Removal of raffinose and stachyose from soy-based beverages using soluble and immobilized MelA α-D-galactosidase from Lactiplantibacillus plantarum WCFS1

Removal of raffinose and stachyose from soy-based beverages using soluble and immobilized MelA α -D-galactosidase from Lactiplantibacillus plantarum WCFS1 Leandro Alves dos Santos a,d , Paloma Delgado-Fern´ andez b , Ana Mu˜ noz-Labrador b , Diego Martin-Gutierrez a , Evelyn C. Romero b , Franzory Cuaran a , Andrea Garcia-Alvarez a , Elena García-Calvo a , Blanca de las Rivas c , Rosario Mu˜ noz c , Nieves Corzo b , F. Javier Moreno b , Cesar Mateo a,* a Departamento de Bioc´ atalisis, Instituto de Cat´ alisis y Petroleoquímica, CSIC, Marie Curie 2, 28049, Madrid, Spain b Departamento de Bioactividad y An´ alisis de Alimentos, Instituto de Investigaci´ on en Ciencias de la, Alimentaci´ on, CIAL (CSIC-UAM), Nicol´ as Cabrera 9, 28049, Madrid, Spain c Instituto de Ciencia y Tecnología de Alimentos y Nutrici´ on, ICTAN (CSIC), Juan de la Cierva 3, 28006, Madrid, Spain d Department of Chemistry, Federal University of Paran´ a, Polytechnic Center, P.O. Box 19032, Curitiba, 81531-980, Paran´ a, Brazil ARTICLE INFO Keywords: Soy-based beverages α -galactosidase Galactose hydrolysis Docking Enzyme immobilization ABSTRACT The hydrolysis of α -galactose-carbohydrates in soy-based beverages using the enzyme MelA α -D-galactosidase (Lp_3485) from Lactiplantibacillus plantarum WCFS1 is presented. MelA is a highly unstable trimeric enzyme whose activity and stability are strongly dependent on the dilution and polarity of the reaction medium. Considering this, the enzyme was immobilized and stabilized using different immobilization protocols. Postimmobilization techniques consisting of cross-linking with different polyfunctional polymers were used. The optimal derivative was used to hydrolyse different oligosaccharides of the raffinose family (raffinose and stachyose) present in soy drink preparations. Remarkably, MelA showed specific activity and high hydrolytic efficiency towards RFOs in soy-based beverages at a relatively low temperature (20 ◦C). Therefore, these results suggest that this enzyme could be preferentially used in the early stage of soybean processing to remove RFOs from soy-based beverages and soy-derived products, thereby improving their nutritional value. 1. Introduction Soybean (Glycine max (L) Merr) is a highly valuable food source, providing essential proteins, fibre, micronutrients and phytochemicals that play an important role in the diets of different regions of the world (Fehily, 2003; Guo et al., 2022). Soybeans are also notable for their carbohydrate composition, which accounts for a significant proportion (30%) of their total content. Within these carbohydrates, soluble sugars make up about 15% of the seed composition. The predominant sugars found in soybean seeds are sucrose, which can range from 40% to 70% of the sugar content, followed by raffinose with 5%–15% and stachyose with 12%–35% (Hou et al., 2009). Soybeans can be used to make various products such as bread, biscuits, cakes, chocolate, etc (Amigo-Benavent, Silv´ an, Moreno, Villamiel, & delCastillo, 2008). Soy products can also be used as a meat substitute in vegetarian diets due to their high protein profile and balanced amino acid pattern (Bhatia, Singh, Batra, & Singh, 2020). In particular, soy-based beverages stand out as a nutritious beverage and a viable alternative to mammalian milk. In addition, soy drinks are lactose-free, making them an ideal choice for people with lactose intolerance (Katrolia, Liu, Li, & Kopparapu, 2019). However, soy-based beverages contain antinutritional factors including trypsin inhibitors, phytic acid, lectins and soluble but indigestible carbohydrates. Among the oligosaccharides of the raffinose family (RFOs) such as raffinose, stachyose, verbascose and ajugose are considered major antinutrients (Padalkar et al., 2023). Structurally, these carbohydrates are α -galactosyl ( α -1,6) derivatives of sucrose, with raffinose, stachyose, verbascose and ajugose containing one, two, three and four galactose units, respectively, linked to sucrose. These α -galactosides are unable to be digested by humans and other monogastric * Corresponding author. E-mail address: [email protected] (C. Mateo). Contents lists available at ScienceDirect LWT journal homepage: www.elsevier.com/locate/lwt https://doi.org/10.1016/j.lwt.2024.116864 Received 1 April 2024; Received in revised form 1 August 2024; Accepted 2 October 2024 LWT - Food Science and Technology 210 (2024) 116864 Available online 3 October 2024 0023-6438/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). animals due to the absence of pancreatic α -D-galactosidase, which is required to hydrolyse the α -1-6 linkages. As a result, RFOs remain undigested in the small intestine and pass through without being absorbed. Subsequently, in the large intestine, gas-producing bacteria such as Clostridium sp. metabolize the RFOs via anaerobic fermentation, producing excess carbon dioxide, hydrogen, traces of short-chain fatty acids (SCFAs) and methane. These flatus accumulation lead to symptoms such as intestinal discomfort, flatulence and diarrhoea in susceptible individuals, which limits the consumption of foods with high concentrations of RFOs (Connes et al., 2004; Mutuyemungu, Singh, Liu, & Rose, 2023; Sanyal, Kumar, Pattanayak, Kar, & Bishi, 2023). Therefore, addressing the removal of RFOs from soy drinks is key to potentially improving their nutritional value and increasing their consumption. Current industrial approaches for removing RFOs from soy-based beverages include the use of enzymes, specifically α -D-galactosidases. These enzymes are widely used in many industrial processes of soyderived products (Falkoski et al., 2006; Çelem & ¨ Onal, 2022). α -D-galactosidases ( α -D-galactoside galactohydrolase, EC 3.2.1.22) are exoglucosidases that catalyse the hydrolysis of α -1,6-linked terminal non-reducing α -D-galactosyl residues from RFOs such as raffinose, stachyose and verbascose, as well as glycoconjugates, glycoproteins, galactomanans and galactolipids using retention mechanisms (Jang et al., 2019; Li, Loman, Coffman, & Ju, 2017; Wang et al., 2014). In addition, α -D-galactosidase cleaves the α -1,6 linkage between galactosyl and glucosyl units from melibiose, hence it is also known as melibiase (Bhatia et al., 2020; Delgado-Fernandez et al., 2020). α -D-galactosidases are widely distributed in plants, animals and microorganisms (fungi and bacteria) and due to the increasing industrial demand for enzymes to remove RFOs, many α -D-galactosidases have been studied (Bhatia et al., 2020; Wang et al., 2019; Çelem & ¨ Onal, 2022). The growing demands of industrial processes and the need to simplify downstream processing and operational stability, have rendered immobilized enzymes highly desirable. More specifically, the use of soluble α -D-galactosidases as catalysts for industrial-scale reactions is not feasible because it is not possible to recover the enzyme at the end of the process. Therefore, enzyme immobilization has emerged as an attractive topic to enable enzyme reuse and continuous operation (Geng et al., 2022; Maghraby, El-Shabasy, Ibrahim, & Azzazy, 2023). Various methods of enzyme immobilization have been described, but their effectiveness varies depending on the nature of the enzymes, particularly their tertiary and quaternary structures in the case of multimeric enzymes (Mateo et al., 2020). In the case of multimeric enzymes, stability depends on the immobilization of all subunits to prevent their dissociation in the reaction medium. Therefore, there is a growing tendency to develop immobilization methods that involve all subunits. However, geometric constraints (e.g, tetrahedral enzymes or enzymes with many subunits) often make this approach infeasible. Therefore, post-immobilization techniques using polymers or polyfunctional compounds capable of cross-linking the non-immobilized subunits to the support become necessary (Fernandez-Lafuente, 2009; Fern´ andez-Lafuente et al., 1999). α -D-galactosidases from different sources have been immobilized on different supports and successfully used for the removal of RFOs from soy drinks (Katrolia et al., 2019). Recently, an efficient hydrolysis (84%) of RFOS (raffinose and stachyose) was achieved using MelA α -D-galactosidase from Lactiplantibacillus plantarum WCFS1, demonstrating that MelA has a high capacity to hydrolyse indigestible RFOS in simple buffered solutions (Delgado-Fernandez et al., 2020). In this work, once the enzyme activity of MelA α -D-galactosidase was optimized, different immobilization supports were prepared to unravel the efficiency of the enzyme immobilization prior to their application. This process was followed by a chemical characterization in order to depict the resulting oligosaccharides after hydrolyzation. Therefore, the main objective of this work aims to investigate the ability of immobilized MelA α -D-galactosidase (Lp_3485) from L. plantarum WCFS1 in comparison with the soluble form, to hydrolyse carbohydrates as RFOs present in real food matrices as a technological strategy for soy-based beverages. 2. Materials and methods In all cases, experiments were performed at least in duplicate and the mean error was never great than 5%. 2.1. MelA α -D-galactosidase activity test The enzyme α -D-galactosidase (MeIA) from Lactobacillus plantarum WCFS1 was purified and characterized as described (Curiel, de las Rivas, Manche˜ no, & Mu˜ noz, 2011). The enzymatic activity was evaluated by measuring the p-nitrophenolate (pNP) released from a solution of p-nitrophenylα -D-galactopyranoside (pNPG). The measurement was performed continuously by kinetics using a JASCO V-730 spectrophotometer. A 5 mM solution of pNPG in 50 mM sodium phosphate at pH 7 was used. The released p-NP was measured at 420 nm and 25 ◦C. One enzymatic unit is defined as the amount of MeIA required to produce 1 μ mol of p-NP per minute under the assay conditions ( ε =9310 M −1 cm −1 ). 2.2. Preparation of the different agarose-based supports The different supports used for the immobilization of the enzyme were derived from agarose previously derivatized with epoxide supports. These epoxide groups were subsequently converted into different reactive groups as described below. 2.2.1. Activation of agarose with epoxide groups 1 g of agarose (4BCL; ABT, Spain) was suspended in 4.4 mL of water and mixed with 0.328 g of NaOH, 0.02 g of NaBH 4 , 1.6 mL of acetone and finally 1.1 mL of epichlorohydrin. This mixture was stirred at 25 ◦C for 16 h. Finally, the support was washed with plenty of water and dried by vacuum filtration. 2.2.2. Activation of epoxy agarose with other functional groups Agarose-CHO (glyoxyl-agarose): 1 g of epoxy agarose was hydrolyzed with 10 mL of 0.5 M aqueous HCl solution for 90 min. After this time, the supports were washed with water and dried. A solution consisting of 0.1 mL of 0.2 M NaIO 4 and 9 mL of water was then added. This suspension was allowed to react for 90 min at 25 ◦C, then washed with water and dried. Triethylamine agarose (TEA): 1 g of agarose epoxide was mixed with 10 mL of a solution consisting of 5 mL acetone, 1.4 mL TEA and 3.6 mL water. This suspension was stirred at 25 ◦C for at least 24 h. Finally, it was washed with water and dried under vacuum. Sodium iminodiacetate (IDA). To 1 g of agarose epoxide, 10 mL of a 0.5 M IDA solution at pH 11 was added. This suspension was allowed to react for at least 24 h with stirring, after which the support was washed and dried. Agarose-IDA/Zn. To 1 g of agarose-IDA support, 10 mL of an aqueous solution of ZnCl 2 (0.2 M) was added. After 30 min of reaction, the support was washed with water and dried for later use. Activation of the different bifunctional supports used: In all cases, the derivatized agarose epoxide supports (IDA, TEA, Zn) were oxidized according to the following protocol: 0.1 mL of a 0.1 M aqueous NaIO 4 solution and 9 mL of water were added to 1 g of each support and allowed to react for 90 min. After this time, they were washed with water and dried under vacuum. Agarose-ethylenediamine (EDA) with glutaraldehyde: 1 g of agaroseCHO was aminated by reaction with 10 mL of a 2 M aqueous solution of ethylenediamine at pH 10 for 2 h. After this reaction time, 10 mg of NaBH 4 was added. After another 2 h, the support was washed with 1 M NaCl and plenty of water. Finally, the support was suspended in a mixture consisting of 1.7 mL of 0.2 M phosphate buffer at pH 7 and 1.1 mL of commercial 25% glutaraldehyde solution. The suspension was L. Alves dos Santos et al. LWT 210 (2024) 116864 2 allowed to stand for 16 h at 25 ◦C with gentle agitation and then washed with plenty of water (Betancor et al., 2006). The other reagents used in the development of the immobilization procedure were purchased from Merck (Stenheim, Germany). 2.3. Immobilization of MeIA on different supports Different buffers were used depending on the support used for immobilization: 5 mM sodium phosphate and 50% glycerol for the agarose-TEA and agarose-IDA activated supports, and 25 mM sodium phosphate with 150 mM NaCl and 15% glycerol for the other supports. Then, 1.0 g of the different supports were suspended in 10 mL of MelA enzyme solution (100 U g-1 support) and gently shaken at 4 ◦C for different times. Immobilization was considered complete when no activity was detected in the supernatant. The immobilization efficiency (IE, %) was calculated as (Equation (1)): IE =Ai−Af Ai ×100% (1) where Ai is the initial enzymatic activity (U) of the enzyme and Af is the enzymatic activity (U) remaining in the supernatant at the end of the immobilization process. The recovered activity (R, %) was calculated as (Equation (2)): R=A0 AT ×100% (2) where A 0 is the quotient between the real activity (U g-1 support) of the immobilized preparation and AT is the theoretical activity (U g-1 support) of the immobilized preparation. At the end of the immobilization process, the enzyme immobilized on supports with aldehyde groups was reduced with 1 mg mL-1 NaBH4 for 30 min at 4 ◦C and then washed with the immobilization buffer previously cooled to 4 ◦C. In the case of the glutaraldehyde support, it was treated with 5 mL of 1M glycine at pH 7 and 4 ◦C for 1 h to block the reactive groups and then washed with the immobilization buffer. 2.4. Post-immobilization treatment 1 g of the preparation immobilized on glutaraldehyde-activated supports was suspended in 5 mL of pH 5.7 buffer containing 15% glycerol and 5 mg/mL of the different polymers (aspartic dextran and glycine dextran) in the presence of 10 mM N-(3-dimethylaminopropyl)- N ′ -ethylcarbodiimide hydrochloride (EDC). The suspension was allowed to stand for 90 min at 4 ◦C with gentle shaking and then washed with 25 mM phosphate buffer, pH 7, containing 150 mM NaCl and 15% glycerol. 2.5. Soydrink based preparation (sample L) The beverage was prepared as was described by (Mulimani and Ramalingam, 1995) using commercially available soybeans. The soybeans were ground and defatted with n-hexane (1:1; v/v). This defatted meal was suspended in 10 vol of distilled water and boiled for 10 min. Insoluble residues were removed by centrifugation at 2000×g for 5 min. 2.6. Treatment of soy-based beverages MelA α -D galactosidase to remove RFOs Eight brands of soy-based beverages, labelled as SA, SL, P, VS, SCa, A, K, and SSA, were purchased from local supermarkets whose selection criteria included their market representation, product diversity and accessibility. Table 1 shows the carbohydrate composition of the different soy-based beverages and the lab-prepared soy-milk. All samples were stored at refrigeration temperature and analyzed in duplicate before the expiry date. Hydrolysis of RFOS present in commercial soy-based beverages (Table S1) and the laboratory prepared soy drink sample were performed in batch reactions using soluble and immobilized enzyme in batch reactions. Briefly, 5 U of free or immobilized MelA α -D-galactosidase were added to 0.460 mL aliquots of soy-based beverages and incubated at 22 ◦C and pH 6.8 for 7 h under agitation. Aliquots were withdrawn periodically 0, 1, 3, 5 and 7 h and kept in a boiling water bath for 10 min to stop the enzyme reaction. 2.7. Determination of pH and dry matter (DM) The pH of the samples was measured at 22 ◦C using an MP225 pH meter with a glass electrode (Mettler Toledo GmbH, Schwerzenbach, Switzerland). The dry matter (DM) was determined in the soy-based beverage previously prepared in the laboratory (L) as described in section 2.5. The samples were heated for 15 h and the dry residue was weighed after cooling in a desiccator. The DM value (%) was about 6.8 for the soy beverages. 2.8. Determination of carbohydrates by gas chromatographyflame ionization detector (GC-FID) Carbohydrates were extracted according to the method of Patil, Praveen Kumar, Mulimani, Veeranagouda, & Lee, 2010 (Patil et al., 2010) with some modifications. First, an optimization of the extraction process of commercial soy-based beverages was performed using two different alcoholic solvents: ethanol and methanol. Regarding the quantitative results of carbohydrates and their coefficient of variation, methanol extraction provided more accurate results for the Nutrition Facts panel of product labels (Table S1) and coefficients of variation <10% compared to ethanol. Once the extraction solvent was optimized, aliquots (3 mL) of soy-based beverage samples were mixed with 7 mL of methanol. The mixture was centrifuged at 3200 g for 15 min at 40 ◦C. The different carbohydrates were analyzed as trimethylsilylated oxime (TMSO) derivatives, prepared as previously described (Garcia Banos, Olano, & Corzo, 2000). 0.4 mL of internal standard (IS) solution (0.5 mg mL-1 phenyl-β-glucoside) was added to different samples containing between 2.5 and 5 mg of sugars. These mixtures were dried in a rotary evaporator at 38–40 ◦C. Sugar oximes were formed by adding 250 μ L of hydroxylamine chloride (2.5%) in pyridine and heating the mixture at 70 ◦C for 30 min, then silylated with hexamethyldisilazane (250 μ L) and TFA (25 μ L) and kept at 50 ◦C for 30 min. Reaction mixtures were Table 1 – Composition in sucrose and raffinooligosaccharides (RFOs; raffinose and stachyose) found in commercial UHT soy-based beverages (n =2). Soy-based beverages labelling Sugar content (mg/100 mL) Sucrose Raffinose Stachyose Total RFOs SA 120.18 ± 6.71 2.28 ±0.27 11.30 ± 0.184 13.58 SL 82.13 ± 4.26 3.17 ±031 14.11 ±1.71 17.28 P 392 ± 36.33 90.2 ±1.64 373 ±8.51 463.2 VS 478.18 ± 21.01 66.22 ± 3.95 383.01 ± 35.22 449.23 SCa 362.76 ± 1.87 10.64 ± 0.14 13.72 ±0.44 23.36 A 648.5 ± 6.37 119 .14 287.2 ±0.71 406.34 K 578.57 ± 67.27 173.47 ± 21.56 375.40 ± 56.64 548.87 L 340.9 ± 32.37 103.94 ± 10.20 436.44 ± 45.42 540.38 SSA (Without added sugars) 19.23 ± 1.86 2.51 ±0.21 10.71 ±0.67 13.22 L. Alves dos Santos et al. LWT 210 (2024) 116864 3 centrifuged at 6708 g for 1 min at room temperature. Supernatants were injected or stored at 4 ◦C prior to analysis. GC-FID analysis of TMSO derivatives was performed on an Agilent Technologies 7820A gas chromatograph (Wilmington, DE, USA) using a DB-5HT fused silica capillary column, cross-linked and bonded phase (5% phenyl-methylpolysiloxane; 30 m ×0.25 mm i.d., 0.10 μ m film thickness) (J&W Scientific, Folson, CA, USA). The initial oven temperature was 120 ◦C, which was increased to 380 ◦C at a rate of 1 ◦C/min and held for 80 min. Injector and detector temperatures were 280 ◦C and 370 ◦C, respectively. Injections were performed in split mode (1:20 or 1:5 in the case of digested samples) using nitrogen at a flow rate of 1 mL/min. Quantification of each carbohydrate was performed by calibration using phenyl-β-glucoside (0.5 mg mL-1) as an internal standard. Mixtures of standard solutions of glucose, galactose, fructose, melibiose, raffinose, and stachyose (Sigma Aldrich; Stenheim, Germany) were prepared in the expected concentration range to calculate response factors for each standard carbohydrate. All analyses were performed in duplicate. Data acquisition was evaluated using Agilent ChemStation software (Wilmington, DE, USA). 3. Results and discussion 3.1. Carbohydrate composition of the soy-based beverages GC-FID analysis was carried out in order to characterize either the commercial soy-based samples and the lab-prepared soy-based sample studied in this manuscript. Moreover, this analytical methodology was used as a screening of the reaction products over time in order to depict the enzyme activity efficiency and recovery of either the soluble enzyme form or within their immobilization supports. Fig. 1 shows the chromatograms of two commercial soy-based beverages, samples K and SSA, with different carbohydrate contents. As can be seen, the quantifiable carbohydrates were sucrose, raffinose and stachyose. Sucrose was the main carbohydrate found in sample K, Fig. 1. GC-FID chromatograms of two commercial soy-based beverages, a) sample K, b) sample SSA (without added sugar). L. Alves dos Santos et al. LWT 210 (2024) 116864 4 followed by stachyose and raffinose. In general, this pattern was found in all analyzed soy beverages (Table 1). Similarly, high levels of sucrose were found in A, VS, P and SCa, whereas it was very low in sample SSA (Fig. 1) in accordance to their commercial label (Table S1). The latter was labelled as having no added sugars, so this soy-based drink may reflect the proportion of carbohydrates originally present in soybeans. Considering RFOs, stachyose was the dominant carbohydrate in all samples. and the total amount of RFOs ranged from 13.22 to 548.87 mg/ 100 mL. Rada-Mendoza and co-workers (Rada-Mendoza, Villamiel, Ramírez, Usuriaga, & Montilla, 2022) analyzed three samples of soy-based beverages and the range of RFOs was within the values reported in our work (i.e., between 30 and 274 mg/100 mL). While some of the soy-based beverages samples exhibit relatively low RFO values (samples SA, SL, SCa and SSA; 13.22–23.36 mg/100 mL), the samples namely: P, VS, A, K and L, still show high quantitative values of 406.34–548.87 mg/100 mL according to the carbohydrate composition in Table 1. The RFO concentrations for the latest samples may lead to flatulence and intestinal issues since humans and monogastric animals lack the enzyme responsible for their hydrolysis in the small intestine (Sanyal et al., 2023). Therefore, enzymatically removing these oligosaccharides is key to allowing the elaboration of soy-based products with enhanced nutritional value while reducing their antinutritional and flatulence-causing compounds (Elango et al., 2022). 3.2. Study of soluble MelA α -D-galactosidase Since the purpose of this study was to suggest an enzyme hydrolyzation of antinutritional factors of soy-based beverages such as α -galactosides, the study of soluble MelA enzyme was carried out. The α -Dgalactosidase enzyme from L. plantarum WCFS1 is an enzyme with a trimeric structure. It is composed of 80 kDa monomers with a total mass of 240.5 kDa (Panwar, Shubhashini, Chaudhari, Prashanth, & Kapoor, 2020). This enzyme gave a hydrolytic activity of 484 U/mg when the synthetic substrate o-NPG was used. However, the enzyme activity was highly variable depending on the reaction medium in which it was used. Therefore, the enzyme activity was measured after dilution in media with different buffers and in the presence of different additives, as shown in Table 2. This variability in catalytic activity is characterized by the multimeric nature of the enzyme, with loss of catalytic activity in media of lower polarity. The presence of an additive such as glycerol had a very positive effect on the stability of the enzyme structure. The optimal conditions, both for the production of the enzyme and for its subsequent use for immobilization, were dilution in phosphate buffer with a concentration ranging from 25 to 150 mM and 15% glycerol. 3.3. Immobilization of MelA α -D-galactosidase The utilization of enzyme immobilization onto different supports serves the purpose of stabilizing the enzyme while preserving its enzymatic activity. Activated agarose supports with different functional groups, monoand hetero-functional, were used for immobilization of the enzyme. Table 3 provides the results of the metrics calculated regarding the IE (%) - Eq. (1) and the R (%) - Eq. (2). Immobilization is achieved by the reaction of the support with different reactive groups on the amino acids of the enzyme and therefore by different regions of the surface where these amino acids are more abundant. The use of different supports can produce catalysts with different properties from those of the soluble enzyme in terms of activity, stability and selectivity (Mateo, Palomo, Fernandez-Lorente, Guisan, & Fernandez-Lafuente, 2007). The best supports for immobilization were those activated with groups capable of immobilizing the enzyme by anion exchange mechanisms (TEA-CHO and EDA-glutaraldehyde) and those activated with metal groups. In the case of metals, this is normal considering that the enzyme is labelled with 6 histidine residues per subunit. The activities obtained retained almost unchanged the initial activity offered, except in the case of supports activated with Cu 2+ , where half of the activity offered was lost after the immobilization process. The high activity recovered in the first stage of the immobilization indicates that there would be no mass transfer limitations. This is consistent with not having catalysts with the maximum possible support load. The stability of the most active derivatives was also studied in Table 2 – Activity of the soluble MelA in different buffers. Dilution (1:5). 100 % activity was considered as the undiluted enzyme (n =3). Composition of the buffer Relative activity (%) Salt Concentration (mM) Salt Concentration (mM) Additive % – Phosphate 50 NaCl 20 – – 1.27 Phosphate 50 NaCl 75 – – 19.3 Phosphate 50 NaCl 150 – – 50 Phosphate 50 NaCl 300 – – 53 Phosphate 50 NaCl 500 – – 40 Phosphate 50 NaCl 150 Glicerol 5 % 84 Phosphate 50 NaCl 150 Glicerol 10 % 85 Phosphate 50 NaCl 150 Glicerol 15 % 98 Phosphate 50 NaCl 20 Glicerol 15 % 95 Phosphate 25 – – Glicerol 15 % 95 Phosphate 25 – – Glicerol 15 % 68 Phosphate 25 – – Glicerol 25 % 68 Phosphate 25 – – Glicerol 50 % 78 Phosphate 5 – – PEG 15 % 2 Phosphate 5 – – PEG 25 % 5 Phosphate 5 – – PEG 50 % 50 MOPS 50 NaCl 20 – – 30 MOPS 50 NaCl 75 – – 56 MOPS 50 NaCl 150 – – 75 MOPS 50 NaCl 300 – – 78 Table 3 Immobilization of MelA α -D-galactosidase from L. plantarum WCFS1 on different agarose activated supports. The offered activity was 75 U per gram of support (n =3). Support Immobilized enzyme (%) Recovered activity (%) TEA 100 100 IDA 10 38 Glyoxyl 10 0 IDA (Cu) 80 49 IDA (Zn) 100 100 IDA – glutaraldehyde 0 0 EDA – glutaraldehyde 100 100 IDA (Zn) – glutaraldehyde 100 100 L. Alves dos Santos et al. LWT 210 (2024) 116864 5 comparison with that of the enzyme in soluble form. The different preparations were incubated at 29 ◦C. Under these conditions, the halflife of the soluble enzyme was less than 30 min (Fig. 2). The most stable derivatives were those immobilized on supports activated with EDAglutaraldehyde. However, the low stabilization factors (around 2-fold increase) suggest that the quaternary structure of the enzyme was not fully stabilized. This is to be expected as it is a multimeric enzyme and it is sometimes difficult for all subunits to be covalently bound to the support surface. With this in mind, a post-immobilization treatment was carried out. This treatment consisted of cross-linking the immobilized catalysts on the optimal supports with different polyfunctional polymers capable of covalently interacting with the different amino acids present on the protein surface. This type of treatment is able to promote the covalent immobilization of all the subunits present in different multimeric proteins. The conserved activity was 35% and 15% when the derivatives were cross-linked with dextran-aspartic acid and dextran-glycine, respectively. The different cross-linked derivatives were incubated at different temperatures and their activity was compared both with that of the soluble enzyme and with that of the non-polymerized derivative. After incubation at 29 ◦C, the activity of the cross-linked derivatives remained unchanged for 24 h. Increasing the temperature up to 35 ◦C promoted the inactivation of the derivative cross-linked with dextranglycine polymers, whereas those cross-linked with dextran-aspartic acid retained 85% of their initial activity after 24 h at this temperature (Fig. 3). 3.4. Elimination of RFOs in soy-based beverages by free and immobilized MelA α -D-galactosidase The capability of MelA α -D-galactosidase from L. plantarum WCFS1 to hydrolyse single buffered solutions of RFOS has already been demonstrated (Delgado-Fernandez et al., 2020), as well as the efficient capacity to synthesize α -GOS (Delgado-Fernandez et al., 2021). However, as a new batch of enzyme was used in this work, the hydrolytic capacity of the enzyme was again tested on raffinose. This trisaccharide was cleaved at the melibiose moiety, releasing sucrose and galactose, achieving 82% degradation after 24 h of reaction. After testing the hydrolytic activity of the enzyme, the following assays were performed to study the hydrolysis of RFOS present in the soy-milk beverage using soluble MelA α -D-galactosidase from L. plantarum WCFS1. Considering the composition of different soy-based beverages, sample L showed the highest content of RFOs whose value is important for comparative purposes after enzyme hydrolysis of these antinutritional factors (Table 1). The pH analysis of the samples was crucial prior to conducting the reactions, as the pH needed to be optimized for either the enzymatic activity of MelA and for each immobilization support. The carbohydrate evolution resulting from the hydrolysis of the RFOs in the soy-based beverage sample treated with the free enzyme form is shown in Fig. 4. In the untreated soy-based beverage, the quantifiable carbohydrates were sucrose, raffinose, stachyose and minor monosaccharides such as fructose, glucose and galactose. During the incubation times studied, complete disappearance of stachyose and raffinose was observed after 1 and 3 h of treatment with free soluble MelA, respectively, together with a concomitant increase in sucrose and release of galactose as a result of hydrolysis of these oligosaccharides, whereas glucose (≈20 mg/100 mL) and fructose (≈10 mg/100 mL) levels remained fairly stable over the course of the reaction time. Stachyose was completely hydrolyzed after 1 h of reaction, while raffinose underwent 53% degradation, reaching almost complete degradation (96%) after 3 h of reaction and no trace could be detected after 5 h of reaction. A similar hydrolysis pattern was observed for all soy beverages described in Table 1. Many α -D-galactosidases have been investigated for the removal of RFOs from soy-based beverages. Gote and coworkers (Gote, Umalkar, Khan, & Khire, 2004) successfully hydrolyzed raffinose and stachyose present in soy beverages at 65 ◦C within 2 h of treatment using a thermostable α -D-galactosidase from Bacillus stearothermophilus (NCIM 51 46). Prashanth and Mulimani (Prashanth & Mulimani, 2005) achieved 93% degradation of RFOs in a soy-based beverage at 50 ◦C after 12 h of reaction using a fungal α -D-galactosidase from Aspergillus oryzae. However, treatment of soy drink with an α -D-galactosidase isolated from Bacillus megaterium VHM1 resulted in complete hydrolysis of RFOs from Fig. 2. Thermal stability of the different MelA α -D-galactosidase preparations. Samples were incubated at 29 ◦C. (n =3) ◆: IDA-Cu; ▴: soluble enzyme; Δ: IDA-Zn; ●: IDA-Cu; □: IDA-Zn-glutaraldehyde; ■: EDA-glutaraldehyde. Fig. 3. Thermal stability of different MelA α -D-galactosidase preparations (n = 3). ◆: soluble enzyme; ■: immobilized on glutaraldehyde supports; ●: immobilized on EDA-glutaraldehyde supports and treated with dextran-glycine; ▴: immobilized on EDA-glutaraldehyde supports and treated with dextranaspartic acid. Fig. 4. Reaction course of L soy-based beverage treated with free MelA (n =2). Δ: Raffinose; ■: Stachyose; ●: Sucrose; ○: Fructose; ◆: Glucose; ▴: Galactose. L. Alves dos Santos et al. LWT 210 (2024) 116864 6 soy-based beverage within 1.5 h of incubation at 55 ◦C and pH 7 (Patil et al., 2010). Treatment of soy-based beverage with an α -D-galactosidase from Bacillus megaterium 3–7 resulted in the complete removal of raffinose and stachyose after 4 h of reaction (Huang et al., 2018). Using a recombinant α -D-galactosidase from Irpex lacteus expressed in Pichia pastoris to hydrolyse RFOs in soy drink, Jang and coworkers (Jang et al., 2019) achieved almost complete hydrolysis of raffinose and stachyose at 60 ◦C in 30 min. In all cases, the optimum temperature for α -D-galactosidases to hydrolyse the RFOs was considerably higher than that used in our work for MelA activity. In the present study, the enzyme is active at 20 ◦C, which is of great interest from the point of view of its application in industry. Our results are in agreement with those obtained by Wang et al. (Wang et al., 2020) who, using an α -D-galactosidase gene from Aspergillus oryzae overexpressed in Pichia pastoris, achieved 97.8% hydrolysis of RFOs after 3 h of incubation at 25 ◦C and with the higher enzyme concentration studied. Our results confirm previous findings indicating a greater preferential activity of α -D-galactosidases towards stachyose than raffinose as substrate. Concerning the hydrolysis of RFOs with immobilization of MelA α -Dgalactosidase, experiments were performed at the same conditions as for the soluble forms. The behavior of the immobilized enzyme was very similar to that of the soluble one, releasing galactose and sucrose throughout the reaction. As in the soy-based drink treated with free soluble MelA, stachyose was completely hydrolyzed after 1 h of reaction, and at this time raffinose was 78% hydrolyzed, reaching complete degradation (100%) after 3 h of reaction. In this case, raffinose degradation was faster than in the soluble enzyme assays. It was found a 79% and 66% reduction of RFOs by free and immobilized α -D-galactosidase from Gibberella fujikuroi, respectively, from soy drink after 3 h of incubation (Thippeswamy & Mulimani, 2002). Using the soluble and immobilized α -D-galactosidase from Aspergillus oryzae, Prashanth and Mulimani (Prashanth & Mulimani, 2005) achieved 93% and 81% degradation of RFOs, respectively, in soy drink at 50 ◦C after 12 h. Katrolia et al., 2019; Katrolia et al., 2019) studied the elimination of RFOs from soy-based beverages using the soluble and immobilized α -D-galactosidase from Aspergillus oryzae over calcium-alginate and chitosan. After 4 h of incubation, the free enzyme hydrolyzed only 30% of the RFOs, whereas the alginate-immobilized and chitosan-immobilized enzyme hydrolyzed 97.6% and 93.7% of the RFOs, respectively. α -D-galactosidase cloned from Bacteroides thetaiotaomicron and expressed in Escherichia coli (iBt_3292) immobilized on alginate supports resulted in the consumption of 98.9% of the RFOs present in soy drink after 48 h of incubation (Shin et al., 2020). The results obtained in the present study show that either in free or immobilized form, α -D-galactosidase from L. plantarum WCFS1 produced a high percentage of hydrolysis of RFOs in soy-based beverage in a short period of time and under mild temperature conditions (20 ◦C). Therefore, the immobilized enzyme could be used on an industrial scale to hydrolyse indigestible RFOs in soy-based beverages, making the process cheaper. 3.5. Reusability of the immobilized α -D-galactosidase One of the major limitations associated with the industrial application of free enzymes is their high cost and operational instability. Therefore, from an economic and sustainable point of view, the heterogeneous biocatalyst must be reused to improve the productivity of the process. For this purpose, the catalyst was reused in successive reaction cycles and its hydrolytic capacity was evaluated. After 10 recycles, the degree of hydrolysis of both stachyose and raffinose was stable, above 80% and 98%, respectively (Fig. 5). These results align with those reported by Çelem and ¨ Onal (2022) and Falkoski et al. (2006) Fig. 5. Effect of the multiple use of immobilized, on support of agarose-EDA-glutaraldehyde, MelA α -D-galactosidase from L. plantarum WCFS1 in different reaction cycles (n =1). The reaction was performed using (0.450 mL) of a commercial soy-based beverage, 5 U/g of immobilized enzyme, at 25 ◦C. The initial concentration of each carbohydrate was, respectively, 14.61 and 23.51 mg/100 mL, for raffinose and stachyose. L. Alves dos Santos et al. LWT 210 (2024) 116864 7 for non-immobilized α -galactosidases, where the degree of hydrolysis for stachyose and raffinose ranges from 78% to 100%. This demonstrates the high efficiency of the various immobilization supports used in this study and highlights the potential for enzyme reusability. 4. Conclusions Considering the antinutritional factors of soy-based beverages and the importance of removing the responsible RFOs for this effect, the industry considers various methodologies for the treatment of these seeds including either conventional or non-conventional procedures. To the best of our knowledge, this is the first study on the carbohydrate composition of soy-based beverages on the Spanish market and the application of MelA α -D-galactosidase from L. plantarum WCFS1 for the removal of RFOs. The activity is strongly dependent on the enzyme concentration and the reaction medium. For this reason, this work also proposes various supports for immobilizing the enzyme in order to stabilize the enzyme for its use in the reactions. MelA α -D-galactosidase from L. plantarum WCFS1 showed specific activity and high hydrolytic efficiency towards RFOs in soy-based beverages at a relatively low temperature (20 ◦C). Therefore, these results suggest that this enzyme could be preferentially used in the early stage of soybean processing to remove RFOs from soy-based beverages and soy-derived products, thereby improving their nutritional value. Compared to other α -D-galactosidases, MelA from L. plantarum WCFS1 has great industrial potential for the removal of RFOs from soy drinks and related foods. Agarose enzyme immobilized catalysts are very useful especially for stirred tank reactors, and considering that the reactions would be carried out at room temperature (20 ◦C), as well as the low catalyst volume/ reaction volume ratio, there would be no major heat transfer problems. Despite these good properties, the limitation of these catalysts would be their use at high temperatures due to the high instability of the soluble starting enzyme. CRediT authorship contribution statement Leandro Alves dos Santos: Software, Methodology, Investigation, Formal analysis. Paloma Delgado-Fern´ andez: Investigation, Formal analysis. Ana Mu˜ noz-Labrador: Formal analysis. Diego MartinGutierrez: Investigation. Evelyn C. Romero: Investigation. Franzory Cuaran: Investigation. Andrea Garcia-Alvarez: Investigation. Elena García-Calvo: Investigation. Blanca de las Rivas: Resources. Rosario Mu˜ noz: Resources. Nieves Corzo: Writing – original draft, Supervision. F. Javier Moreno: Writing – review & editing, Supervision. Cesar Mateo: Writing – review & editing, Validation, Supervision, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability Data will be made available on request. Acknowledgements This work has been financed by the Spanish Ministries of Economy, Industry and Competitiveness (Projects AGL2017-84614-C2-1-R and AGL2017-84614-C2-2-R; PID2021-123862OB-I00. Paloma DelgadoFern´ andez thanks to Ministry of Science, Innovation and Universities of Spain for providing FPI predoctoral fellowship. The authors also wish to acknowledge the funding from European Commission, project “Twinning for intensified enzymatic processes for production of prebioticcontaining functional food and bioactive cosmetics “grant no. 101060130, HORIZON-WIDERA-2021-ACCESS-02-01. Research scholarship was granted to Leandro Alves dos Santos by CAPES - Coordenaç˜ ao de Aperfeiçoamento de Pessoal de Nível Superior. (Finance code 88887.750039/2022-00). Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.lwt.2024.116864. References Amigo-Benavent, M., Silv´ an, J. M., Moreno, F. J., Villamiel, M., & del Castillo, M. D. (2008). Protein quality, antigenicity, and antioxidant activity of soy-based foodstuffs. Journal of Agricultural and Food Chemistry, 56, 6498–6505. https://doi. org/10.1021/jf800697n Betancor, L., L´ opez-Gallego, F., Hidalgo, A., Alonso-Morales, N., Mateo, C., Fern´ andezLafuente, R., et al. (2006). Different mechanisms of protein immobilization on glutaraldehyde activated supports: Effect of support activation and immobilization conditions. Enzyme and Microbial Technology, 39, 877–882. https://doi.org/10.1016/ j.enzmictec.2006.01.014 Bhatia, S., Singh, A., Batra, N., & Singh, J. (2020). Microbial production and biotechnological applications of α -D-galactosidase. Interntional Journal of Biological Macromolecules, 150, 1294–1313. https://doi.org/10.1016/j.ijbiomac.2019.10.140 Çelem, E. B., & ¨ Onal, S. (2022). Removal of raffinose family oligosaccharides from soymilk by α -D-galactosidase immobilized on sepabeads EC-EA and sepabeads ECHA. ACS Food Science and Technology, 2, 1266–1275. https://doi.org/10.1021/ acsfoodscitech.2c00115 Connes, C., Silvestroni, A., Leblanc, J. G., Juillard, V., de Giori, G., Sesma, F., et al. (2004). Towards probiotic lactic acid bacteria strains to remove raffinose-type sugars present in soy-derived products. Le Lait, 84, 207–214. https://doi.org/10.1051/lait: 2003030 Curiel, J. A., de las Rivas, B., Manche˜ no, J. M., & Mu˜ noz, R. (2011). The pURI family of expression vectors: A versatile set of ligation independent cloning plasmids for producing recombinant his-fusion proteins. Protein Expresion and Purification, 76, 44–53. https://doi.org/10.1016/j.pep.2010.10.013 Delgado-Fernandez, P., de las Rivas, B., Mu˜ noz, R., Jimeno, M. L., Doyagüez, E. G., Corzo, N., et al. (2021). Biosynthesis of nondigestible galactose-containing heterooligosaccharides by Lactobacillus plantarum WCFS1 MelA α -galactosidase. Journal of Agricultural and Food Chemistry, 69, 955–965. https://doi.org/10.1021/ acs.jafc.0c06417 Delgado-Fernandez, P., Plaza-Vinuesa, L., Hernandez-Hernandez, O., de las Rivas, B., Corzo, N., Mu˜ noz, R., et al. (2020). Unravelling the carbohydrate specificity of MelA from Lactobacillus plantarum WCFS1: An α -D-galactosidase displaying regioselective transgalactosylation. International Journal of Biological Macromolecules, 153, 1070–1079. https://doi.org/10.1016/j.ijbiomac.2019.10.237 Elango, D., Rajendran, K., Van der Laan, L., Sebastiar, S., Raigne, J., Thaiparambil, N. A., et al. (2022). Raffinose family oligosaccharides: Friend or foe for human and plant health?. In Frontiers in plant science (Vol. 13)Frontiers Media S.A. https://doi.org/ 10.3389/fpls.2022.829118. Falkoski, D. L., Guimar˜ aes, V. M., Callegari, C. M., Reis, A. P., de Barros, E. G., & de Rezende, S. T. (2006). Processing of soybean products by semipurified plant and microbial alpha-galactosidases. Journal of Agricultural and Food Chemistry, 54, 10184–10190. https://doi.org/10.1021/jf0617162 Fehily, A. M. (2003). In Soy (soya) beans: Dietary importance (5398, p. 5392). Academic Press. https://doi.org/10.1016/B0-12-227055-X/01112-3. B. B. T.-E. of F. S. and N. (Second E. Caballero. Fernandez-Lafuente, R. (2009). Stabilization of multimeric enzymes: Strategies to prevent subunit dissociation. Enzyme and Microbial Technololgy, 45, 405–418. https://doi.org/10.1016/j.enzmictec.2009.08.009 Fern´ andez-Lafuente, R., Rodríguez, V., Mateo, C., Fern´ andez-Lorente, G., Arminsen, P., Sabuquillo, P., et al. (1999). Stabilization of enzymes (D-amino acid oxidase) against hydrogen peroxide via immobilization and post-immobilization techniques. Journal of Molecular Catalysis B: Enzymatic, 7, 173–179. https://doi.org/10.1016/S13811177(99)00040-5 Garcia Banos, J. L., Olano, A., & Corzo, N. (2000). Determination of mono and disaccharide content of enteral formulations by gas chromatography. Chromatographia, 52, 221–224. https://doi.org/10.1007/BF02490461 Geng, X., Lei, J., Bau, T., Guo, D., Chang, M., Feng, C., et al. (2022). Purification, characterization, and immobilization of a novel protease-resistant α -galactosidase from oudemansiella radicata and its application in degradation of raffinose family oligosaccharides from soymilk. Foods, 11(19), 3091. https://doi.org/10.3390/ foods11193091. PMID: 36230167; PMCID: PMC9563442. Gote, M., Umalkar, H., Khan, I., & Khire, J. (2004). Thermostable α -D-galactosidase from Bacillus stearothermophilus (NCIM 5146) and its application in the removal of flatulence causing factors from soymilk. Process Biochemistry (Oxford, United Kingdom), 39, 1723–1729. https://doi.org/10.1016/j.procbio.2003.07.008 Guo, B., Sun, L., Jiang, S., Ren, H., Sun, R., Wei, Z., et al. (2022). Soybean genetic resources contributing to sustainable protein production, TAG. Theoretical and Applied Genetics. Theoretische Und Angewandte Genetik, 135, 4095–4121. https:// doi.org/10.1007/s00122-022-04222-9 L. Alves dos Santos et al. LWT 210 (2024) 116864 8 Hou, A., Chen, P., Alloatti, J., Li, D., Mozzoni, L., Zhang, B., et al. (2009). Genetic variability of seed sugar content in worldwide soybean germplasm collections. Crop Science, 49, 903–912. https://doi.org/10.2135/cropsci2008.05.0256 Huang, Y., Zhang, H., Ben, P., Duan, Y., Lu, M., Li, Z., et al. (2018). Characterization of a novel GH36 α -D-galactosidase from Bacillus megaterium and its application in degradation of raffinose family oligosaccharides. International Journal of Biological Macromolecules, 108, 98–104. https://doi.org/10.1016/j.ijbiomac.2017.11.154 Jang, J. M., Yang, Y., Wang, R., Bao, H., Yuan, H., & Yang, J. (2019). Characterization of a high performance α -D-galactosidase from Irpex lacteus and its usage in removal of raffinose family oligosaccharides from soymilk. International Journal of Biological Macromolecules, 131, 1138–1146. https://doi.org/10.1016/j.ijbiomac.2019.04.060 Katrolia, P., Liu, X., Li, J., & Kopparapu, N. K. (2019). Enhanced elimination of nondigestible oligosaccharides from soy milk by immobilized α -D-galactosidase: A comparative analysis. Journal of Food Biochemistry, 43, Article e13005. https://doi. org/10.1111/jfbc.13005 Li, Q., Loman, A., Coffman, A. M., & Ju, L. K. (2017). Soybean hull induced production of carbohydrases and protease among Aspergillus and their effectiveness in soy flour carbohydrate and protein separation. Journal of Biotechnology, 248, 35–42. https:// doi.org/10.1016/j.jbiotec.2017.03.013 Maghraby, Y. R., El-Shabasy, R. M., Ibrahim, A. H., & Azzazy, H. M. E. S. (2023). Enzyme immobilization Technologies and industrial applications. ACS Omega, 8, 5184–5196. https://doi.org/10.1021/acsomega.2c07560 Mateo, C., Palomo, J. M., Fernandez-Lorente, G., Guisan, J. M., & Fernandez-Lafuente, R. (2007). Improvement of enzyme activity, stability and selectivity via immobilization techniques. Enzyme and Microbial Technology, 40, 1451–1463. https://doi.org/ 10.1016/j.enzmictec.2007.01.018 Mateo, C., Pessela, B. C. C., Fuentes, M., Torres, R., Betancor, L., Hidalgo, A., et al. (2020). Stabilization of multimeric enzymes via immobilization and further crosslinking with aldehyde-dextran. In J. M. Guisan, J. M. Bolivar, F. Lopez-Gallego, & J. Rocha (Eds.), Immobilization of enzymes and cells. Methods in molecular biology (clifton, N.J.) (Vol. 2100, pp. 175–187). https://doi.org/10.1007/978-1-0716-02157_11 (2020). Mulimani, V. H., & Ramalingam. (1995). Enzymic hydrolysis of raffinose and stachyose in soymilk by alpha-galactosidase from Gibberella fujikuroi. Biochemistry & Molecular Biology International, 36(4), 897–905. PMID: 8528153. Mutuyemungu, E., Singh, M., Liu, S., & Rose, D. J. (2023). Intestinal gas production by the gut microbiota: A review. Journal of Functional Foods, 100, Article 105367. https://doi.org/10.1016/j.jff.2022.105367 Padalkar, G., Mandlik, R., Sudhakaran, S., Vats, S., Kumawat, S., Kumar, V., et al. (2023). Necessity and challenges for exploration of nutritional potential of staple-food grade soybean. In Journal of food composition and analysis (Vol. 117)Academic Press Inc. https://doi.org/10.1016/j.jfca.2022.105093. Panwar, D., Shubhashini, A., Chaudhari, S. R., Prashanth, K. V. H., & Kapoor, M. (2020). GH36 α -D-galactosidase from Lactobacillus plantarum WCFS1 synthesize Galα -1,6 linked prebiotic α -galactooligosaccharide by transglycosylation. International Journal of Biological Macromolecules, 144, 334–342. https://doi.org/10.1016/j. ijbiomac.2019.12.032 Patil, A. G. G., Praveen Kumar, S. K., Mulimani, V. H., Veeranagouda, Y., & Lee, K. (2010). α -D-galactosidase from Bacillus megaterium VHM1 and its application in removal of flatulence-causing factors from soymilk. Journal of Microbiology and Biotechnology, 20, 1546–1554. https://doi.org/10.4014/jmb.0912.12012 Prashanth, S. J., & Mulimani, V. H. (2005). Soymilk oligosaccharide hydrolysis by Aspergillus oryzae α -D-galactosidase immobilized in calcium alginate. Process Biochemistry (Oxford, United Kingdom), 40, 1199–1205. https://doi.org/10.1016/j. procbio.2004.04.011 Rada-Mendoza, M., Villamiel, M., Ramírez, A., Usuriaga, Y., & Montilla, A. (2022). Quality indicators in lactose hydrolyzed milks and soy beverages from Colombia. Journal of Food Science and Technology, 9, 646–654. https://doi.org/10.1007/ s13197-021-05055-y Sanyal, R., Kumar, S., Pattanayak, A., Kar, A., & Bishi, S. K. (2023). Optimizing raffinose family oligosaccharides content in plants: A tightrope walk. Frontiers of Plant Science, 14, Article 1134754. https://doi.org/10.3389/fpls.2023.1134754. PMID: 37056499; PMCID: PMC10088399. Shin, Y. J., Woo, S. H., Jeong, H. M., Kim, J. S., Ko, D. S., Jeong, D. W., et al. (2020). Characterization of novel α -D-galactosidase in glycohydrolase family 97 from Bacteroides thetaiotaomicron and its immobilization for industrial application. International Journal of Biological Macromolecules, 152, 727–734. https://doi.org/ 10.1016/j.ijbiomac.2020.02.232 Thippeswamy, S., & Mulimani, V. H. (2002). Enzymic degradation of raffinose family oligosaccharides in soymilk by immobilized α -D-galactosidase from Gibberella fujikuroi. Process Biochemistry (Oxford, United Kingdom), 38, 635–640. https://doi. org/10.1016/S0032-9592(02)00010-9 Wang, H., Ma, R., Shi, P., Xue, X., Luo, H., Huang, H., et al. (2014). A new α -Dgalactosidase from thermoacidophilic Alicyclobacillus sp. A4 with wide acceptor specificity for transglycosylation. Applied Biochemistry and Biotechnology, 174, 328–338. https://doi.org/10.1007/s12010-014-1050-8 Wang, H., Ren, P., Mang, L., Shen, N., Chen, J., & Zhang, Y. (2019). In vitro fermentation of novel microwave-synthesized non-digestible oligosaccharides and their impact on the composition and metabolites of human gut microbiota. Journal of Functional Foods, 55, 156–166. https://doi.org/10.1016/j.jff.2019.02.030 Wang, J., Yang, X., Yang, Y., Liu, Y., Piao, X., & Cao, Y. (2020). Characterization of a protease-resistant α -D-galactosidase from Aspergillus oryzae YZ1 and its application in hydrolysis of raffinose family oligosaccharides from soymilk. International Journal of Biological Macromolecules, 158, 708–720. https://doi.org/10.1016/j. ijbiomac.2020.04.256 L. Alves dos Santos et al. LWT 210 (2024) 116864 9