulsion, because saliva largely contains negatively charged electrolytes and anionic mucin. Broadly, neutral and negatively charged emulsions undergo reversible depletion flocculation whereas cationic emulsions show irreversible associative electrostatic interactions with mucin and salivary salts

54

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57

Fig. 4. Confocal micrographs of lactoferrin emulsions as they pass through the in vitro physiological model: before digestion (A), after oral treatment for 5 min (B), after gastric digestion for 2 h (C) and after duodenal digestion for 2 h (D). Scale bar corresponds to 10 μm. Emulsion–physiological fluid mixtures was stained with 1.0% (w/v) Nile Blue (fluorescent dye), covered with a cover slip and finally observed with a 100× magnification lens in a confocal microscope (Leica DM6000 B, Heidelberg, Germany) using an Ar/Kr laser with an excitation wavelength of 488 nm.

when mixed with saliva. However, little is known about the influence of salivary component (e.g. proline-rich proteins) other than mucin on emulsion droplet behaviour during consumption. In the stomach, the emulsion–saliva mixtures undergo a drastic shift in pH because of the highly acidic gastric environment (pH 1–3)

and are also exposed to a proteolytic enzyme (pepsin). As the pH is well below the isoelectric point of most food proteins, most of the emulsions acquire a net positive charge under gastric conditions. Hence, rather than the initial charge of the emulsifier, it is the nature of the protein coated on the droplet surface and its susceptibility to 350

Free fatty acid (µmol/mL emulsion)

60

-potential (mV)

40 20 0 -20 -40 -60

Initial

Oral

Gastric

300 250 200 150 100 50

Duodenal

Simulated physiological conditions Fig. 5. ζ-Potentials of lactoferrin (white) and β-lactoglobulin (black) emulsions as they pass through the in vitro physiological model. Dilution of emulsion-simulated physiological fluid mixture sample was carried out to approximately 0.005 wt.% droplet concentrations using the corresponding inorganic buffer solutions (10 mM phosphate buffer at pH 7 was used for the initial, oral and small intestinal steps, while 10 mM citrate buffer at pH 1.5 was used for the gastric step) [97].

0 0.0

20.0

40.0

60.0

80.0

100.0

120.0

Time of digestion (min) Fig. 6. Amount of free fatty acid (μmol) released from each millilitre of lactoferrin (white) and β-lactoglobulin (black) emulsions on the addition of SIF containing 1.0 wt.% of pancreatin and 0.5 wt.% of added bile salts as a function of time.

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57

55

Fig. 7. Schematic illustration of the possible changes in emulsions as they pass through the in vitro physiological model (reproduced with the permission of Elsevier Inc. [2]).

pepsin that drive the emulsion destabilisation process. Although globular proteins, such as β-lactoglobulin, are highly resistant to pepsin digestion in their native state, the unfolding of the protein structure during emulsion formation renders the proteolytic sites susceptible to pepsin attack. As a result of proteolysis of the interfacial protein layer, the peptides generated are not strong enough to provide emulsion stability, thus resulting in coalescence. Human gastric fluid contains acid-stable gastric lipase that could act on emulsions, resulting in the generation of fatty acids and monoglycerides from the hydrophobic lipid core. As some of these digestion products are surface-active, they have the potential to displace protein/peptides from the interface of protein-stabilised emulsions. There are no reported studies on how protein-stabilised emulsions are affected by the action of gastric lipase, and whether or not this action influences emulsions stability. Human stomach also contains highly glycosylated mucin which forms a self associated gel-like structure at gastric pH. It is not known how protein/peptide-stabilised emulsion droplets interact with gastric mucin gelled layers or solubilised mucin. On entering the intestine, the gastric-digested emulsion droplets are mixed with lipases, proteases, surface-active bile salts and bicarbonates. The pH in the small intestine becomes close to neutral because of the mixing of the alkaline intestinal fluid with the gastricdigested emulsions. In protein/peptide-stabilised emulsions, the action of serine proteases (trypsin and chymotrypsin) on the adsorbed protein layer would be expected leads to emulsion instability. However, molecular details of sequential actions of pepsin and trypsin/chymotrypsin on adsorbed proteins in emulsions are unknown. The action of bile salts on the composition of the adsorbed layer is also very important, as these surface active compounds displace the adsorbed protein/peptides from the surface of emulsion droplets. This action appears to promote the accessibility of active site

of pancreatic lipase to the hydrophobic lipid core. Pancreatic lipase adsorbs to the droplet interface usually via complexation with colipase and/or bile salts. Pancreatic lipase cleaves triacylglycerols to form 2-monoacylglycerols and fatty acids; some of these digestion products are surface active and could potentially displace the initial adsorbed material from the droplet surface. The various lipid digestion products are then incorporated within bile/phospholipid micelles and vesicles that are transported to the enterocyte cells, where they are absorbed. Hence, the initial charge and composition of the protein layers plays an insignificant role, when considering the overall complexities of lipid digestion phenomena, possibly because they are partially or completely displaced by bile salts and lipid digestion products and/or are digested by the proteolytic enzymes. Further research is required on understanding the competitive displacement mechanisms occurring in the intestine, characterization of the final state of droplets and the products of lipid hydrolysis. 5. Concluding remarks Understanding of lipid digestion and its relationship to food structure is an emerging area of scientific research. As lipid digestion is an interfacial process dependent on the adsorption of lipases on the surface of emulsified droplets, it seems possible to alter lipid digestion by modification of emulsion structure and stability. Our knowledge of the specific interfacial aspects of lipid digestion is rather limited at this point in time. One of the crucial challenges is how to simulate the exact complexities of the physiological processes occurring in the human digestive system using in vitro models. The in vitro models used in most recent literature [90,100,103] are static; therefore, the gradual changes and physical processes that occur in vivo (e.g. gastric

56

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57

emptying, peristalsis, hydration etc.) are not generally mimicked [104]. To obtain an overall understanding of total digestion and gastric transit, dynamic models may need to be used. The TNO in vitro gastrointestinal model (TIM), developed at TNO Nutrition and Food Research (The Netherlands), and the dynamic gastric model (DGM), developed at the Institute of Food Research (United Kingdom) [104], represent computerised dynamic digestion processes. Clearly, in vitro studies offer several advantages over in vivo studies because they are usually less time consuming, less expensive and more reproducible (no biological variations), have no ethical constraints, are easy to perform and allow the collection of samples at any level of the gastrointestinal tract at any time of digestion. Both static and dynamic in vitro models help to provide an improved understanding of the fundamental physicochemical mechanisms by unravelling the interactions between food structures (emulsions) and individual or total physiological variants (including real time measurements as in the case of dynamic digestion models). However, in reality, the dilution with saliva and the salivary flow, the amount of gastric juice released and the quantities of surface-active agents and enzymes present in the gastrointestinal tract may vary markedly according to the type of food (chemical composition, constituents, quantity, physical state etc.), according to the individual's physiology (age, gender, appetite, blood group etc.) and within the same human subject at different times of the day. Thus, one of the major drawbacks of using in vitro models is that the composition and the quantity of the physiological fluids, surfactants, enzymes etc. involved is not always comparable with real physiological conditions. For example, it is frequently debated that the pepsin:substrate ratios in most of the in vitro studies vary from 1:3 to 1:10 w/w, which are far ahead of those likely to be found in the stomach, even under fed state conditions [104]. It should be noted here that it is extremely difficult to select an optimal protein:pepsin ratio in an in vitro digestion model that exactly mimics the secretion found physiologically in humans because a wide variation (up to about 10,000 fold) in gastric and pancreatic secretions, depending on the individual's health and the type of food intake, has been suggested [105]. Because food systems are more complex than model oil-in-water emulsions, the physiological interactions in multi-component systems are likely to be extremely complex. Determining the interactions of emulsified lipids containing mixed biopolymer systems (proteins and carbohydrates) in the presence of simulated physiological fluids would provide better understanding of the macromolecular interactions within the nutrients together with their influence on controlling lipid digestion. For example, the presence of polysaccharides might significantly saturate the protein-stabilised lipid droplets, thereby preventing the enzymes from coming in contact with the lipid hydrophobic core, and thus influencing the overall lipid digestion. Manipulation of food structures together with mapping of their physical, chemical and biological fates during physiological lipid digestion (using in vitro, in vivo and clinical trials) will help to rationally fabricate future foods with designed functional behaviours in the body, such as controlled release of lipid bioactive molecules. Such foods might have the potential benefits of fighting against obesity, cardiovascular diseases and other food-linked health issues and/or of providing improved satiety responses. References [1] Kritchhevsky D. In: Akoh CC, Min DB, editors. Food lipids: chemistry, nutrition, and biotechnology. New York: Marcel Dekker; 2002. p. 543–58. [2] Singh H, Ye A, Horne D. Prog Lipid Res 2009;48:92. [3] McClements DJ, Decker EA, Park Y. Crit Rev Food Sci Nutr 2009;49:48. [4] Le Révérend BJD, Norton IT, Cox PW, Spyropoulos F. Curr Opin Colloid Interface Sci 2010;15:84. [5] Doyen K, Carey M, Linforth RST, Marin M, Taylor AJ. J Agric Food Chem 2001;49:804. [6] Malone ME, Appelqvist IAM, Norton IT. Food Hydrocoll 2003;17:763. [7] Malone ME, Appelqvist IAM, Norton IT. Food Hydrocoll 2003;17:775. [8] Bayarri S, Taylor AJ, Hort J. J Agric Food Chem 2006;54:8862.

[9] McClements DJ, Decker E, Park Y, Weiss J. Food Biophys 2008;3:219. [10] Golding M, Wooster TJ. Curr Opin Colloid Interface Sci 2010;15:90. [11] McClements DJ. Food emulsions: principles, practices and techniques. 2nd ed. Florida: CRC Press; 2005. [12] Dickinson E. Food Hydrocoll 2003;17:25. [13] Morr CV. In: Fox PF, editor. Developments in dairy chemistry—I. London: Applied Science Publishers; 1982. p. 375. [14] Mulvihill DM, Fox PF. In: Fox PF, editor. Developments in dairy chemistry. London: Elsevier Applied Science; 1989. p. 131. [15] Euston SR, Hirst RL. Int Dairy J 1999;9:693. [16] Srinivasan M, Singh H, Munro PA. J Food Sci 2001;66:441. [17] Holt C, Sawyer L. Prot Eng 1988;2:251. [18] Dalgleish DG. Food Res Int 1996;29:541. [19] Dalgleish DG. Colloids Surf 1990;46:141. [20] Dickinson E. J Chem Soc Faraday Trans 1998;94:1657. [21] Fang Y, Dalgleish DG. Food Hydrocoll 1998;12:121. [22] Dickinson E, McClements DJ. Advances in food colloids. London: Blackie Academic & Professional; 1995. p. 27. [23] Dalgleish DG. In: Dickinson E, Lorient D, editors. Food macromolecules and colloids. Cambridge: Royal Society of Chemistry; 1995. p. 23. [24] Dalgleish DG, Leaver J. In: Dickinson E, Walstra P, editors. Food colloids and polymers: stability and mechanical properties. Cambridge: Royal Society of Chemistry; 1993. p. 113. [25] Mackie AR, Mingins J, Dann R. In: Dickinson E, Walstra P, editors. Food colloids and polymers: stability and mechanical properties. Cambridge: Royal Society of Chemistry; 1993. p. 96. [26] McClements DJ, Monahan FJ, Kinsella JE. J Food Sci 1993;58:1036. [27] Dickinson E, Matsumura Y. Int J Biol Macromol 1991;13:26. [28] Kinsella JE. Crit Rev Food Sci Nutr 1984;21:197. [29] Wong DWS, Camirand WM, Pavlath AE. Crit Rev Food Sci Nutr 1996;36:807. [30] Dickinson E. Colloids Surf B 2001;20:197. [31] Dickinson E. Int Dairy J 1999;9:305. [32] Dalgleish DG. Food Hydrocoll 2006;20:415. [33] Morr CV, Ha EYW. Crit Rev Food Sci Nutr 1993;33:431. [34] Baker EN, Baker HM. Cell Mol Life Sci 2005;62:253. [35] Ye A, Singh H. J Colloid Interface Sci 2006:295. [36] Ye A, Singh H. Food Biophys 2007;2:125. [37] Srinivasan M, Singh H, Munro PA. Int Dairy J 1999;9:337. [38] Dickinson E. Colloids Surf A 2006;288:3. [39] MacRitchie F, Lafiandra D. In: Damodaran S, Paraf A, editors. Food proteins and their applications. New York: Marcel Dekker; 1997. p. 293. [40] Örnebro J, Nylander T, Eliasson AC. J Cereal Sci 2000;31:195. [41] Gueguen J. J Agric Food Chem 1989;37:1236. [42] Kinsella J. J Am Oil Chem Soc 1979;56:242. [43] Utsumi S, Matsumura Y, Mori T. In: Damodaran S, Paraf A, editors. Food proteins and their applications. New York: Marcel Dekker; 1997. p. 257. [44] Ducel V, Richard J, Popineau Y, Boury F. Biomacromolecules 2004;5:2088. [45] Gueguen J, Chevalier M, And JB, Schaeffer F. J Sci Food Agric 1988;44:167. [46] Gharsallaoui A, Cases E, Chambin O, Saurel R. Food Biophys 2009;4:273. [47] Peng IC, Quass DW, Dayton WR, Allen CE. Cereal Chem 1984;61:480. [48] Dickinson E. J Food Eng 1994;22:59. [49] Dickinson E, James JD. Food Hydrocoll 2000;14:365. [50] Güzey D, Kim HJ, McClements DJ. Food Hydrocoll 2004;18:967. [51] Güzey D, McClements DJ. Food Biophys 2006;1:30. [52] Moreau L, Kim H-J, Decker EA, McClements DJ. J Agric Food Chem 2003;51:6612. [53] Mun S, Decker E, Park Y, Weiss J, McClements DJ. Food Biophys 2006;1:21. [54] Laplante S, Turgeon SL, Paquin P. Food Hydrocoll 2005;19:721. [55] Hong Y-H, McClements D. Food Biophys 2007;2:46. [56] Hattori M, Nagasawa K, Amenati A, Kaminogawa S, Takahashi K. J Agric Food Chem 1994;42:2120. [57] Hattori M, Yang W, Takahashi K. J Agric Food Chem 1995;43:2007. [58] Kato A, Sato T, Kobayashi K. Agric Biol Chem 1989;53:2147. [59] Nagasawa K, Takahashi K, Hattori M. Food Hydrocoll 1996;10:63. [60] Dickinson E, Euston SR. In: Dickinson E, editor. Food polymers, gels and colloids. Cambridge: Woodhead Publishing; 1991. p. 132. [61] Dickinson E, Flint FO, Hunt JA. Food Hydrocoll 1989;3:389. [62] Dickinson E, Galazka VB. Food Hydrocoll 1991;5:281. [63] Kato A, Mifuru R, Matsudomi N, Kobayashi K. Biosci Biotechnol Biochem 1992;56: 567. [64] Neirynck N, Van der Meeren P, Bayarri Gorbe S, Dierckx S, Dewettinck K. Food Hydrocoll 2004;18:949. [65] Al-Hakkak J, Kavale S. Int Congr Ser 2002;1245:491. [66] McClements DJ, Decker EA. J Food Sci 2000;65:1270. [67] Dickinson E, Golding M, Povey MJW. J Colloid Interface Sci 1997;185:515. [68] Mun S, Decker EA, McClements DJ. Food Res Int 2007;40:770. [69] Vingerhoeds MH, Blijdenstein TBJ, Zoet FD, van Aken GA. Food Hydrocoll 2005;19:915. [70] de Wijk RA, Prinz JF. Food Qual Prefer 2005;16:121. [71] de Wijk RA, Prinz JF, Engelen L, Weenen H. Physiol Behav 2004;83:81. [72] de Wijk RA, Engelen L, Prinz JF. Appetite 2003;40:1. [73] Engelen L, de Wijk RA, Prinz JF, van der Bilt A, Bosman F. Physiol Behav 2003;78: 165. [74] Kilcast D, Clegg S. Food Qual Prefer 2002;13:609. [75] Metcalf KL, Vickers ZM. J Sens Stud 2002;17:379. [76] de Roos KB. Int Dairy J 2003;13:593. [77] van Ruth SM, King C, Giannouli P. J Agric Food Chem 2002;50:2365.

H. Singh, A. Sarkar / Advances in Colloid and Interface Science 165 (2011) 47–57 [78] van Ruth SM, Roozen JP. Eur Food Res Technol 2000;210:258. [79] van Ruth SM, Roozen JP. Food Chem 2000;71:339. [80] Miettinen S-M, Tuorila H, Piironen V, Vehkalahti K, Hyvonen L. J Agric Food Chem 2002;50:4232. [81] van Aken GA, Vingerhoeds MH, De Hoog EHA. In: Dickinson E, editor. Food Colloids: interactions, microstructure and processing. Cambridge: Royal Society of Chemistry; 2005. p. 356. [82] Vingerhoeds MH, Silletti E, de Groot J, Schipper RG, van Aken GA. Food Hydrocoll 2009;23:773. [83] Sarkar A, Goh KKT, Singh H. Food Hydrocoll 2009;23:1270. [84] Silletti E, Vingerhoeds MH, Norde W, van Aken GA. J Colloid Interface Sci 2007;313:485. [85] Silletti E, Vingerhoeds MH, Norde W, van Aken GA. Food Hydrocoll 2007;21:596. [86] Dresselhuis DM, de Hoog EHA, Cohen Stuart MA, Vingerhoeds MH, van Aken GA. Food Hydrocoll 2008;22:1170. [87] van Aken GA. In: Friberg SE, Larsson K, Sjöblom J, editors. Food emulsions. New York: Marcel Dekker; 2004. p. 299. [88] Sarkar A, Goh KKT, Singh RP, Singh H. Food Hydrocoll 2009;23:1563. [89] Sarkar A, Goh KKT, Singh H. Food Hydrocoll 2010;24:534. [90] Macierzanka A, Sancho AI, Mills ENC, Rigby NM, Mackie AR. Soft Matter 2009;5:538. [91] Maldonado-Valderrama J, Woodward NC, Gunning AP, Ridout MJ, Husband FA, Mackie AR, et al. Langmuir 2008;24:6759.

57

[92] Sarkar A, Horne DS, Singh H. Food Hydrocoll 2010;24:142. [93] Gargouri Y, Julien R, Bois AG, Verger R, Sarda L. J Lipid Res 1983;24:1336. [94] Armand M, Borel P, Ythier P, Dutot G, Melin C, Senft M, et al. J Nutr Biochem 1992;3:333. [95] Sarkar A, Horne DS, Singh H. Int Dairy 2010;20:589. [96] Armand M. Curr Opin Clin Nutr Metab Care 2007;10:156. [97] Armand M, Borel P, Dubois C, Senft M, Peyrot J, Salducci J, et al. Am J Physiol Gastrointest Liver Physiol 1994;266:G372. [98] Carriere F, Barrowman JA, Verger R, Laugier R. Gastroenterology 1993;105:876. [99] Pafumi Y, Lairon D, de la Porte PL, Juhel C, Storch J, Hamosh M, et al. J Biol Chem 2002;277:28070. [100] Hur SJ, Decker EA, McClements DJ. Food Chem 2009;114:253. [101] de Zwart LL, Haenen HEMG, Versantvoort CHM, Wolterink G, van Engelen JGM, Sips AJAM. Regul Toxicol Pharmacol 2004;39:282. [102] Versantvoort CHM, Oomen AG, Van de Kamp E, Rompelberg CJM, Sips AJAM. Food Chem Toxicol 2005;43:31. [103] Beysseriat M, Decker EA, McClements DJ. Food Hydrocoll 2006;20:800. [104] Wickham M, Faulks R, Mills C. Mol Nutr Food Res 2009;53:952. [105] da Silva Gomes RA, Batista RP, de Almeida AC, da Fonseca DN, Juliano L, Hial V. Anal Biochem 2003;316:11.