Agarose-based freeze-dried capsules prepared by the oil-induced biphasic hydrogel particle formation approach for the protection of sensitive probiotic bacteria

   1  Agarose-based freeze-dried capsules prepared by 1 the oil-induced biphasic hydrogel particle 2 formation approach for the protection of sensitive 3 probiotic bacteria 4 5 Ali Alehosseini1,2, Eva-María Gomez del Pulgar3, Maria José Fabra2, Laura G. Gómez-6 Mascaraque2, Alfonso Benítez-Páez3, Mahboobe Sarabi-Jamab4, Behrouz Ghorani1, Amparo 7 Lopez-Rubio2* 8 9 1 Department of Food Nanotechnology, Research Institute of Food Science and Technology 10 (RIFST), Km 12 Mashhad-Quchan Highway, P.O. Box: 91895/157/356, Mashhad, Iran 11 2 Food Preservation and Food Quality Department, IATA-CSIC, Avda. Agustin Escardino 7, 12 46980 Paterna, Valencia, Spain 13 3 Microbial Ecology, Nutrition and Health Group, IATA-CSIC, Avda. Agustin Escardino 7, 14 46980 Paterna, Valencia, Spain 15 4 Department of Food Biotechnology, Research Institute of Food Science and Technology 16 (RIFST), Km 12 Mashhad-Quchan Highway, P.O. Box: 91895/157/356, Mashhad, Iran 17 18 *Corresponding author: E-mail address: [email protected] (A. López-Rubio) 19 Tel.: +34 963900022; fax: +34 963636301 20    2  ABSTRACT 21 This work reports on a simple, fast and food-grade encapsulation approach with potential for 22 probiotic protection, which consists on the formation of agarose-based hydrogel particles. 23 These were generated by the supramolecular self-assembly of the probiotic-containing 24 carbohydrate solutions by dripping them into a biphasic bath with an upper oil layer, which 25 acted as the particle-forming antisolvent, and a lower aqueous layer, where the hydrogel 26 particles were collected. This technique, which we have named “oil-induced biphasic 27 hydrogel particle formation”, has been used to encapsulate the sensitive strain 28 Bifidobacterium pseudocatenulatum CECT 7765. In order to avoid agarose gelling at 40ºC 29 before the encapsulation process, this seaweed-derived carbohydrate was combined with 30 other hydrocolloids (alginate, whey protein concentrate, and gelatin), and the obtained 31 probiotic-containing hydrogel particles were subsequently freeze-dried. The protection ability 32 of this method versus directly freeze-drying the probiotic-containing solutions was 33 demonstrated during storage and simulated in-vitro digestion. Both the formation of a 34 continuous layer surrounding the bacteria and the optimal combination of materials (agarose 35 providing suitable oxygen barrier and WPC with proven probiotic affinity) rendered 36 encapsulation systems keeping viability levels required for commercial applications. 37 38 Keywords: agarose, encapsulation, probiotics, freeze-drying, GIT, WPC 39    3  1. INTRODUCTION 40 Recent years have witnessed a growing interest in the development of encapsulation 41 techniques and/or formulations for preserving probiotic bacteria viability during 42 commercialization and gastrointestinal tract (GIT) passage, given their potential for the 43 prevention and treatment of several pathologies (Bruce-Keller, Salbaum, & Berthoud, 2018; 44 Tonucci et al., 2017). Amongst probiotics, Lactobacillus and Bifidobacterium are the main 45 genera of probiotic bacteria used in the food industry since, apart from the proven health 46 benefits of some strains (Kabeerdoss et al., 2011; Pitino et al., 2010), they are inhabitants of 47 the normal intestinal microbiota of animals and humans (de Vrese & Schrezenmeir, 2008; 48 Murguia-Peniche et al., 2013). However, probiotics should be alive, metabolically active, and 49 abundant (at least 107 CFU/g) in the final product and, desirably, during the GIT passage to 50 improve or guarantee their efficacy (Ying et al., 2010). Bifidobacteria are particularly 51 sensitive to stress conditions during processing, showing a low survival rate during storage 52 and consumption of food products (Simoes et al., 2017). 53 A broadly explored approach for probiotic protection has been the development of food-grade 54 micro-hydrogels as encapsulation matrices (Ghibaudo, Gerbino, Campo Dall’ Orto, & 55 Gómez-Zavaglia, 2017; González-Ferrero, Irache, & González-Navarro, 2018; Huq, Khan, 56 Khan, Riedl, & Lacroix, 2013; McClements, 2017; Ramos et al., 2018; Zhao et al., 2018). 57 Alginate-based micro-hydrogels have been the most widely explored and they have 58 demonstrated to efficiently increase the viability of different probiotic strains (Shaharuddin & 59 Muhamad, 2015; Smidsrod & Skjak-Braek, 1990; Sohail, Turner, Coombes, & Bhandari, 60 2013; Sousa et al., 2015; Yeung, Ucok, Tiani, McClements, & Sela, 2016; Yeung, Arroyo-61 Maya, McClements, & Sela, 2016). Recently, pectin and cellulose-based micro-hydrogels 62 have also shown to increase the viability of Lactobacillus strains (Li, Luo et al., 2016; Li, 63 Zhang et al., 2016). Agarose, a structural polysaccharide occurring in different species of red 64    4  seaweed (Rhodophyceae), is another polysaccharide with gel forming ability and with a great 65 and unexplored potential for probiotic encapsulation, given its high oxygen barrier properties 66 and its long track use for culturing anaerobic bacteria (Katopo, Kasapis, & Y. Hemar, 2012; 67 Yokoyama, Kishida, Uchimura, & Ichinole, 2006). The ability of agarose to prevent the entry 68 of oxygen in liquid media is especially relevant when manipulating sensitive strains with 69 beneficial health effects, such as Bifidobacterium pseudocatenulatum CECT 7765, isolated 70 from the stools of a breast-fed infant (Benítez-Páez, Moreno, Sanz, & Sanz, 2016). This 71 bifidobacteria has been reported to reduce obesity-associated inflammation by restoring the 72 lymphocyte-macrophage balance and gut microbiota structure in high-fat diet-fed mice 73 (Gauffin, Santacruz, Trejo, & Sanz, 2013; Moya-Perez, Neef, & Sanz, 2015). 74 Therefore, the aim of this work was to explore the potential of agarose as the main matrix for 75 hydrogel particle formation to be used for the protection of the sensitive probiotic strain B. 76 pseudocatenulatum CECT 7765. In order to minimize moisture and oxygen exposure during 77 hydrogel particle formation, a new strategy was developed in which the probiotic-containing 78 agarose-based solutions were dripped into a biphasic bath with an upper oil layer, which 79 acted as an antisolvent, forcing the self-assembly of the agarose-based solutions, and a lower 80 water phase, in which the homogeneous spherical hydrogel capsules were collected. The 81 principle underlying this method was the supramolecular self-assembly of the probiotic-82 containing carbohydrate solutions when they were dispersed in a medium where they could 83 not solubilize, thus forming spherical hydrogels (Trivedi, Rao, & Kumar, 2014). 84 Advantageously, this new encapsulation method, which we have named “oil-induced biphasic 85 hydrogel particle formation”, is simple, fast, cytocompatible, it did not require the use of 86 organic solvents and all the materials used were food-grade. Moreover, it can be easily 87 scaled-up. In order to facilitate solution handling before hydrogel particle formation, initially, 88 different combinations of agarose with other biopolymers were tested and the encapsulation 89    5  structures were subsequently freeze-dried. The protection ability of the developed structures 90 was compared with directly freeze-dried solutions using the same formulations and with a 91 freeze-dried control using maltodextrin. The viability of probiotic bacteria was evaluated 92 during storage at two different temperatures (4 and 25 ºC) and the protection ability of the 93 best formulations was also tested under a simulated in-vitro gastrointestinal digestion process. 94 95 96 97 2. MATERIALS AND METHODS 98 2.1. Materials 99 Commercial agarose (type D2-LE) (Ag) was kindly donated by Hispanagar Inc. (Burgos, 100 Spain). Whey protein concentrate (WPC), under the commercial name of Lacprodan® DI-101 8090 and with a composition of ∼80 wt. % protein, ∼9 wt.% lactose and ∼8 wt.% lipids, was 102 obtained from ARLA (ARLA Food Ingredients, Viby, Denmark). Type A gelatin from 103 porcine skin (Gel) with reported gel strength of 175 g Bloom, maltodextrin (with dextrose 104 equivalent 16.5-19.5), alginic acid sodium salt from brown algae (medium viscosity) (Alg), 105 soybean oil, pepsin (porcine gastric mucosa, P-7000), pancreatin (porcine pancreas), bile 106 extract (porcine), phosphate buffered saline (PBS), 5-cCFDA (5-Carboxyfluorescein 107 diacetate), L-cysteine (97%), hydrochloric (HCL) acid and sodium bicarbonate (NaHCO3) 108 were supplied by Sigma-Aldrich (Barcelona, Spain). All products were used as received 109 without further purification. 110 111 2.2. Culture conditions of Bifidobacterium pseudocatenulatum – CECT 7765 112 Bifidobacterium pseudocatenulatum CECT 7765 was isolated from the stools of a breast-fed 113 infant, identified (Benítez-Páez et al., 2016) and deposited in the Spanish Cell Culture 114    6  Collection (CECT). The bacterial strain was routinely grown in Man, Rogosa and Sharpe 115 (MRS) broth (Scharlau, Barcelona, Spain) supplemented with cysteine (0.05% w/v) for 24 h 116 at 37°C in microaerophilic conditions (AneroGen; Oxoid, Basingstoke, UK). The bacterial 117 cells were then collected by centrifugation at 4000 rpm for 5 min using an Eppendorf 118 Centrifuge 5804R equipped with an Eppendorf Rotor S-4-72, obtaining a pellet that was 119 subsequently washed twice with PBS and cysteine (0.05% wt.) and re-suspended in the 120 biopolymeric dispersions. 121 122 123 2.3 Encapsulation 124 The encapsulation of probiotic bacteria was carried out by a novel technique named “oil-125 induced biphasic hydrogel particle formation”. This method consists on dripping the 126 probiotic-containing agarose-based solutions into a biphasic bath containing an upper oil 127 layer, acting as an antisolvent, and a lower water layer, for the collection of the hydrogel 128 particles. A scheme of the process is illustrated in Figure 1. More specifically, 0.5 wt.% 129 agarose stock solutions were prepared by dissolving 0.005 g mL-1 of the agarose powder in 130 PBS (130 mM sodium chloride, 10mM sodium phosphate, pH 7.4) and heating the solutions 131 at 90ºC under constant stirring until complete dissolution. Then, the agarose stock solution 132 was cooled down to 40 ºC and mixed with different selected biopolymer solutions. To this 133 end, WPC (5 wt. %), Gel (1 or 5 wt.%) or Alg (0.5 wt. %) solutions were prepared by 134 dispersing these hydrocolloid matrices into distilled water under magnetic stirring at 40 ºC 135 and blended with agarose solutions in different ratios (1:2, 1:1 and 2:1 agarose:biopolymer). 136 The ratio 1:2 agarose:biopolymer was found to be the most satisfactory as it prevented the 137 gelling of the blended solutions (data not shown). The final pH of the mixed solutions was 138 adjusted at 7.0 with 1M NaOH and, to avoid gelation, all the solutions were kept in a warm 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154   bath (a t solutio n probiot i (immer respect i of 0.2 M quick g compil e Fi g ure Table 1 S t 40ºC) unti n and gen t i c-containi n sed in an i i vely. In th e M CaCl 2 aq u g elation of t e d in Table 1. Scheme 1 . Nomencl a S ample N a l use. Subs e t ly stirred n g blend b i ce bath), c e case of u s u eous soluti he encapsu 1. of the prob a ture and c o a me A e quently, t h f or 10 mi n b iopolyme r c ontaining 2 s ing CaCl 2 on was use d l ating matr i i otic-conta i o mposition Ag arose / B  h e cell disp e n . Finally, solutions 2 00 mL oi l as a crossl i d instead o f i x. The dif f i ning sampl of the biop B lended e rsion was a capsules w into the b l /wate r (1:1 i nker of th e f water. Th e f erent form u es preparat i o lymer she l Volume a dded to th e w ere forme b iphasic o i ) as upper e agarose-A e ice bath w u lations tes t i on protoco l l l materials . Ratio e mixed bio p e d by drip p i l/aqueous and lower A lg capsule s w as used to t ed in this w l. . Total so  7 p olymer p ing the solution phases, , 10 mL favour a w ork are l ids    8  biopolymer concentration (% w/v) Agarose: blended biopolymer content (% w/v) Agarose:WPC5 0.5-5 1:2 3.5 Agarose:Gel1 0.5-1 1:2 0.8 Agarose:Gel5 0.5-5 1:2 3.5 Agarose:Alg0.5 0.5-0.5 1:2 0.5 Agarose:Alg0.5 (CaCl2*) 0.5-0.5 1:2 0.5 Control (Maltodextrin) - - 10 (*) 100 mL of 0.2M CaCl2 aqueous solution was used instead of water in the oil/water 155 bath. 156 157 158 2.3. Freeze-dried samples 159 Three different types of freeze-dried samples were prepared by using a Genesis 35-EL freeze-160 dryer (VirTis, USA). On one hand, the capsules prepared by the oil-induced biphasic 161 hydrogel particle formation methodology (as explained in Section 2.2) were subsequently 162 freeze-dried. Moreover, freeze-dried samples were also obtained by directly freeze-drying the 163 probiotic bacteria dispersed into the agarose-based mixtures. 164 And finally, for comparative purposes, control samples were prepared by dispersing the 165 twice-washed pellets in PBS solution containing 0.1 g mL-1 of maltodextrin, which were then 166 freeze-dried under the same conditions (López-Rubio, Sanchez, Wilkanowicz, Sanz, & 167 Lagaron, 2012). 168 169 170 2.4. Viability of Bifidobacterium pseudocatenulatum – CECT 7765 at different storage 171 temperatures 172    9  The viability of the probiotic bacteria in the different freeze-dried samples was evaluated 173 during storage at refrigerated conditions (4 ºC) and room temperature (25 ºC). Release of the 174 living cells was carried out by mechanically disrupting the microcapsules via high-speed 175 homogenization for 60s using a high-speed homogenizer (MICCRA D-9, Germany) equipped 176 with a VARIO (DS-20/PG SMIR) dispersing element (stator diameter of 20 mm). The 177 viability of the bacteria was determined by plate counting Samples (free cells and 178 encapsulated bacteria) were serially diluted in Ringer’s solution (pH 7.0) and plated on MRS 179 agar. After 48h incubation at 37°C, the cell counts were expressed in CFU g-1. Tests were 180 made in triplicate. 181 182 183 2.5. Survival of freeze-dried Bifidobacterium pseudocatenulatum CECT 7765 after 184 exposure to simulated gastrointestinal conditions. 185 The gastrointestinal digestion process was simulated as previously described (Laparra & 186 Sanz, 2010; Olivares, Laparra, & Sanz, 2011) using porcine pepsin (P7000, Sigma) (800-187 2500 units/mg protein), pancreatin (P1750, Sigma) (activity 4 x USP specifications), and bile 188 (B3883, Sigma). Briefly, the simulated gastric fluid was prepared with pepsin and the pH was 189 adjusted to 3 with HCl (0.1N). Similarly, simulated intestinal fluid was prepared with 190 pancreatin bile salt and the pH was adjusted to 6.8-7.0 with NaHCO3 (0.1N). The assay was 191 performed by adding 0.03g of freeze-dried material to 10 mL of saline solution (140 mM 192 NaCl, 5 mM KCl adjusted to pH=3). The mixture was immersed in a water bath (37 ºC) for 193 30 min with a gentle agitation. Once the samples were at room temperature, gastric and 194 intestinal digestions were conducted on a rocking platform shaker placed in an incubator (37 195 ºC, 5% CO2, 95% relative humidity). The viability of the bacteria was assessed after 60 min 196 of incubation in the simulated gastric fluid (SGF) and after 120 min incubation in the 197 335 336 337 338 339 340 341 342   Jobbeh d effecti v amino (simila r d ar, Yone k v e material s acid resid u r ly to a mil k A k ura, Parm e s for the pr u es that pr o k environm e e nter, & F o tection o f o vide a pr o e nt) (Rodri g  F isk, 2014) . f b ifidobact e o tective bu f g ues et al., 2 . Indeed, w e ria due to f fering env i 2 011). w hey prote i their high i ronment f o ins are co n content of o r probioti c  16 n sidered specific c strains 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357   Fi g ure freezed (freeze - In cont r capsul e the via b dried c concen t surviva shell m the m e refrige r b acteri a 3. Viabilit y d rying and ( - dried caps u r ast, this i m e s did not o c b ility was f c ounterpart s t ration of t l up to 69 aterial whi c e thodology r ated condi t a l survival B y loss unde r ( B) sample s u les). Valu e m provemen t c cur for th o f ound to be s (4 log t he formul a % (2 log c h contribu t used wer e t ions. Whe r when c o r controlled s obtained b e s are mean s t in the via b o se sample s similar fo r units red u a tions (up units reduc t ed to form e crucial i n eas the co m o mpared w  conditions b y encapsul s and stand a b ility of th e s prepared w r the freez e u ction). Int to 5 wt. % tion) proba b denser str u n the surv i m position c o w ith the m (4 ºC). (A) a tion and s u a rd deviati o e B ifidobac t w ith agaros e -dried cap s erestingly, % ) signifi c b ly d ue to t ctures. Bot h i val of th e o ntaining 1 m altodextri n Samples o b u bsequent fr o n of triplic a t erium cells e and Gel a s ules and t h an incre a c antly imp r t he higher s h the gelati n probiotic wt % gela t n control, b tained by d fr eeze-dryi n a tes. s in the fre e a t 1 wt.%, i h eir directl y a se in the r oved the b s olids conte n n concentr a bacteria s t t in did not increasing  17 d irect n g e ze-dried i n which y freezegelatin b acterial n t in the a tion and t ored at i mprove gelatin    18  concentration in the formulation resulted in improved bacterial viability. Moreover, 358 differences were also observed regarding the processing method, being encapsulation prior to 359 freeze-drying an excellent strategy to enhance cell viability during storage. 360 The enhanced viability observed when the cells were encapsulated via the oil-induced 361 biphasic hydrogel particle formation prior to freeze-drying could be ascribed to the formation 362 of a denser and more continuous matrix, which fully entrapped the bacteria in the coating 363 shell material and protected them from the oxygen or other detrimental conditions. In 364 contrast, the probiotic bacteria were more exposed to the adverse conditions when they were 365 obtained by directly freeze-drying the samples because of the more porous structures 366 obtained in this case, as it will be discussed below. As it has been previously reported, 367 various conditions affect the quality of the microcapsules, such as physico-chemical 368 characteristics of the capsule materials and the encapsulation methodology used (Prisco & 369 Mauriello, 2015), being the porosity of the wall material one of the key factors in the 370 protection of bacterial cells (Mu et al., 2018; Shaharuddin & Muhamad, 2015; Tonon, 371 Grosso, & Hubinger, 2011). 372 Addition of CaCl2 to the water phase in the bath was also evaluated for the agarose-based 373 formulations containing alginate, as this salt is known to act as a crosslinking agent for this 374 polysaccharide and, thus, it could contribute to the development of a more tightly packed 375 biopolymer network with expected improved barrier properties. However, no viability 376 improvement was observed in these samples, which could be due to changes in agarose-377 alginate interactions in the presence of the calcium ions, which deserve further study but were 378 out of the scope of this work. 379 380    19  3.3. Comparison of protective effect of agarose-based formulations with WPC or gelatin 381 during B. pseudocatenulatum CECT 7765 storage at ambient conditions and during 382 in-vitro digestion 383 As observed in the previous section, probiotic survival during storage depended on both the 384 matrix formulation used and the processing method employed, being encapsulation prior to 385 freeze-drying generally beneficial. In order to investigate if the differences in protective 386 ability were kept when the cells were exposed to different stress conditions (i.e. storage at 387 room temperature and during in-vitro digestion), the agarose-based formulation which was 388 more efficient in probiotic protection during storage in refrigerated conditions (0.5Ag-5WPC) 389 and the one which showed a greater viability loss (0.5Ag-5Gel) were selected for further 390 experiments. First, the viability of B. pseudocatenulatum within these two matrices upon 391 storage at ambient conditions (more cost-effective conditions for commercialization 392 purposes) was investigated. 393 The survival of freeze-dried probiotic bacteria stored at 25 ºC for 35 days is represented in 394 Figure 4. When compared with the viability loss during refrigerated storage (cf. section 3.2), 395 greater viability loss was observed in all cases, although the presence of WPC in the 396 formulation resulted in prolonged viability in all the freeze-dried samples. Just as it was 397 observed for samples stored at 4 ºC, the agarose-Gel freeze-dried samples were the least 398 effective in probiotic protection during storage at ambient conditions. Specifically, after 35 399 days at 25 ºC, the viability of B. pseudocatenulatum experienced a significant decrease 400 (p<0.05) of about 4-5 log units for all the agarose-Gel samples, being even significantly 401 lower than the control freeze-dried sample prepared with maltodextrin. These less favorable 402 results reported for agarose-Gel showed that microencapsulation is not always synonymous 403 of cell protection, and highlighted the relevance of an adequate material selection. More 404 interestingly, the bacterial survival for the freeze-dried samples obtained from agarose-WPC 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419   decrea s encaps u WPC f observ e general prepar e amino protein fact, m i techniq u Rodrig u Fi g ur e control s ed about 3 . u lation and f or probioti e d betwee n , gelatin m e d with wh e aci d resid u concentrat e i lk protein s u es (Anant a u es et al., 2 0 e 4. Viabilit y (freeze-dri e . 6 and 2.8 l subseque n c protectio n both hyd r atrices sho w e y protein u es of whe y e (i.e. lipid s s have been a , Volkert, 0 11). y loss duri n e d with ma l o g dependi n n t freeze-dr y n . One fa c r ocolloid m w e d higher ( Wang, A u y protein a s , lactose) m widely us e & Knorr, 2 0 n g storage u l todextrin) a  n g of the m y ing, respe c c tor that m i m atrices is oxygen p e u ty, & Ker r a nd the pre s m ight have a e d as prote c 005; Gardi n u nder ambi e a nd the fre e m ethodolog y c tively an d i ght have c related to e rmeability r y, 2010). F s ence of o t a lso contri b c tive agent s n er et al., 2 0 e nt conditio n e ze-dried sa m y used, dire c d confirmin g c ontributed the oxyge values tha n F urthermor e t her comp o b uted to pro t s using dif fe 0 02; López - n s (25 ºC/5 0 m ples obta i ct freeze-d r g the suita b to the di f e n permea b n their cou n e , b oth the o nents in t h o tect the ba c f erent enca p -Rubio et a 0 % RH) of i ned by bot h  20 r ying vs. b ility of f ferences b ility. In n terparts specific h e whey c teria. In p sulation l ., 2012; the h    21  methodologies (direct freeze-drying vs. encapsulation and subsequent freeze-drying). Values 420 are means and standard deviation of triplicates. 421 422 Therefore, the combination of agarose and WPC proved to be an excellent wall material, 423 which provided an effective physical barrier against adverse environmental conditions. 424 Conversely, the presence of gelatin in the formulations detrimentally affected the viability of 425 probiotics upon storage when compared with the maltodextrin control. Once again, the 426 encapsulation method here proposed and subsequent freeze-drying protected the viability of 427 probiotic bacteria more efficiently, probably due to the denser and more continuous matrix 428 obtained in this case. 429 In order to investigate the differences observed in the bacterial survival, the morphology of 430 the selected freeze-dried compositions (agarose-WPC and agarose-Gel, respectively) was 431 examined by SEM and compared with those obtained with maltodextrin. Representative 432 images of the freeze-dried samples produced by direct freeze-drying and by encapsulation 433 and subsequent freeze-drying are shown in Figure 5. In general, both maltodextrin and 434 agarose-based samples obtained through the direct freeze-drying methodology exhibited 435 nearly parallel, separated and porous sheets structure (see Figure 5A), although the 436 interconnection between the sheets was denser when agarose-based biopolymers were used as 437 protecting matrices. However, the morphology of the samples obtained by means of the 438 encapsulation and subsequent freeze-drying method exhibited a more continuous and denser 439 matrix. 440 By comparing the SEM images of agarose-WPC (Figure 5D) and agarose-Gel (Figure 5E) 441 biopolymeric matrices, the greater protection efficiency could be partly ascribed to the more 442 compact and continuous matrix (without cracks in the surface) obtained for the freeze-dried 443 agarose-WPC capsules, which provided a protective barrier for probiotics, decreasing their 444    22  damage by exposure to external agents (Tonon et al., 2011). In contrast, some cracks and 445 pinholes were distinguished in their counterparts containing gelatin, which could favor 446 oxygen diffusion through the capsules. Even though freeze-dried agarose-WPC capsules 447 presented a coarser surface than their counterparts prepared with gelatin, less porous regions 448 were observed in the surface of the microcapsules. 449 450 451 Thus, the SEM results further demonstrated that both the formulation of the starting solutions 452 and the processing method employed had an effect on the morphology of the probiotic-453 containing structures, which, at least partially, helped explaining the improved protection 454 attained in the agarose-WPC freeze-dried capsules. 455 456 457 458 459 460 461 462 463 464 465   Fi g ure obtaine freezed and (E) drying. materi a Given b iopol y 5. SEM i m d by direct d ried agaro s freeze-dri e Main scal e a ls(scales c o the differ e y mer matri c m ages of fre e freeze-dryi n s e-WPC ca p e d agaroseG e s correspo n o rrespond t o e nces in c e c es containi e ze-dried s a n g, (C) aga r p sules obta i G el capsule n d to 1 mm. o 100 μm). e ll viabilit y n g Gel or W  a mples. (A) r ose-gelati n i ned by enc s obtained b The insets y observed W PC, the s with malto d n obtained b apsulation a b y encapsu l show a det a during st o s urvival of f d extrin, (B ) b y direct fre a nd subseq u l ation and s u a il of the c r o rage b et w f reeze-drie d ) agaroseW e eze-drying , u ent freeze - u bsequent f r oss-sectio n w een agaro s d probiotic  23 W PC , (D) - drying f reezen s of the s e-based bacteria    24  was also evaluated under simulated GIT conditions, for the samples obtained using both 466 methodologies (direct freeze-drying vs. encapsulation and subsequent freeze-drying) as 467 compared to freeze-dried samples with maltodextrin (control). 468 Figure 6 shows the survival of the Bifidobacterium strain after exposure to simulated in-vitro 469 digestion conditions. After 90 min at pH 3.0 (simulating stomach environment) a reduction of 470 3 logs in the control samples (freeze-dried with maltodextrin) (p<0.05) was observed when 471 compared with the initial bacterial counts. As expected, these results demonstrated that the 472 exposure of freeze-dried samples with maltodextrin to acidic conditions and gastric enzymes 473 resulted in a decreased viability of probiotic bacteria. 474 When comparing the viability of freeze-dried probiotic bacteria in the two agarose-based 475 formulations, it was found that the samples containing WPC were more efficient in protecting 476 the cells in acidic conditions. As explained previously and also demonstrated by previous 477 works, WPC is able to better protect the bacteria by creating a microenvironment within the 478 hydrocolloid matrix surrounding the bacteria, thus isolating the probiotic cells from the 479 stresses of the external low pH environment (Ashwar, Gani, Gani, Shah, & Masoodi, 2018; 480 Champagne, Reid, Gardner, Fustier, & Vuillemard, 2006; Dianawati, Mishra, & Shah, 2013), 481 fact that was not observed for gelatin. These differences can be ascribed to the presence of 482 lipids and lactose in the WPC as well as the specific aminoacid residues of whey protein, 483 which may contribute to protect the bacteria. In fact, the survival of probiotics in the freeze-484 dried agarose-Gel matrices was reduced by approximately 2 log CFU g-1 when they were 485 prepared by the oil-induced biphasic hydrogel particle formation approach and subsequent 486 freeze-drying method (freeze-dried capsules) and 3 log CFU g-1 reduction when they were 487 obtained by the direct freeze-drying process. In contrast, the viability of probiotic bacteria 488 only decreased about 1 log CFU g-1 in agarose-WPC matrices regardless of the methodology 489 used. 490    25  After incubation in SIF for 210 min, the number of bacterial counts remained constant in 491 freeze-dried agarose-WPC matrices obtained by the encapsulation and subsequent freeze-492 drying method. Comparing both procedures after exposure to the SIF, the bacterial counts 493 were found to be about 5.4 and 6.6 log CFU g-1 in agarose-WPC samples obtained by the 494 direct freeze-drying process, and by encapsulation and subsequent freeze-drying method, 495 respectively. In contrast, the recovery of probiotic bacteria from the agarose-Gel capsules was 496 about 4 log CFU g-1 at the end of the simulated in-vitro digestion process, which was in the 497 same order as those obtained for the freeze-dried samples with maltodextrin. 498 It is interesting to highlight that the bacterial survival from freeze-dried samples with 499 maltodextrin and freeze-dried samples in agarose-Gel matrices decreased more than 3 logs 500 after the in-vitro digestion process, and this reduction was significantly greater than that 501 observed for their counterparts prepared with agarose-WPC matrices (1 log reduction after 502 the in-vitro gastrointestinal assays). In this regard, the differences observed by the 503 methodology used could be related with the microstructure of the freeze-dried samples (see 504 Figure 5). In fact, freeze-dried samples obtained in one-step (direct freeze-drying process) 505 reduced the cultivable probiotic bacteria to a greater extent than their counterpart freeze-dried 506 capsules (oil-induced biphasic hydrogel particle formation and subsequent freeze-drying 507 method). It is worth mentioning the potential of the developed freeze-dried agarose-WPC 508 capsules to develop probiotic functional foods, since the amount of B. pseudocatenulatum 509 survival after the simulated digestion process was greater than 6 log CFU g-1, in accordance 510 with the requirements for the probiotics to exert their beneficial effects (FAO/WHO, 2001). 511 Therefore, the entrapment of B. pseudocatenulatum bacteria in agarose-WPC matrices by 512 encapsulation and subsequent freeze-drying method is a promising approach for the 513 preparation of probiotic functional foods, since it provided an efficient physical barrier 514    32  Simoes, L.S., Madalena, D.A., Pinheiro, A.C., Teixeira, J.A., Vicente, A.A., & Ramos, O.L. 646 (2017). Microand nano bio-based delivery systems for food applications: In vitro behavior. 647 Advances in Colloid and Interface Science, 243, 23-45. 648 Smidsrod, O., & Skjak-Braek, G. (1990). Alginate for cell immobilization. Trends in Food 649 Science and Technology 8, 71–75. 650 Sohail, A., Turner, M.S., Coombes, A., & Bhandari, B. (2013). The Viability of Lactobacillus 651 rhamnosus GG and Lactobacillus acidophilus NCFM Following Double Encapsulation in 652 Alginate and Maltodextrin. Food and Bioprocess Technology, 6 (10), 2763-2769. 653 Soukoulis, C., Behboudi-Jobbehdar, S., Macnaughtan, W., Parmenter, C., & Fisk, I.D. 654 (2017). Stability of Lactobacillus rhamnosus GG incorporated in edible films: Impact of 655 anionic biopolymers and whey protein concentrate. Food Hydrocolloids 70, 345-355 656 Soukoulis, C., Behboudi-Jobbehdar, S., Yonekura, L., Parmenter, C., & Fisk, I. (2014). 657 Impact of milk protein type on the viability and storage stability of microencapsulated 658 Lactobacillus acidophilus NCIMB 701748 using spray drying. Food and Bioprocess 659 Technology 7 (5), 1255-1268. 660 Sousa, S., Gomes, A.M., Pintado, M.M., Silva, J.P., Costa, P., Amaral, M.H., Duarte, A.C., 661 Rodrigues, D., Rocha-Santos, T.A.P., & Freitas, A.C. (2015). Characterization of freezing 662 effect upon stability of, probiotic loaded, calcium-alginate microparticles. Food and 663 Bioproducts Processing, 93, 90-97. 664 Tonon, R.V., Grosso, C.R.F., & Hubinger, M.D. (2011). Influence of emulsion composition 665 and inlet air temperature on the microencapsulation of flaxseed oil by spray drying. Food 666 Research International, 44 (1), 282-289. 667 Tonucci, L., Dos Santos, K.M.O., Ferreira, C.L.D.F., Ribeiro, S.M.R., De Oliveira, L.L., & 668 Martino, H.S.D. (2017). Gut microbiota and probiotics: focus on diabetes mellitus. Critical 669 Reviews in Food Science and Nutrition, 57, 2296-2309. 670    33  Trivedi, T.J., Rao, K.S., & Kumar, A. (2014). Facile preparation of agarose-chitosan hybrid 671 materials and nanocomposite ionogels using an ionic liquid via dissolution, regeneration and 672 sol-gel transition. Green Chemistry, 16 (1), 320-330. 673 Wang, L., Auty, M.A.E., & Kerry, J.P. (2010). Physical assessment of composite 674 biodegradable films manufactured using whey protein isolate, gelatin and sodium alginate. 675 Journal of Food Engineering, 96 (2), 199-207. 676 Yeung, T.W., Arroyo-Maya, I.J., McClements, D.J., & Sela, D.A. (2016). 677 Microencapsulation of probiotics in hydrogel particles: enhancing Lactococcus lactis subsp. 678 cremoris LM0230 viability using calcium alginate beads. Food & Function, 7, 1797-1804. 679 Yeung, T.W., Ucok, E.F., Tiani, K.A., McClements, D.J., & Sela, D.A. (2016). 680 Microencapsulation in alginate and chitosan microgels to enhance viability of 681 Bifidobacterium longum for oral delivery. Frontiers in Microbiology 7, 494. 682 Ying, D.Y., Phoon, M.C., Sanguansri, L., Weerakkody, R., Burgar, I., & Augustin, M.A. 683 (2010). Microencapsulated Lactobacillus rhamnosus GG powders: relationship of powder 684 physical properties to probiotic survival during storage. Journal of Food Science, 25(9), 685 E588-E595. 686 Yokoyama, E., Kishida, K., Uchimura, M., & Ichinole, S. (2006). Comparison between 687 agarose gel electrophoresis and capillary electrophoresis for variable numbers of tandem 688 repeat typing of Mycobacterium tuberculosis. Journal of Microbiological Methods, 65, 425-689 431. 690 Zhao, M., Wang, Y., Huang, X., Gaenzle, M., Wu, Z., Nishinari, K., Yang, N., & Fang, Y. 691 (2018). Ambient storage of microencapsulated: Lactobacillus plantarum ST-III by complex 692 coacervation of type-A gelatin and gum Arabic. Food and Function, 9 (2), 1000-1008. 693 694