Influence of methylcellulose on attributes of β-carotene fortified starch-based filled hydrogels: Optical, rheological, structural, digestibility, and bioaccessibility properties

1 Influence of methylcellulose on attributes of β-carotene fortified 2 starch-based filled hydrogels: Optical, rheological, structural, 3 digestibility, and bioaccessibility properties 4 Saehun Munab*, Shinjae Park a, Yong-Ro Kima, & David Julian McClements b* 5 6 7 a 8 Science and Engineering, Seoul National University, Seoul, 151-742, Republic of Korea 9 b Center for Food and Bioconvergence, and Department of Biosystems and Biomaterials Department of Food Science, University of Massachusetts, Amherst, MA 01003 10 11 12 13 Journal: Food Research International 14 Submitted: March 23, 2016 15 Revised: June 9, 2016 16 17 Corresponding Author: David Julian McClements, Department of Food Science, University 18 of Massachusetts, Amherst, MA 01003, mcclements@foodsci.umass.edu. 19 20 1 © 2016. This manuscript version is made available under the Elsevier user license http://www.elsevier.com/open-access/userlicense/1.0/ 21 22 ABSTRACT There is considerable interest in controlling the gastrointestinal fate of nutraceuticals to 23 improve their efficacy. In this study, the influence of methylcellulose (an indigestible 24 polysaccharide) on lipid digestion and β-carotene bioaccessibility was determined. 25 carotenoids were encapsulated within lipid droplets that were then loaded into rice starch 26 hydrogels containing different methylcellulose levels. Incorporation of 0 to 0.2% of 27 methylcellulose had little impact on the dynamic shear rheology of the starch hydrogels, 28 which may be important for formulating functional foods with desirable textural attributes. 29 The microstructure, lipid digestion, and β–carotene bioaccessibility of the filled hydrogels 30 were measured as the samples were passed through simulated oral, gastric, and small 31 intestinal phases. The lipid digestion rate and carotenoid bioaccessibility decreased with 32 increasing methylcellulose. This effect was attributed to the ability of the methylcellulose to 33 inhibit molecular diffusion, promote droplet flocculation, or bind gastrointestinal components 34 thereby inhibiting triacylglycerol hydrolysis at the lipid droplet surfaces. This information 35 may be useful for rationally designing functional foods with improved nutritional benefits. 36 37 Keywords: methylcellulose; -carotene; lipid digestion; emulsions; nanoemulsions; 38 hydrogels; gastrointestinal tract; starch 39 40 41 42 43 44 45 2 The 46 47 1. Introduction The food industry is attempting to fortify many types of food and beverage products with 48 nutraceuticals. Nutraceuticals are bioactive agents found in foods that are not required for 49 normal human functioning (unlike nutrients, minerals, and vitamins), but that may improve 50 human health, wellbeing, and performance through their biological effects (McClements, 51 2013). Nutraceuticals include food components such as -3 oils, carotenoids, polyphenols, 52 phytosterols, and curcumin. 53 functional foods in their pure form because of physicochemical constraints, such as limited 54 solubility, chemical instability, and poor bioavailability (Tang & Zhong-Gui, 2007; Liang, 55 Shoemaker, Yang, Zhong, & Huang, 2013; McClements, 2013; Reboul, 2013). 56 challenges can often be overcome using food-grade colloidal delivery systems, such as 57 emulsions, nanoemulsions, microemulsions, solid lipid nanoparticles, and filled hydrogels 58 (Tokle, Lesmes, Decker, & McClements, 2012; Nik, Langmaid, & Wright, 2012; Verrijssen, 59 Balduyck, Christiaens, Loey, Buggenhout, & Hendrickx, 2014; Mun, Kim, Shin, & 60 McClements, 2015; Zhang, Xu, Jin, Shah, Li, & Li, 2015). 61 encapsulated within lipid droplets that can readily be dispersed in water, that can protect the 62 bioactives from degradation, and that are digested in the gastrointestinal tract (GIT) to form 63 mixed micelles that enhance bioaccessibility (McClements, 2013). 64 Many of these nutraceuticals cannot simply be introduced into These Lipophilic nutraceuticals can be Many commonly consumed food products have gel-like properties, such as some desserts, 65 confectionary, and meat-substitutes. These products are often prepared using food-grade 66 biopolymers that form a three-dimensional network that traps water and leads to a material 67 with viscoelastic properties. 68 encapsulated in lipid droplets and then dispersed in macroscopic starch hydrogels (Mun, Kim, 69 Shin, & McClements, 2015; Mun, Kim, & McClements, 2015). This research showed that β- 70 carotene bioaccessibility was higher in filled hydrogels (lipid droplets dispersed in hydrogels), 71 than in lipid droplets or in hydrogels alone. 72 starch hydrogels to prevent the lipid droplets from aggregating in the gastrointestinal tract, 73 which increased the accessibility of the lipase to the emulsified lipid phase. 74 75 In a previous study, our group showed that β-carotene could be This effect was attributed to the ability of the In the current study, the influence of the addition of methylcellulose on the microstructure, lipid digestibility, and β-carotene bioaccessibility of the filled starch hydrogels was examined 3 76 using a simulated GIT. Methylcellulose is an indigestible water-soluble polysaccharide that 77 is widely used in the food industry for its physicochemical and physiological properties 78 (Chawla & Patil, 2010). 79 foods, as well as to prevent constipation in supplements (Li & Nie, 2016). Previous studies 80 have shown that the microstructure of starch hydrogels can be appreciably altered by the 81 addition of indigestible polysaccharides such as guar gum, xanthan gum, and cellulose 82 derivatives (Techawipharat, Suphantharika, & BeMiller, 2008; Ptaszek, Beriski, Ptaszek, 83 Witczak, Repelewicz, & Grzesik, 2009; Gladkowska-Balewicz, Norton, & Hamilton, 2014). 84 Other studies have shown that interactions between starch and indigestible polysaccharides 85 may appreciably alter the functional properties of starch hydrogels, such as their appearance, 86 rheology, and stability (Baranowska, Sikora, Krystyjan, & Tomasik, 2011; Samutsri & 87 Suphantharika, 2012; Sikoral, Tomasik, & Krystyjan, 2010; Krystyjan, Adamczyk, Sikora, & 88 Tomasik, 2013). We therefore hypothesized that the addition of methylcellulose to filled 89 starch hydrogels would alter their microstructure, rheology, and gastrointestinal fate, which 90 could alter the bioaccessibility of any encapsulated nutraceuticals. 91 useful in the design of functional foods with enhanced health benefits, such as controlled 92 lipid digestion or nutraceutical bioaccessibility. 93 2. Materials and methods 94 2.1. Materials Methylcellulose is commonly used as a thickener and emulsifier in This information may be 95 The rice starch used for this study was isolated from native rice (Ilmi byeo, Korea) in a 96 laboratory using a traditional alkaline method (Lumdubwong & Seib, 2000). β-carotene, 97 pancreatin (from porcine pancreas) and bile extract (porcine) were purchased from Sigma 98 Aldrich (St. Louis, MO). 99 Foods International Inc. (BiPRO, Le Sueur, MN, USA). Corn oil was purchased from a Whey protein isolate (WPI) was kindly provided by Davisco 100 local supermarket. All other chemicals were of analytical grade. Double-distilled water 101 was used to prepare all solution and emulsions. 102 i.e., g of specified component per 100 g of sample. 103 2.2. Viscoelastic behavior 104 Most concentrations are reported as wt%, The dynamic viscoelastic properties of the samples were determined using a dynamic 4 105 shear rheometer (AR 1500 ex, TA instruments Ltd, New Castle, DE, USA) operating in 106 oscillatory mode with parallel plate geometry (20 mm diameter, 1 mm gap). 107 were prepared by dispersing rice starch powder in distilled water (10 wt% starch) and then 108 heating at 90 °C for 10 min in the absence and presence of emulsion and methyl cellulose. 109 After thermal treatment, the resulting hot paste was loaded between the parallel plates of the 110 rheometer that had previously been equilibrated at 4 °C and a thin layer of paraffin oil was 111 applied to the outer edges of the sample to prevent evaporation during measurement. A 112 dynamic frequency sweep test was conducted by applying a constant strain of 0.5%, which 113 was within the linear viscoelastic region (established in a preliminary test), over a frequency 114 range between 0.63 and 63 rad/s. 115 were then reported over a range of frequencies, and the moduli of all samples were reported 116 at a constant frequency of 10 rad/s. 117 2.3. Emulsion preparation 118 Starch pastes The storage (G’) and loss (G”) moduli of selected samples An oil phase was prepared by dispersing β-carotene (0.3%, w/w) in corn oil using a 119 sonicating water bath (Model 250, Ultrasonic cleaner, E/MC RAI research, Long Island, New 120 York) for 5 min and then heating (60 °C for 30 min) to ensure complete dissolution. An 121 aqueous emulsifier solution was prepared by dispersing WPI (1.06 wt%) in 10 mM phosphate 122 buffer (pH 7.0). 123 emulsifier solution (1.06 wt% WPI, pH 7.0) using a high speed blender for 2 min 124 (M133/1281-0, Biospec Products, Inc., ESGC, Switzerland) and then passing through a 125 microfludizer four times at 10 kpsi (Model 110L, Microfluidics, Newton, MA). 126 2.4. Preparation of filled hydrogels containing methylcellulose 127 A stock emulsion was prepared by homogenizing 6 wt% corn oil and 94 wt% Initially, the desired amount of methylcellulose was dissolved in 10 mM phosphate buffer 128 (pH 7.0) and then stock WPI-stabilized emulsion and rice starch were added into the 129 methylcellulose solution. The final concentrations of lipid, rice starch, and methylcellulose 130 contained in the mixtures were 4 wt% lipid, 10 wt% of rice starch, and 0, 0.05 wt%, 0.1 wt%, 131 or 0.2 wt% of methylcellulose. 132 a range of values that might be used in commercial products to modify their properties (Dar 133 and Light, 2014). These mixtures were then heated at 90 °C for 10 min to gelatinize the 134 starch granules. The heated mixtures were loaded into a flat, cylindrically shaped vessel, These levels of methylcellulose were selected so as to cover 5 135 and then stored at 4 °C overnight to allow the hydrogels to set. 136 2.5. Simulated gastrointestinal tract model 137 Samples were passed through an in vitro GIT model consisting of mouth, stomach, and 138 small intestine phases, which has been described in detail in previous studies (Salvia-Trujillo, 139 Qian, Martin-Belloso, & McClements, 2013a,b; Mun, Kim, & McClements, 2015; Lopez- 140 Pena, Zheng, Sela, Decker, Xiao, & McClements, 2016). Hence, the method is only briefly 141 described here. 142 and is fairly similar to the standardized method proposed recently (Minekus et al 2014). 143 For the mouth phase, the samples were mixed with simulated saliva fluid (SSF) This method has been widely used by our laboratory in previous studies, 144 containing mucin and various salts (Sarkar, Goh, & Singh, 2009) at a 50:50 vol/vol ratio. 145 The pH of the mixture was then adjusted to pH 6.8 and then incubated at 37 °C for 10 min 146 with continuous agitation. 147 typically spend in the mouth, but was used to ensure consistency from sample to sample. 148 For the stomach phase, the sample was mixed with simulated gastric fluid (SGF), which was 149 prepared by dissolving 2 g NaCl, and 7 mL of HCl (37%) in 1l of water and then adding 3.2 g 150 of pepsin (Sarkar, Goh, Singh, & Singh, 2009). 151 incubated at 37 °C for 2 h with continuous agitation at 100 rpm. 152 phase, 30 mL of samples from the gastric phase was placed in a temperature-controlled 153 (37 °C) chamber and the system was set at pH 7.0. 154 (187.5 mg/3.5 mL) and 1.5 mL of salt solution (10 mM of calcium chloride and 150 mM of 155 sodium chloride) were added to the samples and the mixture was adjusted to pH 7.0. 156 Afterwards, 2.5 mL of freshly prepared pancreatin suspension (187.5 mg/2.5 mL) dissolved 157 in phosphate buffer was added into the mixture and incubated at 37 °C for 2 h. 158 the sample was monitored using a pH-stat automatic titration unit (Metrohm USA Inc., 159 Riverview, FL) and the solution was maintained at pH 7.0 by adding 0.25 M NaOH to 160 neutralize any free fatty acids (FFA) released due to lipid digestion. The percentage of 161 FFAs released was calculated from the volume of alkaline solution required to neutralize the 162 samples using the following equation: This incubation period is longer than the time a food would The sample was then adjusted to pH 2.5 and For the small intestine Then 3.5 mL of bile extract solution VNaOH  mNaOH  Mlipid 6 The pH of % FFA = 100  163 wlipid  2 164 165 166 167 where VNaOH is the volume of titrant in liters, mNaOH is the molarity of sodium hydroxide, 168 Mlipid is the molecular weight of corn oil (872 g/mol), and wlipid is the weight of oil in the 169 digestion system in grams. Blanks (samples without oil) were run, and the volume of titrant 170 used for these blank samples was subtracted from the corresponding test samples that 171 contained oil. 172 2.6. Microstructure 173 To determine structural changes that occurred within different phases of the GIT model, 174 confocal scanning laser microscopy with a 60 objective lens and 10 eyepiece were used 175 (Nikon D-Eclipse C1 80i, Nikon, Melville, NY). 176 fat-soluble fluorescent dye that was previously dissolved at 0.1% (w/v) in ethanol. An air- 177 cooled argon ion laser (Model IMA 1010BOS, Melles Griot, Carlsbad, CA) was used to 178 excite Nile red at 488 nm. All images were taken and processed using the instrument 179 software program (EZ-CS1 version 3.8, Nikon, Melville, NY) 180 2.7. β-carotene bioaccessibility Samples were dyed by adding Nile red, a 181 After the small intestinal stage, the bioaccessibility of β–carotene was determined 182 using a method described previously (Qian, Decker, Xiao, & McClements, 2012; Salvia- 183 Trujillo, Qian, Martin-Belloso, & McClements, 2013a). 184 centrifuged at 2647g for 40 min at 25 °C (CL10 centrifuge, Thermo Scientific, Pittsburgh, PA, 185 USA). This process led to the generation of tubes that contained a sediment at the bottom 186 and a supernatant above. 187 phase containing the bioactive component. 188 phase were mixed with 5 mL of chloroform, vortexed and then centrifuged at 1158g for 189 10 min at 25 °C. The bottom layer containing the solubilized β-carotene was collected, 190 while the top layer was mixed with an additional 5 mL of chloroform and the same procedure 191 was repeated. The bottom chloroform layer was added to the previous one and analyzed Digesta were collected and The supernatant was collected and assumed to be the “micelle” Aliquots of 5 mL of the digesta or the micelle 7 192 spectrophotometrically (Ultraspec 3000 pro, GE Health Sciences, USA) at 450 nm. A 193 cuvette containing pure chloroform was used as a reference cell. 194 The concentration of β-carotene extracted from a sample was determined from a 195 calibration curve of absorbance versus β-carotene concentration in chloroform. The 196 bioaccessibility was then calculated using the following equation: 197 198 Bioaccessibility = 100  ( CMicelle/CDigesta) 199 200 where Cmicelle and CDigesta are the concentration of β-carotene in the micelle fraction and in 201 the raw digesta after the pH-stat experiment, respectively (Qian, Decker, Xiao, & 202 McClements, 2012; Salvia-Trujillo et al., 2013 ab). 203 not take into account any chemical changes of the β-carotene, and so the calculated 204 bioaccessibility is only based on the ratios of carotenoids that can be measured by UV-visible 205 spectrometry in the micelle phase and digesta. It should be noted that this method does 206 207 208 209 2.8. Statistical analysis All experiments were performed in at least triplicate using freshly prepared samples. Means and standard deviations were calculated from these data. 210 211 3. Results and discussion 212 3.1. Influence of methylcellulose on rheological and optical properties 213 The rheological properties of filled starch hydrogels were characterized in the absence 214 and presence of methylcellulose. The storage modulus (G') of all the samples was much 215 larger than the loss modulus (G'') across the entire frequency range tested (Fig. 1a), which 216 indicated that the samples were predominantly elastic rather than fluid. Both G’ and G” 217 increased with increasing measurement frequency indicating that the gels became stiffer at 218 higher frequencies, which was presumably because they had a characteristic relaxation time 219 within the frequency range used in our measurements. The influence of methylcellulose 220 concentration on the rheological properties of the filled starch hydrogels was determined at a 8 There was little change in the storage modulus (G’) of 221 fixed frequency (10 rad/s) (Fig. 1b). 222 the hydrogels with increasing methylcellulose concentration, but there was a slight increase 223 in the loss modulus (G”) from 30.0 Pa (0% MC) to 39.4 Pa (0.2% MC). 224 suggest that addition of relatively low levels of methylcellulose to the hydrogels only had a 225 modest impact on their dynamic shear rheology, which may be important when designing 226 functional foods fortified with methylcellulose. 227 unfilled starch hydrogels containing methylcellulose were similar to those of the filled starch 228 hydrogels (data not shown), indicating that the fat droplets also did not make a major 229 contribution to the overall rheology of the filled hydrogels. 230 small-strain oscillatory measurements used in this study do not provide information about the 231 rheological response of the starch hydrogels under conditions that they may experience in 232 practice, such as during food manufacturing, food preparation, or mastication. 233 These results The measured rheological properties of the It should be noted that the The formation of starch hydrogels involves the swelling and bursting of starch granules 234 during heating followed by the release of amylose and amylopectin molecules (Krystyjan, 235 Adamczyk, Sikora, & Tomasik, 2013). The released starch molecules may then form a 236 three-dimensional network with some elastic-like properties due to cross-linking of the 237 amylose molecules through hydrogen bonding of helical regions (Eidam, Kulicke, Kuhn, & 238 Stute, 1995; Yang, Irudayaraj, Otgonchimeg, & Walsh, 2004; Samutsri & Suphantharika, 239 2012). Previous studies have reported that indigestible polysaccharides may disrupt the 240 formation of starch hydrogels by interfering with the formation of these cross-links (Eidam, 241 Kulicke, Kuhn, & Stute, 1995). 242 may have been because the methylcellulose concentration used was relatively low or because 243 the methylcellulose remained in the aqueous phase between the starch molecule cross-links. However, we did not observe this effect in our study, which 244 The β–carotene enriched emulsions and filled starch hydrogels had a bright yellowish- 245 orange color, which can be attributed to the selective absorption of light by the carotenoids 246 (Mun, Kim, Shin, & McClements, 2015; Mun, Kim, & McClements, 2015). 247 contain numerous conjugated double bonds that enable them to strongly absorb visible light. 248 Visual observation of the samples indicated that they had a uniform appearance throughout 249 the study (data not shown), suggesting that the β–carotene did not form crystals that sediment 250 and that the lipid droplets did not cream. 9 Carotenoids 251 252 3.2. Influence of methylcellulose on microstructure in GIT The initial mean particle diameter (d4,3) of the WPI-stabilized emulsions used to prepare 253 the filled hydrogels was 0.26 μm, and the initial particle size distribution was monomodal 254 (data not shown). The changes in the particle size and microstructure of these emulsions as 255 they passed through the simulated GIT were similar to those observed using a similar system 256 in our previous studies (Mun, Kim, Shin, & McClements, 2015; Mun, Kim, & McClements, 257 2015), and so are not described in detail here. 258 droplets was observed in the oral and gastric phases (Fig. 2a), which can be attributed to a 259 number of effects: depletion and/or bridging flocculation by mucin in the simulated saliva; 260 changes in pH and ionic strength in the different GIT regions; and hydrolysis of adsorbed 261 proteins by pepsin in the gastric phase. 262 digested by lipase and therefore largely disappeared from the confocal images (Fig. 2a). Briefly, extensive aggregation of the fat In the small intestine phase the fat droplets were 263 The semi-solid structure of all the filled hydrogels remained intact after exposure to the 264 oral and gastric phases, so it was not possible to measure the particle size distribution of these 265 samples using light scattering. Consequently, only confocal microscopy was used to 266 monitor changes in the microstructure of these samples as they passed through the simulated 267 GIT. 268 methylcellulose are shown in Figure 2b. 269 dispersed throughout the starch hydrogels. Upon exposure to simulated oral and gastric 270 fluids, the fat droplets remained relatively small and were still fairly evenly dispersed 271 throughout the hydrogels, which is in contrast to the extensive fat droplet aggregation 272 observed in the emulsions (Fig. 2a). 273 inhibited the aggregation of the fat droplets, presumably by forming a highly viscous aqueous 274 phase that retarded droplet movement. The microstructural changes in the filled starch hydrogels in the absence of Initially, fine fat droplets were fairly evenly This suggests that the presence of the starch hydrogel 275 The incorporation of 0.05% methylcellulose into the filled starch hydrogels did not 276 affect their initial microstructure. Again, the fat droplets remained relatively small and evenly 277 distributed throughout the hydrogel when the samples were exposed to simulated oral and 278 gastric conditions (Fig. 2c). 279 hydrogels, there appeared to be some fat droplet aggregation in the initial samples (Fig. 2d), 280 which may have been due to thermodynamic incompatibility of the biopolymers or due to However, when 0.2% methylcellulose was present in the filled 10 281 depletion flocculation of the fat droplets. 282 fairly evenly dispersed within the mouth and stomach phases, suggesting that they were still 283 trapped in a hydrogel network. 284 Nevertheless, the fat droplets still appeared to be After exposure to small intestinal conditions, the hydrogel structure disintegrated and the 285 samples became fluid-like, which may have been due to dilution effects and/or amylase 286 activity (Mun, Kim, Shin, & McClements, 2015). 287 biopolymer concentration, whereas amylase activity will have led to hydrolysis of some of 288 the starch molecules, thereby leading to breakdown of the hydrogel network. 289 after exposure to the small intestine phase there appeared to be an increase in the number of 290 large lipid-rich particles within the samples as the methylcellulose level increased (Fig. 2d). 291 Results from previous studies suggest that these lipid-rich particles could be undigested Dilution will have decreased the overall Interestingly, 292 fat droplets, micelles, and/or vesicles (McClements, Decker, & Park, 2009; Singh, Ye, & 293 Horne, 2009). 294 Lipase molecules adsorb to the fat droplet surfaces and hydrolyze the triacylglycerols (TAGs) 295 into free fatty acids (FFAs) and monoacylglycerols (MAGs). 296 MAGs are then solubilized by bile salts or precipitated by calcium ions, which helps remove 297 them from the fat droplet surfaces during lipid digestion. 298 presence of the methylcellulose may have interfered with this process, thereby inhibiting the 299 digestion of the emulsified TAGS. 300 that may have inhibited the diffusion of lipase molecules to the fat droplet surfaces, or that 301 inhibited the removal of FFAs and MAGs from the lipid droplet surfaces. 302 methylcellulose may have altered the structure of the colloidal particles in the mixed micelle 303 phase by binding to FFAs, MAGs, or bile salts. 304 formation of large vesicles rather than small micelles. 305 to establish the precise molecular and physicochemical origin of the ability of 306 methylcellulose to alter the gastrointestinal fate of filled hydrogels. 307 3.3. Influence of methylcellulose on lipid digestibility Usually, fat droplets are hydrolyzed by lipase in the small intestine phase. The resulting FFAs and Our results suggest that the The methylcellulose is an indigestible polysaccharide Alternatively, the These interactions may have promoted the Clearly, further research is required 308 In this series of experiments the impact of incorporating methylcellulose into the filled 309 starch hydrogels on the rate and extent of lipid digestion was measured under simulated small 310 intestinal conditions (Fig. 3). At the end of 2 h digestion, the overall extent of lipid digestion 11 311 was fairly similar in the presence and absence of methylcellulose, regardless of its 312 concentration. 313 filled hydrogels containing 0.2% methylcellulose than for the ones containing lower levels. 314 These results suggested that sufficiently high levels of methylcellulose were able to delay 315 lipid digestion, possibly by restricting the access of lipase to the droplet surfaces or by 316 inhibiting the removal of FFAs and MAGs from the droplet surfaces as mentioned previously. 317 However, the rate of lipid digestion appeared to be somewhat slower for the Tokle et al. (2012) examined the influence of dietary fiber on the digestion of fat droplets 318 under simulated small intestinal conditions. They also reported that the dietary fiber did not 319 prevent lipid digestion, but that it did slow down the rate of lipid digestion. In their study, 320 anionic dietary fibers (pectin) were used, which may have interacted with cationic calcium 321 ions and formed a gel-network that inhibited lipase diffusion or prevented calcium soaps of 322 FFAs being formed. Other studies have suggested that dietary fibers may impact lipid 323 digestion by binding to various components in the GIT, such as calcium ions, bile salts, free 324 fatty acids, and lipase (Palafox-Carlos, Ayala-Zavala, & Gonzalez-Aguilar, 2011; Verrijssen, 325 Balduyck, Christiaens, Loey, Buggenhout, & Hendrickx, 2014; Qiu, Zhao, Decker, & 326 McClements, 2015). A differential scanning calorimetry study of cellulose ester - bile salt 327 mixtures showed that the bile salts affected the thermal gelation of the cellulose derivatives, 328 which suggested that there was some kind of binding interaction (Torcello-Gόmez, & Foster, 329 2014). Consequently, methylcellulose may have interfered with the normal role of bile salts 330 in the lipid digestion process, i.e., displacing other surface-active molecules from the fat 331 droplet surfaces or solubilizing lipolysis products such as FFAs and MAGs (Bellesi, Ruiz- 332 Henestrosa, & Pilosof, 2014; Euston, Baird, Campbell, & Kuhns, 2013; Li, Hu, & 333 McClements, 2011; Mun, Decker, & McClements, 2007; Nik, Wright, & Corredig, 2011). 334 Thus, the retardation of lipid digestion observed at high methylcellulose levels may have 335 been partially due to binding of bile salts. Alternatively, it may have been due to the ability 336 of the methylcellulose to increase the viscosity of the aqueous phase around the lipid droplets, 337 thereby slowing down the diffusion of reactants, catalysts, and products involved in the lipid 338 digestion process (Palafox-Carlos, Ayala-Zavala, & Gonzalez-Aguilar, 2011; Tokle, Lesmes, 339 Decker, & McClements, 2012; Verrijssen, Balduyck, Christiaens, Loey, Buggenhout, & 340 Hendrickx, 2014). 12 341 342 3.4. Influence of methylcellulose on β–carotene bioaccessibility Finally, the impact of the composition of the hydrogel matrix on β–carotene 343 bioaccessibility was measured (Fig. 4). 344 ingested compound that is solubilized in the gastrointestinal fluids in a form available for 345 absorption (Hedrén, Diaz, & Svanberg, 2002). 346 taken to be the fraction of the β–carotene that was released from the hydrogels and 347 solubilized within the mixed micelles formed after lipid digestion. In general, bioaccessibility is the fraction of an In the current study, the bioaccessibility was 348 In our previous study, the bioaccessibility of β–carotene in protein-coated lipid droplets 349 was found to be higher when they were incorporated into a starch hydrogel than when they 350 were in a free state, which was attributed to the ability of the hydrogel to inhibit extensive 351 droplet flocculation in the GIT (Mun, Kim, Shin, & McClements, 2015; Mun, Kim, & 352 McClements, 2015). 353 droplet surfaces and convert the TAGs to MAGs and FFAs that could form mixed micelles 354 capable of solubilizing the carotenoids. 355 the bioaccessibility of β–carotene was 16% and 66% in emulsions and filled hydrogels 356 containing no methylcellulose, respectively (Fig. 4). These values are in good agreement 357 with those found in our previous studies (Mun, Kim, Shin, & McClements, 2015; Mun, Kim, 358 & McClements, 2015). 359 methylcellulose concentration in the filled hydrogels increased (Fig. 4), despite the fact that 360 the final extent of lipid digestion did not appear to be strongly affected by methylcellulose 361 level (Fig 3). There are a number of possible reasons for this phenomenon. 362 As a result the lipase molecules could more easily adsorb to the lipid A similar finding was observed in the current study: Interestingly, the bioaccessibility of β–carotene decreased as the First, methylcellulose molecules might have altered the nature of the mixed micelles 363 formed by lipid digestion, i.e., by interacting with bile salts, FFAs, or MAGs. 364 confocal microscopy images of the samples indicated that there were more large lipid-rich 365 particles formed after lipid digestion in the presence of high levels of methylcellulose (Fig. 366 2d). These lipid-rich particles may have been large liposomes, rather than small micelles, 367 which could have sedimented to the bottom of the samples during the centrifugation step 368 carried out as part of the bioaccessibility analysis. Second, the methylcellulose may have 369 directly bound to the released β–carotene molecules and formed a dense molecular complex 370 that sedimented during the centrifugation step. Further study is clearly needed to determine 13 Indeed, the 371 the physicochemical origin of the ability of methylcellulose to reduce the bioaccessibility of 372 β–carotene in filled starch hydrogels. 373 4. Conclusions 374 The impact of the composition of filled hydrogels on lipid digestion and β–carotene- 375 bioaccessibility was determined. 376 methylcellulose concentrations, however, the final extent of lipid digestion was unaffected. 377 The bioaccessibility of β–carotene was lower in protein-coated lipid droplets when they were 378 in a free state than when they were encapsulated in filled hydrogels, which was attributed to 379 the ability of the starch hydrogel to inhibit extensive lipid droplet flocculation. 380 incorporation of methylcellulose into the filled starch hydrogels reduced β–carotene 381 bioaccessibility, which could be due to the ability of this dietary fiber to interact with specific 382 components in the GIT or alter their diffusion. The information obtained from this study 383 might be helpful for designing functional foods that perform in a specific way within the 384 gastrointestinal tract. 385 5. Acknowledgements 386 The rate of lipid digestion was reduced at higher The This material was partly based upon work supported by the 387 Cooperative State Research, Extension, Education Service, USDA, Massachusetts Agricultur 388 al Experiment Station (MAS00491) and USDA, NRI Grants (2013-03795). This research 389 was partly supported by Basic Science Research Program through the National Resear 390 ch Foundation of Korea (NRF) grants funded by the Ministry of Science, ICT, and F 391 uture Planning (NRF-2015R1A1A3A04001485). 392 393 394 395 396 14 397 6. References 398 Baranowska, H., Sikora, M., Krystyjan, M., & Tomasik, P. (2011). Analysis of the formation 399 of starch — hydrocolloid binary gels and their structure based on the relaxation times of 400 the water molecules. Polimery, 56, 478-483. 401 402 403 404 405 406 407 408 Bellesi, F. A., Ruiz-Henestrosa, V. M. P., & Pilosof, A. M. R. (2014). Behavior of protein interfacial films upon bile salts addition. Food Hydrocolloids, 36, 115-122. Chawla, R., & Patil, G. R. (2010). Soluble dietary fiber. Comprehensive Reviews In Food Science And Food Safety, 9, 178-196. Dar, Y.L. and Light, J.M. (2014). Food Texture Design and Optimization. John Wiley and Sons, New York, N.Y., pp 128-159. Eidam, D., Kulicke, W. M., Kuhn, K., & Stute, R. (1995). Formation of maize starch gels selectively regulated by the addition of hydrocolloids. Starch, 10, 378-384. 409 Euston, S. R., Baird, W. G., Campbell, L., & Kuhns, M. (2013). Competitive adsorption of 410 dihydroxy and trihydroxy bile salts with whey protein and casein in oil-in-water 411 emulsions. Biomacromolecules, 14, 1850-1858. 412 Gladkowska-Balewicz, I., Norton, I. T., & Hamilton, I. E. (2014). Effect of process 413 conditions, and component concentrations on the viscosity of κ–carrageenan and 414 pregelatinised cross-linked waxy maize starch mixed fluid gels. Food Hydrocolloids, 42, 415 355-361. 416 Hedrén, E., Diaz, V., & Svanberg, U. (2002). Estimation of carotenoid accessibility from 417 carrots determined by an in vitro digestion method. European Journal of Clinical 418 Nutrition, 56, 425-430. 419 Krystyjan, M., Adamczyk, G., Sikora, M., & Tomasik, P. (2013). Long-term storage stability 420 of selected potato starch-Non-starchy hydrocolloid binary gels. Food Hydrocolloids, 31, 421 270-276. 422 423 Li, Y., Hu, M., & McClements, D. J. (2011). Factors affecting lipase digestibility of emulsified lipids using an in vitro digestion model: Proposal for a standardised pH-stat 15 424 425 426 427 method. Food Chemistry, 126, 498-505. Li, J. M., & Nie, S. P. (2016). The functional and nutritional aspects of hydrocolloids in foods. Food Hydrocolloids, 53, 46-61. Liang, R., Shoemaker, C. F., Yang, X., Zhong, F., & Huang, Q. (2013). Stability and 428 bioaccessibility of -carotene in nanoemulsions stabilized by modified starches. 429 Journal of Agricultural and Food Chemistry, 61, 1249-1257 430 Lopez-Pena, C. L., Zheng, B., Sela, D. A., Decker, E. A., Xiao, H., & McClements, D. J. 431 (2016). Impact of ε-polylysine and pectin on the potential gastrointestinal fate of 432 emulsified lipids: In vitro mouth, stomach and small intestine model. Food Chemistry, 433 192, 857-864. 434 435 436 Lumdubwong, N., & Seib, P. A. (2000). Rice starch isolation by alkaline protease digestion of wet-milled rice flour. Journal of Cereal Science, 31, 63–74. McClements, D. J. (2013). Utilizing food effects to overcome challenges in delivery of 437 lipophilic bioactives: structural design of medical and functional food. Expert Opinion 438 on Drug Delivery, 10, 1621-1632. 439 McClements, D. J., Decker, E. A., & Park, Y. (2009). Controlling lipid bioavailability through 440 physicochemical and structural approaches. Critical Reviews in Food Science and 441 Nutrition, 49, 48-67. 442 Minekus, M., M. Alminger, P. Alvito, S. Ballance, T. Bohn, C. Bourlieu, F. Carriere, R. 443 Boutrou, M. Corredig, D. Dupont, C. Dufour, L. Egger, M. Golding, S. Karakaya, B. 444 Kirkhus, S. Le Feunteun, U. Lesmes, A. Macierzanka, A. Mackie, S. Marze, D. J. 445 McClements, O. Menard, I. Recio, C. N. Santos, R. P. Singh, G. E. Vegarud, M. S. J. 446 Wickham, W. Weitschies and A. Brodkorb (2014). A standardised static in vitro 447 digestion method suitable for food - an international consensus. Food & Function 5, 448 1113-1124. 449 Mun, S., Decker, E. A., & McClements, D. J. (2007). Influence of emulsifier type on in vitro 450 digestibility of lipid droplets by pancreatic lipase. Food Research International, 40, 451 770-781. 16 452 453 454 Mun, S. H., Kim, Y. R., & McClements, D. J. (2015). Control of β–carotene bioaccessibility using starch-based filled hydrogels. Food Chemistry, 173, 454-461. Mun, S. H., Kim, Y. R., Shin, M. S., & McClements, D. J. (2015). Control of lipid digestion 455 and nutraceutical bioaccessibility using starch-based filled hydrogels: Influence of 456 starch and surface type. Food Hydrocolloids, 44, 380-389. 457 Nik, A. M., Wright, A. J., & Corredig, M. (2011). Impact of interfacial composition on 458 emulsion digestion and rate of lipid hydrolysis using different in vitro digestion models. 459 Colloids and Surfaces B-Biointerfaces, 83, 321-330. 460 Nik, A.M., Langmaid, S., & Wright, A. J. (2012). Digestibility and β–carotene release from 461 lipid nanodispersions depend on dispersed phase crystallinity and interfacial properties. 462 Food & Function, 3, 234-245. 463 Palafox-Carlos, H., Ayala-Zavala, J. F., & Gonzalez-Aguilar, G. A. (2011). The role of dietary 464 fiber in the bioaccessibility and bioavailability of fruit and vegetable antioxidants. 465 Journal of Food Science, 76, R6-R15. 466 Ptaszek, A., Beriski, W., Ptaszek, P., Witczak, T., Repelewicz, U., & Grzesik, M. (2009). 467 Viscoelastic properties of waxy maize starch and selected non-starch hydrocolloids gels. 468 Carbohydrate Polymers, 76, 567-577. 469 Qian, C., Decker, E. A., Xiao, H., & McClements, D. J. (2012). Nanoemulsion delivery 470 systems: Influence of carrier oil on β–carotene bioaccessibility. Food Chemistry, 135, 471 1440-1447. 472 Qiu, C., Zhao, M., Decker, E. A., & McClements, D. J. (2015). Influence of anionic dietary 473 fibers (xanthan gum and pectin) on oxidative stability and lipid digestibility of wheat 474 protein-stabilized fish oil-in-water emulsion. Food Research International, 74, 131-139. 475 476 Reboul, E. (2013). Absorption of vitamin A and carotenoids by the enterocyte: focus on transport proteins. Nutrients, 5, 3563-3581. 477 Salvia-Trujillo, L., Qian, C., Martin-Belloso, O., & McClements, D. J. (2013a). Modulating 478 β-carotene bioaccessibility by controlling oil composition and concentration in edible 479 nanoemulsions. Food Chemistry, 139, 878-884. 17 480 Salvia-Trujillo, L., Qian, C., Martin-Belloso, O., & McClements, D. J. (2013b). Influence of 481 particle size on lipid digestion and β–carotene bioaccessibility in emulsions and 482 nanoemulsions. Food Chemistry, 141, 1472-1480. 483 Samutsri, W., & Suphantharika, M. (2012). Effect of salts on pasting, thermal, and 484 rheological properties of rice starch in the presence of non-ionic and ionic hydrocolloids. 485 Carbohydrate Polymers, 87, 1559-1568. 486 487 488 Sarkar, A., Goh, K. K. T., & Singh, H. (2009). Colloidal stability and interactions of milkprotein-stabilized emulsions in an artificial saliva. Food Hydrocolloids, 23, 1270–1278. Sarkar, A., Goh, K. K. T., Singh, R. P., & Singh, H. (2009). Behaviour of an oil-in-water 489 emulsion stabilized by β-lactoglobulin in an in vitro gastric model. Food Hydrocolloids, 490 3, 1563–1569. 491 492 493 494 Sikoral, M., Tomasik, P., & Krystyjan, M. (2010). Characterization of potato starch fractions and their interaction with hydrocolloids. Starch, 63, 341-349. Singh, H., Ye, A. Q., & Horne, D. (2009). Structuring food emulsions in the gastrointestinal tract to modify lipid digestion. Progress in Lipid Research, 48, 92-100 495 Tang, J. J. S., & Zhong-Gui, H. (2007). Self-emulsifying drug delivery systems: strategy for 496 improving oral delivery of poorly soluble drugs. Current Drug Therapy, 2, 85-93. 497 Techawipharat, J., Suphantharika, M., & BeMiller, J. N. (2008). Effects of cellulose 498 derivatives and carrageenans on the pasting, paste, and gel properties of rice starches. 499 Carbohydrate Polymers, 73, 417-426. 500 Tokle, T., Lesmes, U., Decker, E. A., & McClements, D. J. (2012). Impact of dietary fiber 501 coating on behavior of protein-stabilized lipid droplets under simulated gastrointestinal 502 conditions. Food & Function, 3, 58-66. 503 Torcello-Gόmez, A., & Foster, T. J. (2014). Interactions between cellulose ethers and a bile 504 salt in the control of lipid digestion of lipid-based systems. Carbohydrate Polymers, 113, 505 53-61. 506 Verrijssen, T. A .J., Balduyck, L. G., Christiaens, S., Loey, A. M. V., Buggenhout, S. V., & 18 507 Hendrickx, M. E. (2014). The effect of pectin concentration and degree of methyl- 508 esterification on the in vitro bioaccessibility of β–carotene-enriched emulsions. Food 509 Research International, 57, 71-78. 510 511 512 Yang, H., Irudayaraj, J., Otgonchimeg, S., & Walsh, M. (2004). Rheological study of starch and dairy ingredient-based food systems. Food Chemistry, 86, 571-578. Zhang, C., Xu, W., Jin, W., Shah, B. R., Li, Y., & Li, B. (2015). Influence of anionic alginate 513 and cationic chitosan on physicochemical stability and carotenoids bioaccessibility of 514 soy protein isolate-stabilized emulsions. Food Research International, 77, 419-425. 515 19 Figure Captions Figure Captions Figure 1a. Influence of methylcellulose addition and measurement frequency on the dynamic shear properties of filled hydrogels measured at 4 °C using oscillatory viscoelastic analysis. The samples contained either 0 or 0.2% methylcellulose. The black points represent the storage modulus (G’) and the white points represent the loss modulus (G”). Figure 1b. Influence of methylcellulose level on the storage and loss components of the shear modulus of filled hydrogels measured at 4 °C using oscillatory viscoelastic analysis (10 rads/sec). Figure 2. Confocal micrographs showing microstructural changes during passage through simulated oral, gastric, and intestinal phases; (A) emulsions; (B) filled starch hydrogels without methylcellulose, (C) filled starch hydrogels containing 0.05% methylcellulose; and (D) filled starch hydrogels containing 0.2% methylcellulose Figure 3. Calculated FFA released from rice starch-based filled hydrogels containing different concentration of methylcellulose. Digestion was retarded in the presence of the highest methylcellulose level Figure 4. Bioaccessibility (%) of β-carotene incorporated in WPI-stabilized emulsion and rice starch-based filled hydrogels containing methylcellulose (MC) after in vitro digestion. Figures 0%MC G' 0%MC G'' 0.2%MC G' 0.2% MC G'' G' & G'' (Pa) 200 150 100 50 0 1 10 Frequency (rad/s) 100 Figure 1a. Influence of methylcellulose addition and measurement frequency on the dynamic shear properties of filled hydrogels measured at 4 °C using oscillatory viscoelastic analysis. The samples contained either 0 or 0.2% methylcellulose. The black points represent the storage modulus (G’) and the white points represent the loss modulus (G”). 140 G' G" Shear Modulus (Pa) 120 100 80 60 40 20 0 0 0.05 0.1 0.2 Methylcellulose Level (wt%) Figure 1b. Influence of methylcellulose level on the storage and loss components of the shear modulus of filled hydrogels measured at 4 °C using oscillatory viscoelastic analysis (10 rads/sec). Before digestion Mouth Stomach Intestine (A) (B) (C) (D) Figure 2. Confocal micrographs showing microstructural changes during passage through simulated oral, gastric, and intestinal phases; (A) emulsions; (B) filled starch hydrogels without methylcellulose, (C) filled starch hydrogels containing 0.05% methylcellulose; and (D) filled starch hydrogels containing 0.2% methylcellulose 90 FFA Released (%) 80 70 60 50 40 30 FH-4% 20 MC-0.05 MC-0.1 10 MC-0.2 0 0 20 40 60 80 100 120 Digestion Time (min) Figure 3. Calculated FFA released from rice starch-based filled hydrogels containing different concentration of methylcellulose. Digestion was retarded in the presence of the highest methylcellulose level. 80 Bioaccessibility (%) 70 60 50 40 30 20 10 0 WPI Emulsion FH(MC-0) FH(MC-0.05) FH(MC-0.1) FH(MC-0.2) Figure 4. Bioaccessibility (%) of β-carotene incorporated in WPI-stabilized emulsion and rice starch-based filled hydrogels containing methylcellulose (MC) after in vitro digestion. *Graphical Abstract Manuscript title: Influence of methylcellulose on bioaccessibility of β-carotene incorporated within starch-based filled hydrogels, by Mun et al Graphical Abstract Before digestion Mouth Stomach Intestine (A) (B) (C) (D) Confocal micrographs showing microstructural changes during passage through simulated oral, gastric, and intestinal phases; (A) emulsions; (B) filled starch hydrogels without methylcellulose, (C) filled starch hydrogels containing 0.05% methylcellulose; and (D) filled starch hydrogels containing 0.2% methylcellulose