Influence of a non-dairy probiotic matrix based on wheat bran and soybean meal on lactic acid bacteria growth

Abstract

Wheat bran and soybean meal are rich sources of bioactive compounds that are beneficial in preventing disease and stimulating probiotic growth. From the point of view of healthy diets and resource conservation, the objective of this paper was to prepare a functional beverage by selecting an optimal non-dairy probiotic matrix. By comparing the main nutrients, sensory qualities and growth of lactic acid bacteria (LAB) on different substrates, the most suitable non-dairy probiotic matrix was obtained. The matrix comprised the composite enzyme hydrolysate produced by boiled wheat bran and an Aspergillus oryzae culture in a base of wheat bran and soybean meal. The results showed that various kinds of LAB could grow on the matrix, and the viable count could range from 8.91 × 10⁸ to 9.23 × 10⁹ cfu ml⁻¹. The phytic acid in the matrix had a slight effect on the viable probiotic count, and the xylan hydrolysate in the matrix stimulated the growth of LAB. Furthermore, LAB fermentation further decomposed the peptides and saccharides in the matrix, thus increasing the nutritional value of this non-dairy probiotic matrix.

---

1. Introduction

In recent times, the development of non-dairy probiotic food has signalled an important advance in the probiotic food industry. Compared to fermented dairy products, non-dairy probiotic beverages have the advantages of no lactose intolerance and low cholesterol content.¹ The key to the investigation of non-dairy probiotic beverages is matrix selection and the use of an enzyme mixture to modify the substrate.

There is increasing interest in grain as a probiotic substrate, mainly because it provides the benefits of both probiotics and cereals, such as oligosaccharides, arabinoxylans, soluble fibre, phytochemicals and other bioactive components.^2–4 However, two main problems limit the development of grain-based probiotic beverage: one is the digestibility of the nutrients, the other is the lack of certain essential amino acids, especially lysine,⁵ making the protein quality poorer than that of milk. Molin et al.⁶ described the production process for an oat-based probiotic product in which the oat nutrients were hydrolysed by adding α-amylase, β-glucanase and protease to successfully produce a health drink.

It is well known that both wheat bran and soybean meal are by-products of food processing, and are high yield, inexpensive, and high in nutritional value. Wheat bran is well known for being rich in carbohydrates, especially xylan, and soybean meal contains complete protein that can compensate for the shortcomings of grain proteins.^4,7,8 The objective of this study was to obtain a suitable non-dairy probiotic matrix, which could be used to produce a non-dairy probiotic beverage with high nutritional value and efficient use of the resources. The growth of various kinds of LAB on this matrix, the influence of matrix composition on the LAB growth and the distribution of the peptide molecular weight and oligosaccharide content both before and after LAB fermentation were studied to provide a theoretical basis for the development of a non-dairy probiotic beverage.

2. Materials and methods

2.1. Materials

Wheat bran was purchased from the local markets in Wuxi, Jiangsu Province, China. An A. oryzae culture was produced by the Guangdong ZhuJiangQiao Biotechnology Company. As an A. oryzae culture is always used as an intermediate in soy sauce production, we obtained the raw materials directly from a soy sauce factory, expecting it to be the best formulation with high nutritional content and enzyme activity. And the preparation method of the A. oryzae culture was shown in Fig. 1. LAB (L. bulgaricus, L. thermophilus, L. casei, L. thermophilus, L. plantarum, L. casei Zhang, L. rhamnosus) was provided by the food biotechnology research centre at Jiangnan University.

Fig. 1 The substrate compositions of the non-dairy probiotic matrixes. ^aThe centrifugation condition was 4000 rpm × 5 min, and collected the supernatant. ^bThe sterilization condition was 100 °C, 30 min.

The α-amylase (4000 U g⁻¹) and protease (70 000 U g⁻¹) were obtained from the Xuemei Enzyme Preparation Science and Technology Co. Ltd., Wuxi, China. Xylanase (120 KNU g⁻¹) and phytase (11 506 U g⁻¹) were obtained from the Sunhy Biology Co., Ltd., Wuhan, China.

2.2. Preparation of non-dairy prebiotic matrix

The substrate compositions of the non-dairy probiotic matrixes we compared were shown in Fig. 1 (preparation of 1.0 L substrate was shown).

2.3. Preparation of samples with different phytic acid levels

The preparation method was similar to that of substrate 3, with the following differences: for sample 1, the added wheat bran was not pre-boiled; sample 2 had no extra phytase; and for samples 3–5, different quantities (0.1%, 1.0%, 5.0%) of extra phytase were added during enzymolysis, so that the final phytic acid values were all different.

2.4. Preparation of samples with different xylanase levels

The preparation method was similar to that of substrate 3, with the addition of extra xylanase at 0.0%, 0.1%, 1.0% and 5.0%.

2.5. Preparation of samples with different xylan hydrolysate levels

Wheat bran was first treated with amylase and protease, and the residue washed three times with distilled water to remove the starch, protein and other soluble ingredients. Then the insoluble residue was decomposed with xylanase to acquire an appropriate concentration of xylan hydrolysate solution. The samples were MRS medium supplemented with xylan hydrolysate in range of 0–0.6% (v/v).

2.6. The growth of LAB on the non-dairy probiotic matrix

The treated matrixes were inoculated with 2% (v/v) LAB at 37 °C for 24 h, after which the pH, viable bacterial count⁹ or OD600 nm (with the initial blank absorbance deducted), were determined.

2.7. Analysis methods

Reducing sugars were analysed using the DNS method.¹⁰ Total sugars were analysed using the phenol-sulfuric acid method.¹⁰ The amino nitrogen content was analysed by methanol potentiometric titration. The phytic acid content determination used the procedure described by Latta and Eskin¹¹ based on the reaction between ferric ion and sulfosalicylic acid.

The peptide molecular weight distribution was assessed using high speed liquid chromatography (HPLC) (Waters Company, USA) on a TSK_gelG2000_swxl column (7.8 mm id × 300 mm), with 20 μl injected and eluted with acetonitrile : water : trichloroacetic acid (20 : 80 : 0.1) at a flow rate of 0.5 μl min⁻¹.

The oligosaccharide content was determined using a HPLC system (Waters Company, USA) with a Sugarpak1 column (6.5 mm id × 300 mm), with 10 μl injected and eluted with pure water at a flow rate of 0.4 μl min⁻¹.¹²

2.8. Statistical analysis

The SPSS 20.0 package was used to analyse the means and standard deviations of the data.

3. Results and discussion

3.1. Main nutrients and L. rhamnosus growth of three non-dairy probiotic matrixes

As basic components, carbon and nitrogen sources have an important effect on the growth of LAB.¹³ A comparison of the reducing sugar, total sugar and amino nitrogen content of the different substrates is shown in Table 1. The pure wheat bran, substrate 1, had the highest carbon but lowest amino nitrogen content. Due to the addition of soybean meal, the amino nitrogen content of substrates 2 and 3 increased significantly by 163% and 178%, respectively. The sugar content of substrate 3 was higher than that of substrate 2, mainly because of the extra wheat bran (the optimum additive amount and processing method were obtained in our previous study). The results showed that the viable bacterial count in substrates 1 and 2 were adjacent, while that of substrate 3 was higher and reached 3.23 × 10⁹ cfu ml⁻¹ after fermentation. This may be due to adequacy of sugar and amino nitrogen (Table 1).

Table 1 Main nutrient content of different substrates and the growth of L. rhamnosus after fermentation^a

Content	Substrate 1	Substrate 2	Substrate 3
a All values shown are means ± standard deviation.
Reducing sugar (g/100 ml)	3.72 ± 0.05	1.12 ± 0.03	2.02 ± 0.05
Total sugar (g/100 ml)	4.39 ± 0.06	1.32 ± 0.05	2.16 ± 0.09
Amino nitrogen (g/100 ml)	0.060 ± 0.003	0.158 ± 0.004	0.167 ± 0.004
Viable bacterial count (×10^9 cfu ml^−1)	1.88 ± 0.13	2.20 ± 0.19	3.23 ± 0.15
pH	3.79 ± 0.02	3.94 ± 0.02	3.90 ± 0.02

---

However, only amylase, xylanase and protease were used to decompose the nutrients in substrate 1, and the end product was singularly lacking in flavour. In particular, the addition of protease produced the following adverse effects: (1). the production of bitter materials, such as a hydrophobic peptide in the hydrolysate¹⁴ reduced the flavour; (2). the large quantity of sugar intensified the Maillard reaction,¹⁵ producing a black colour with low appeal; and (3). the enzymatic hydrolysate was acidic, so the pH after fermentation was lower than other substrates. In summary, the wheat bran in substrate 1 was unsuitable as a non-dairy probiotic matrix.

Substrates 2 and 3 utilised various enzymes produced by A. oryzae to decompose nutrients, such as amylase, cellulase, xylanase, protease, lipase, glutaminase,^16–18 so that there were none of the bad effects mentioned above, and the enzymatic hydrolysate obtained had the natural aroma of wheat bran and soybean meal. The materials were processed into a liquid, resolving the problem of mouth puckering caused by eating it directly. Moreover, the extra wheat bran added to substrate 3 was boiled before enzymolysis, promoting dissolution of the total sugars and forming a gel from the starch, which benefited enzymolysis. This accelerated the leaching and enzymolysis of nutrients and shortened the time required to achieve balance (with the extension of time, the concentration of nutrients was no longer increased). Compared to that of substrate 2, the reaction time of substrate 3 decreased by 4 h, nevertheless, the reducing sugar and total sugar content increased by 80% and 64%, respectively (Table 1). In brief, substrate 3 showed the most potential as a non-dairy probiotic matrix in this study. The following investigation was based on substrate 3.

3.2. The growth of various kinds of LAB on the non-dairy probiotic matrix

To investigate the applicability of the substrate, the growth of different kinds of LAB on the non-dairy probiotic matrix is studied, and the result is shown in Table 2. It is obvious that all of the bacteria we used reached 10⁸ to 10⁹ cfu ml⁻¹ on this matrix after fermentation, which confirmed that this matrix could support the growth of a variety kind of LAB. The WHO/FAO defines probiotics as “live microorganisms, which when administered in adequate amounts confer a health benefit on the host”, and LAB is a well-known probiotic,^19–21 but their health benefits in this non-dairy matrix need to be researched in future. The final pH ranged from 3.84 to 3.97, conforming to sensory requirements. As L. rhamnosus and L. plantarum grew better than others, they were used for the following exploration.

Table 2 The growth of different kinds of LAB on the non-dairy probiotic matrix^a

Bacteria	Viable bacterial count (log cfu ml^−1)			pH
	0 h	24 h	Increased	0 h	24 h	Decreased
a All values shown are means ± standard deviation.
L. bulgaricus	7.23 ± 0.04	8.96 ± 0.02	1.73 ± 0.05	6.03 ± 0.01	3.85 ± 0.03	2.18 ± 0.02
L. casei Zhang	7.25 ± 0.03	8.95 ± 0.08	1.70 ± 0.10	6.04 ± 0.02	3.84 ± 0.03	2.21 ± 0.04
L. thermophilus	7.19 ± 0.01	8.99 ± 0.03	1.80 ± 0.03	6.01 ± 0.01	3.86 ± 0.02	2.15 ± 0.01
L. acidophilus	7.36 ± 0.03	9.06 ± 0.07	1.71 ± 0.07	6.04 ± 0.01	3.89 ± 0.02	2.14 ± 0.02
L. casei	7.28 ± 0.02	9.09 ± 0.02	1.81 ± 0.03	6.00 ± 0.01	3.96 ± 0.02	2.04 ± 0.02
L. plantarum	7.36 ± 0.02	9.42 ± 0.02	2.07 ± 0.03	6.03 ± 0.01	3.97 ± 0.02	2.06 ± 0.02
L. rhamnosus	7.37 ± 0.03	9.51 ± 0.02	2.14 ± 0.02	6.02 ± 0.02	3.89 ± 0.02	2.12 ± 0.01

---

3.3. The effect of phytic acid on the growth of LAB

Chavan et al. reported that the presence of phytic acid, tannins, and polyphenols that bind to proteins can make the protein indigestible. As the phytase produced by A. oryzae cannot completely remove the phytic acid from the matrix, extra phytase was added to study the influence of phytic acid on the growth of LAB. The only difference between samples 2–5 was the addition of different amounts of phytase, with final phytic acid contents decreased. After LAB fermentation, the viable bacterial count reduced with the decrease of the phytic acid content as a whole (Tables 3 and 4); however, the differences were not very significant, especially for L. plantarum. In addition, phytic acid has anti-aging, antioxidant, and even anticancer effects,^22,23 therefore, we see no need to add extra phytase.

Table 3 The growth of L. rhamnosus in different samples^a,b

Samples	8 h	10 h	12 h	16 h	24 h
a All values shown are means ± standard deviation.b Date with the same letter (a–d) with a column are not significantly different (p < 0.05).
1	5.11 ± 0.04^a	6.31 ± 0.01^a	6.98 ± 0.02^a	6.99 ± 0.04^a	6.97 ± 0.12^a
2	5.26 ± 0.02^b	6.40 ± 0.02^a	7.16 ± 0.04^b	7.14 ± 0.05^a,b	7.18 ± 0.13^a
3	5.27 ± 0.04^b	6.57 ± 0.09^b	7.49 ± 0.02^c	7.52 ± 0.05^b,c	7.70 ± 0.08^b
4	5.23 ± 0.04^b	6.55 ± 0.04^b	7.41 ± 0.14^c	7.45 ± 0.26^a,b,c	7.95 ± 0.22^b
5	5.38 ± 0.04^c	6.74 ± 0.01^c	7.66 ± 0.07^d	7.69 ± 0.06^c	7.79 ± 0.11^b

---

Table 4 The growth of L. plantarum in different samples^a,b

Samples	10 h	12 h	16 h	24 h
a All values shown are means ± standard deviation.b Date with the same letter (a–b) with a column are not significantly different (p < 0.05).
1	5.91 ± 0.25^a	6.75 ± 0.18^a	6.76 ± 0.07^a	6.91 ± 0.28^a
2	5.98 ± 0.23^a	6.86 ± 0.15^a,b	6.85 ± 0.13^a,b	7.00 ± 0.17^a
3	6.23 ± 0.29^a	7.10 ± 0.10^b	7.03 ± 0.03^a,b	7.15 ± 0.20^a
4	6.24 ± 0.14^a	7.09 ± 0.08^b	7.10 ± 0.08^b	7.35 ± 0.24^a
5	6.35 ± 0.06^a	7.18 ± 0.05^b	7.17 ± 0.07^b	7.20 ± 0.05^a

---

The difference between samples 1 and 2 was that the extra wheat bran in sample 1 had not been boiled before enzymolysis. As shown in Tables 3 and 4, we found that sample 1 had the lowest viable bacterial count regardless of the type of bacteria, and although the phytic acid content of sample 1 (33.56 ± 1.79 mg/100 ml) was lower than for sample 2 (53.27 ± 0.82 mg/100 ml), the dissolution of sugars, amino nitrogens and other ingredients were also lower for sample 1, and its starch had no chance of being gelatinised. Thus, we could deduce that the influence of material dissolution and starch gelatinisation was greater than that of phytic acid in preparing the probiotic matrix, so the extra wheat bran should be boiled before enzymolysis.

3.4. The effect of xylan hydrolysates on the growth of LAB

Prebiotics are a non-digestible food component that can selectively stimulate the growth of beneficial gut microbiota and inhibit harmful bacteria in a host, thus contributing to health.²⁴ Arabinoxylans are one of the main constituents of the cell wall of wheat bran, which has demonstrated prebiotic activity when used as an ingredient in a ready-to-eat wheat-based cereal.²⁵ To investigate the influence of xylan hydrolysate on the growth of lactic acid bacteria in this non-dairy probiotic matrix, different ratios of xylanase were added to the matrix. However, although the addition of xylanase slightly increased the reducing sugar content, the growth of LAB remained almost the same (the data was no significant difference (p < 0.05), and not listed here), and the data showed no special regulation. The reason may be that the xylanase produced by A. oryzae²⁶ decomposed xylan relatively thoroughly, and the xylan hydrolysate in the primary matrix thus met the needs of the LAB. Thus, wheat bran was used alone for further study.

The growth of LAB in different xylan hydrolysate contents ranging from 0.0% to 0.6% was studied, as shown in Tables 5 and 6 (the unlisted data showed no significant difference (p < 0.05)). In the initial stage of fermentation, there were no significant differences between different samples. When grew up to 6 h (L. rhamnosus) and 8 h (L. plantarum), differences began to appear. The most significant differences were seen in 12 h (L. rhamnosus), 10 h and 12 h (L. plantarum). With the extension of time, strains grew into the stationary phase, the numbers were no longer increasing, and the significant differences were slighter. On the whole, it is obvious that the higher the xylan hydrolysate concentration, the better the growth of LAB. This study clearly shows that the xylan hydrolysate in wheat bran had a positive effect on the probiotic. The optimal matrix in this study could be used as a combination of probiotics and prebiotics after fermentation, offering a general beneficial effect on human health. As the substrate 3 was the optimal matrix, and L. rhamnosus grew better than other discussed LAB on this matrix, the following research was based on the fermentation products used L. rhamnosus and substrate 3.

Table 5 The growth of L. rhamnosus in different concentrations of xylan hydrolysate^a,b

Samples	6 h	8 h	10 h	12 h	16 h	24 h
a All values shown are means ± standard deviation.b Date with the same letter (a–e) with a column are not significantly different (p < 0.05).
0.00%	2.78 ± 0.11^a	5.15 ± 0.10^a	7.10 ± 0.07^a	8.01 ± 0.04^a	8.95 ± 0.04^a	9.42 ± 0.47^a
0.15%	2.83 ± 0.05^a,b	5.34 ± 0.13^a,b	7.11 ± 0.20^a	8.50 ± 0.02^b	9.31 ± 0.01^b	9.41 ± 0.28^a
0.30%	2.93 ± 0.06^b,c	5.63 ± 0.13^b,c	7.52 ± 0.02^a,b	8.64 ± 0.02^c	9.49 ± 0.14^b	9.64 ± 0.13^a,b
0.45%	2.98 ± 0.02^c	5.84 ± 0.23^c,d	7.89 ± 0.02^b,c	8.96 ± 0.01^d	9.89 ± 0.06^c	9.85 ± 0.07^a,b
0.60%	3.25 ± 0.02^d	6.20 ± 0.25^d	8.11 ± 0.35^c	9.18 ± 0.04^e	10.06 ± 0.10^c	10.23 ± 0.14^b

---

Table 6 The growth of L. plantarum in different concentrations of xylan hydrolysate^a,b

Samples	8 h	10 h	12 h	16 h	24 h
a All values shown are means ± standard deviation.b Date with the same letter (a–e) with a column are not significantly different (p < 0.05).
0.00%	5.14 ± 0.11^a	6.46 ± 0.09^a	7.69 ± 0.02^a	8.02 ± 0.21^a	8.46 ± 0.20^a
0.15%	5.36 ± 0.12^a,b	6.91 ± 0.01^b	7.87 ± 0.01^b	8.18 ± 0.38^a	8.46 ± 0.05^a
0.30%	5.35 ± 0.37^a,b	7.27 ± 0.04^c	8.12 ± 0.01^c	8.68 ± 0.10^b	8.69 ± 0.03^a,b
0.45%	5.75 ± 0.17^b	7.60 ± 0.12^d	8.46 ± 0.03^d	8.88 ± 0.11^b,c	8.76 ± 0.07^b
0.60%	5.91 ± 0.30^b	8.02 ± 0.13^e	8.78 ± 0.05^e	9.36 ± 0.14^c	9.03 ± 0.14^c

---

3.5. The influence of lactic acid fermentation on peptide molecular weight

Proteins are more easily absorbed after decomposition, and the distribution of the peptide molecular weight in the matrix can be used to reflect the decomposition of the proteins. As seen in Table 7, there was no peptide molecular weight higher than 5000, and more than 50% of the peptide molecular weight was below 180 both before and after fermentation. This is evidence that the A. oryzae culture had high peptidase activity, and that all dissolved protein was broken down into smaller molecules, even to amino acids. Moreover, the peptide molecular weight was further reduced after fermentation (Table 7), and the molecular weight below 180 increased markedly to as high as 75.33%, mainly due to the peptidase in LAB.²⁷ The results show that the problem of soybean meal protein being poorly digested and absorbed was solved.

Table 7 The distribution of peptide molecular weight before and after LAB fermentation^a

Molecular weight	Before fermentation%	After fermentation%
a All values shown are means ± standard deviation.
>5000	0	0
5000–3000	0.64	0.01
3000–2000	1.01	0.3
2000–1000	2.76	3.84
1000–500	6.22	5.46
500–180	37.24	15.06
<180	52.13	75.33

---

3.6. The influence of lactic acid fermentation on oligosaccharide content

Due to enzymolysis of the A. oryzae culture, the monosaccharide content before LAB fermentation was more than 50%, which indicates the effectiveness of enzymolysis (Table 8). During fermentation, most of the monosaccharides were used by LAB as carbon sources, so the monosaccharide content reduced markedly. Where enzymolysis also occurred during LAB fermentation, for example, it is reported that L. rhamnosus was able to produce a high beta-glucosidase^28,29 activity, which could further decompose macromolecule sugars in the matrix. Therefore, the polysaccharide content was further decreased, and the oligosaccharide content increased, supplied by the polysaccharides. Moreover, 0.59 g/100 ml disaccharides to tetrasaccharides were left in the matrix, which could be further used by probiotics in the human intestines.

Table 8 Oligosaccharide content before and after LAB fermentation^a

Name	Before fermentation	After fermentation
a All values shown are means ± standard deviation.
Other sugars (g/100 ml)	0.73 ± 0.03	0.39 ± 0.01
Disaccharide to tetrasaccharide (g/100 ml)	0.34 ± 0.01	0.59 ± 0.05
Monosaccharide (g/100 ml)	1.16 ± 0.03	0.40 ± 0.05

---

4. Conclusion

The A. oryzae culture based on wheat bran and soybean meal is a suitable non-dairy probiotic matrix. The addition of extra boiled wheat bran can significantly increase the growth of LAB and also improve the nutrient content of the matrix. Various kinds of LAB can grow well in this matrix. Although the phytic acid content of the matrix had an adverse effect on the growth of LAB, because this effect is small and phytic acid has health benefits, we did not deliberately take steps to reduce the phytic acid. Furthermore, the xylan hydrolysate in the matrix can stimulate the growth of probiotics in the matrix and also in the human body. After LAB fermentation, the peptides and polysaccharides were further broken down to produce more amino acids, oligosaccharides and monosaccharides. This is not only beneficial for digestibility and absorption, but also increases the nutritional value. As a consequence, this matrix has the potential to enhance consumer health via its probiotic, prebiotic and bioactive components.

References

1. F. C. Prado, J. L. Parada, A. Pandey and C. R. Soccol, Food Res. Int., 2008, 41, 111–123.
2. D. Charalampopoulos, R. Wang, S. S. Pandiella and C. Webb, Int. J. Food Microbiol., 2002, 79, 131–141.
3. L. M. E. Cohen, Increasing Whole, 2005, 59, 24–31.
4. K. Katina, A. Laitila, R. Juvonen, K. H. Liukkonen, S. Kariluoto, V. Piironen, R. Landberg, P. Aman and K. Poutanen, Food Microbiol., 2007, 24, 175–186.
5. J. K. Chavan and S. S. Kadam, Crit. Rev. Food Sci. Nutr., 1989, 28, 349–400.
6. N. Molin, C. E. Albertsson, S. Bengmark and K. Larsson, US Pat., 5190755, 1993.
7. R. M. J. W. Cartera and F. Padula, Nutr. Res., 2006, 26, 33–38.
8. X. J. Yang, J. S. Wang and J. J. Guan, Food Feed Ind, 2013, 3, 44–46.
9. M. Rakin, M. Vukasinovic, S. Siler-Marinkovic and M. Maksimovic, Food Chem., 2007, 100, 599–602.
10. W. Zhang, Glycoconjugates biochemical research techniques, Zhejiang University Press, Zhejiang, 2003.
11. M. Latta and M. Eskin, J. Agric. Food Chem., 1980, 28, 1313–1315.
12. S. J. Binbin Gan, Anal. Instrum., 2000, 03, 38–42.
13. A. Hansen and B. Hansen, J. Cereal Sci., 1994, 19, 185–190.
14. C. Hyun and K. Miryung, Food Biotechnol., 1997, 6, 244–249.
15. J. Shibao and D. H. Markowicz Bastos, Rev. Nutr. Braz. J. Food Nutr., 2011, 24, 895–904.
16. M. Machida, O. Yamada and K. Gomi, DNA Res., 2008, 15, 173–183.
17. K. Koibuchi, H. Nagasaki, A. Yuasa, J. Kataoka and K. Kitamoto, Appl. Microbiol. Biotechnol., 2000, 54, 59–68.
18. T. Nemoto, J.-i. Maruyama and K. Kitamoto, Biosci., Biotechnol., Biochem., 2012, 76, 1477–1483.
19. A. J. Marsh, C. Hill, R. P. Ross and P. D. Cotter, Trends Food Sci. Technol., 2014, 38, 113–124.
20. B. P. Lamsal and J. M. Faubion, Food Rev. Int., 2009, 25, 103–114.
21. B. Zhou, H. Gao and z. Chang, Food Sci., 2005, 255–259.
22. I. Vucenik and A. M. Shamsuddin, Nutr. Cancer, 2006, 55, 109–125.
23. E. Graf, J. R. Mahoney, R. G. Bryant and J. W. Eaton, J. Biol. Chem., 1984, 259, 3620–3624.
24. P. L. Conway, J. Nutr., 2001, 45, 13–21.
25. K. C. Maki, G. R. Gibson, R. S. Dickmann, C. W. C. Kendall, C. Y. O. Chen, A. Costabile, E. M. Comelli, D. L. McKay, N. G. Almeida, D. Jenkins, G. A. Zello and J. B. Blumberg, Nutrition, 2012, 28, 1115–1121.
26. Z. A. Chipeta, J. C. du Preez and L. Christopher, J. Ind. Microbiol. Biotechnol., 2008, 35, 587–594.
27. E. R. S. Kunji, I. Mierau, A. Hagting, B. Poolman and W. N. Konings, Antonie van Leeuwenhoek, 1996, 70, 187–221.
28. J. A. Marazza, M. S. Garro and G. Savoy de Giori, Food Microbiol., 2009, 26, 333–339.
29. S. G. Hwang, Korean Journal for Food Science of Animal Resources, 2005, 25, 357–364.

---