Protein digestibility and biochemical characteristics of soybean in three conditions of 2 fermentation processing step: An in vitro study

Protein digestibility of soybean (soaked "S", boiled "B", and fermented "F" soybeans) was 13 changed as 20.58 ± 0.25%, 48.71 ± 0.04%, and 50.21 ± 0.45%, respectively in the 14 preparation of soybean fermentation. After simulated digestion, the increment rate of protein 15 digestibility of both B and F was comparable and higher than that of S accompanying by the 16 accumulations of small protein sub-fractions and essential amino acids. Interestingly, 17 bioactivity parameters of all digested fractions increased by around 2 to 4-fold when 18 digestion stages were progressed with overall F showed the maximum values. Processing 19 not only improves the palatability but also increases protein utilization, the bioavailability of 20 nutrients, and healthy support. The study verified the effect of processing and the benefits of 21 soybean and fermented soybean beyond their basic nutrients which could be claimed as 22 functional foods with higher protein digestibility and indispensable amino acids as well as 23 potential bioactivities.


5
digestion. The higher nitrogen content in F than that in S may have been generated from 79 hydrolysis by bacterial fermentation. Previous studies have shown that an increase in 80 soluble and dialyzable material is caused by biochemical changes in the substrate from 81 fermentation of soybeans using several Bacillus spp. 6 Investigations of the level of protein 82 and carbohydrate breakdown in fermented soybean by Lee et al. 14  of Bacillus fermentation on food nutrient bioavailability in the stimulated digestion. Not only 90 nitrogen content but also digestible material should be determined to confirm digestibility 91 because TSN cannot represent amino acids and very short peptides. Thus, quantitative 92 measurement of the short-chain peptides produced during simulated digestion performed 93 using TCA-soluble peptides can be used to assess protein digestibility. 94 The progress of hydrolysis was also confirmed by the determination of the TCA-soluble 95 peptide content and is depicted in Figure 1B. Besides that, peptides have been reported 96 7 digestibility increased as the digestion stage progressed. F showed the highest protein 133 digestibility (17.05-50.21%) followed by B (4.75-48.71%) and S (1.46-20.58%) after the 134 digestion. The rate of increment of protein digestibility, B showed a better increment (2.64-135 fold) followed by F (1.86-fold) and S (1.13-fold) compared to that of stage 1. Even though B 136 showed the highest protein digestibility increment rate, the protein digestibility of F had a 137 similar rate to that of B (around 1.2-fold) at the final digestion stage. The stable protein 138 digestibility was observed in S from stage 1 until the end of the digestion. During simulated 139 digestion, the protein in F and B was digested better than in S. The increased in protein 140 digestibility shows the ability of digestive enzymes in the simulated gastric digestion stage to 141 digest the protein in F and B. It could be assumed that both boiling and fermentation 142 contributes to the improvement of protein digestibility. Digested F could improve oligopeptide 143 fractions from bacterial fermentation, likewise as aid by heating thus improve protein 144 digestibility. Moreover, the fermentation process results in a higher protein digestibility and 145 available lysine content, especially when it is combined with microbial enzymes to 146 significantly improve protein utilization. However, the slight difference between F and B could 147 be explained by the effect of natto texture that has a slime-coated appearance which could 148 hamper or delay the penetration of digestive enzymes. 19 Besides, the presence of slime or 149 mucilage could form sticky solutions or gels, and impact passage rate, stickiness as well as 150 interactions with digestive enzymes and buffer solution in the stomach and small intestine. 20, 151 21 . 152

Soluble protein fractions and distribution by SDS-PAGE 153
The soluble protein distribution profile at each stage of simulated digestion for S, B and F is 154 shown in Figure 2. Comparing S0, the protein-based anti-nutrient factors, for example TI and 155 allergens, were almost completely broken down and hydrolyzed into low-MW peptides in B0 156 and F0 due to the proteolysis that occurred during fermentation with Bacillus spp. var. natto. enzymes by fermentation, Figure 2 shows a change in the distribution of proteins during 160 simulated digestion. In B0-4 and F0-4, the high-MW proteins were obviously decreased, 161 and the accumulation of small proteins was increased by the progress of the digestion stage. 162 Our results showed that the ratio of the small protein fraction in stage 4 is the highest. During 163 pepsin digestion stages (1-2), the intensity of the protein bands corresponding to 7S and 164 11S fractions was decreased and many peptide bands appeared at < 36 kDa, indicating 165 protein hydrolysis (B1-B4). Simulated digestion of F caused complete degradation of 166 polypeptides > 20 kDa and increased the abundance of oligopeptides with MW < 10 kDa 167 (F4, lane 17, Figure 2 digestion. This result suggests that digestive enzymes containing active proteases 172 decompose the larger proteins. In the case of pepsin hydrolysis, the subunits of fermented 173 soybean protein were partially digested within 1-2 h, whilst in intestinal digestion, most 174 disappearance of larger molecules was found after 2 h. Besides that, fermentation improves 175 digestibility due to the reduction of the presence of anti-nutritional factors such as protease 176 inhibitors (trypsin/chymotrypsin inhibitors), tannins and lectins and the degradation of 177 soybean allergens by microbial proteolytic enzymes. 3,22,23 This study verified that protein 178 was mainly digested to smaller MW size fragments that could be a prime contributing factor 179 to superior bioavailability and benefit to human health. 180

Total phenolic content (TPC) 181
Changes in TPC at each digestion stage of F, B and S are shown in Figure 3A antioxidant content, which means that the antioxidant activity of peptides is higher than that 202 of phenolic compounds. All antioxidant properties improved between before and after 203 digestion. F showed better DPPH, FRAP and MIC activity but not ABTS activity than B. The 204 digested fraction of S showed higher DPPH activity than B; however, it was stable even 205 when the digestion stage progressed. Besides that, the antioxidant activity of S was stable 206 as the progress of the digestion. The digested fraction of F showed the highest DPPH 207 activity (24.12-68.00 µmol TE/g protein) followed by S (40.03-51.52 µmol TE/g protein) and 208 B (4.19-22.12 µmol TE/g protein) ( Figure 3B). According to Yadav et al., 25 a reduction in 209 DPPH activity was observed in two of the studied cultivars (EC4216 and BL2) of cowpea 210 seeds after thermal treatment as a result of leaching out of phenolic compounds from seeds 211 due to heat application. Besides, phenolic compounds, especially tannins, are also likely to activity than S and B, it may be because the activity of S comes only from phenolic 215 compound reactions not of both phenolic compounds and oligopeptides as in F. An increase 216 in DPPH activity in F may corresponding to higher TPC content because microbial enzymes. 217 Higher molecular weight phenolic compounds were depolymerized to simple phenolic 218 monomers like catechins by metabolic activity of microbes. Additionally, fermentation can 219 also change the level of bioactive compounds and can further breakdown cell walls of seed 220 leading to liberation or synthesis of various bioactive compounds. 26, 27 221 The DPPH activity increased with the digestion stage progressed except that it was stable in 222 S. The increment of DPPH activity during gastric digestion was 2.42-and 1.96-fold whereas 223 that after the intestinal stage was 1.17-and 2.69-fold in F and B, respectively. Even though 224 the DPPH activity in digested fractions of F was higher than that of B, the big change in 225 increment was found in B. The same trend as DPPH was observed for ABTS: the activity 226 increased in the digestion stage progressed. ABTS ranged from 0.84 to 3.62, 2.07 to 2.99 227 and 0.97 to 1.08 mg AA/g protein in F, B and S, respectively, as shown in Figure 3C In the digested fractions of F, the biggest increment antioxidant activity was found in ABTS 234 which calculated to almost 2-fold of increment (P < 0.05) whereas it increased slightly for B 238 (9.96 to 18.95 µmol FeSO4/g protein) and was stable for S (7.13 to 9.22 µmol FeSO4/g 239 protein) ( Figure 3D). The augmentation in FRAP of digests shows that F proteins can be 240 more effective in donating electron after simulated digestion. The chelation of transition 241 metals such as Fe 2+ and Cu 2+ helps to delay peroxidation and subsequently prevent food 242 rancidity. The change of MIC activity during simulated digestion of F, B and S was also 243 investigated ( Figure 3E). For the digested fractions of F, the MIC value was drastically 244 increased from 2.76 to 5.47 and 6.10 µmol EDTA/g protein for F0, F3 and F4, respectively 245 (P < 0.05), which totally 2.2-fold of increment. Surprisingly, the same increment in MIC was 246 also found in S (0.92 to 4.28 µmol EDTA/g protein). A 2.13-fold increase was observed in 247 the gastric stage, and a 2.00-fold increment in the intestinal stage in S. In contrast, MIC was Apart from phenolic compounds, antioxidant activity changes during simulated digestion, 258 suggesting that generated bioactive peptides might play a main role. In the digestion, pepsin 259 possibly disrupted the spatial structure of soybean peptides conducive to binding and 260 trapping of metal ions and free radicals, resulting in reduced chelation and free-radical 261 scavenging activity. In combination with, pancreatic digestion fully exposed or newly formed 262 the high-affinity metal-binding groups including imidazole and carboxylic groups, thus ionic 263 and electrostatic interactions with metal ions were likely imposed. 264 This study showed that antioxidant activity increased (in F) or was stable (in B) as the 265 digestion stage progressed. The results distinctly stipulate that F in the last stage of 266 simulated digestion showed the highest activity. This may be due to short chain peptides 267 convert free radicals into more stable products by donating electron atom to cease the 268 radical chain reaction. 30

Free amino acid composition 283
The change of free amino acid profile was presented in terms of increment between before 284 (stage 0) and after simulated digestion (stage 4) and is shown in Tables 1 and 2. When 285 compared to S before digestion (stage 0, BD), the total increment of free amino acids in S, B 286 and F after digestion (stage 4 or G2I2) was 2.15-, 6.78-and 21.10-fold, respectively. The 287 distinctive amino acids found in the soybeans were Ile, Glu, Val, Leu, Tyr, Phe, Lys and Asg 288 (> 20 nmol/mL). All other amino acid contents also increased; in particular, B4 and F4 289 showed the biggest increment (Table 1). When comparing between F0 and F4, most 290 represented amino acids were increased significantly, around 1.5-to 3.5-fold. For F, the 291 smallest change and the maximum changes were found in Pro (0.95-fold) and Arg (49.43-292 fold), respectively, in F4. For B, there was a big difference in some amino acids between 293 before and after digestion. There was drastic increase in Leu, Tyr, Phe, Lys and Arg, around 294 5-to 28-fold, found in B4. 295 Table 2 shows the change in content of free amino acids by different groups. The maximum 296 change of 9.52-fold in B4 and 4.19-fold was observed in F4 in the essential amino acid 297 (EAA) group. It can be assumed that the simulated digestion improves the generation of 298 EAA as observed from the increment number (Table 2). We also found a remarkable 299 increment in other groups of amino acids, for example, hydrophobic amino acids (HAA),

In vitro anti-inflammatory activity 314
The digested soybeans fractions showed different trends in anti-inflammatory activity 315 tested by NO production and egg albumin denaturation inhibition as shown in Figure 4A

Sample preparation 368
Soybean samples during fermentation process were collected after soaking (soaked 369 soybeans "S"), boiling (boiled soybeans "B") and fermentation (fermented soybeans "F") 370 following the method of Ketnawa and Ogawa. 9 In summary, dehulled yellow-seeded 371 soybean samples (900 g) were washed using tap water and soaked in distilled water 372 (soybeans/water ratio of 1 : 3, w/v) for 18 h at 20 °C. Then, the soaked samples were 373 separated from the soaking water. At this stage, part of the soaked samples was collected 374 as soaked soybeans (S). Subsequently, soaked samples were washed again with tap water, 375 boiled with the same ratio of fresh distilled water using a household pressure cooker (H-376 5040, Pearl Metal Co., Ltd., Niigata, Japan) under approximately 100 kPa gauge pressure 377 (approx. 120 °C) for 90 min, and then cooled down at room temperature. At this stage, a part 378 of the boiled samples was collected as boiled soybeans (B). Fermented soybeans were 379 prepared according to Ketnawa and Ogawa. 9 In summary, boiled soybeans (150 g) were 380 transferred into a glass beaker and inoculated with 50 mL of the diluted culture of Bacillus 381 spp. natto from a commercial natto product S-903 (Takanofoods Co., Ltd, Tokyo, Japan). 382 After inoculation, the soybeans (37.5 g) were packed into a paper cup (205 mL), the top 383 surface covered with polyvinylidene chloride wrap film, and incubated at 40 °C for 18 h. The 384 products were collected and considered as fermented soybeans (F). 385

Simulated in vitro gastrointestinal digestion 386
The simulated static in vitro gastrointestinal digestion model described by Ketnawa and 387 Ogawa 9 was used with minor adjustment and carried out in duplicate. Sampling of 5 sample 388 sets was performed separately from each reactor. The sample was named as follows: 1) 0 or 389 BD for samples before digestion, 2) 1 or G1 for samples from gastric digestion for 1 h, 3) 2 390 or G2 for samples from gastric digestion for 2 h, 4) 3 or G2I1 for samples after gastric 391 digestion for 2 h and intestinal digestion for 1 h, and 5) 4 or G2I2 for samples after gastric 392 digestion for 2 h and intestinal digestion for 2 h (after digestion). The following analyses were 393 carried out in triplicate.
soybeans. The protein digestibility was calculated using the following formula:

Protein digestibility (%) = B/A × 100%
(1) 404 where A is the total protein content of soaked soybeans, and B is the TCA-soluble 405 peptide content at each digestion stage. 406

Soluble protein fractions and distribution by electrophoretic analysis 407
Before studying protein patterns, Biuret method according to Gornall    Trp n/a n/a n/a n/a n/a n/a Each value represents the mean of three replications ± standard deviation. Means in a 574 column with a different letter are significantly different (P < 0.05). Different capital letters in 575 the same column indicate a significant difference (P < 0.05) among the same conditions of 576 digestion. 577 **Cysteine (Cys) is determined in the form of cystathionine (Cysta). Cysta is a dipeptide 578 formed by serine and homocysteine; the trans-sulfuration of methionine yields homocysteine