The Effect of Fermentation with Kefir Grains on the Physicochemical and Antioxidant Properties of Beverages from Blue Lupin (Lupinus angustifolius L.) Seeds

Plant derived fermented beverages have recently gained consumers’ interest, particularly due to their intrinsic functional properties and presence of beneficial microorganisms. Three variants containing 5%, 10%, and 15% (w/w) of sweet blue lupin (Lupinus angustifolius L. cv. “Boregine”) seeds were inoculated with kefir grains and incubated at 25 °C for 24 h. After processing, beverages were stored in refrigerated conditions (6 °C) for 21 days. Changes in microbial population, pH, bioactive compounds (polyphenolics, flavonoids, ascorbic acid), reducing sugars, and free amino acids were estimated. Additionally, viscosity, firmness, color, and free radicals scavenging properties were determined. Results showed that lactic acid bacteria as well as yeast were capable of growing well in the lupin matrix without any supplementation. During the process of refrigeration, the viability of the microorganisms was over the recommended minimum level for kefir products. Hydrolysis of polysaccharides as well as increase of free amino acids was observed. As a result of fermentation, the beverages showed excellent DPPH, ABTS+·, ·OH, and O2− radicals scavenging activities with a potential when considering diseases associated with oxidative stress. This beverages could be used as a new, non-dairy vehicle for beneficial microflora consumption, especially by vegans and lactose-intolerant consumers.


Introduction
Increasing consumers' demand for healthy foods and awareness of the impact of dietary habits in human well-being has prompted the efforts of developing novel foods with defined health benefits [1][2][3][4][5]. Nowadays, a great number of novel functional foods are available on the market with dairy foods and beverages representing an important segment [1]. One of the most popular dairy foods with Bifidobacterium sp., Lactobacillus sp., Pediococcus sp.) or spontaneously [27,[39][40][41][42][43]. There are no studies available concerning lupin seeds fermented by kefir grains.
No studies have been published on the use of blue lupin (L. angustifolius) seeds for development of kefir-like fermented semi-solid beverages. We hypothesized that fermentation of lupin seeds with kefir grains will allow to obtain new products with high added values in comparison with unprocessed seeds. Thus, the aim of the presented study is to produce beverage based on various concentrations of lupin fermented by kefir grains and evaluation of its bioactivity, microbiological, and physiochemical properties during refrigerated storage for 21 days.

The Lactic Acid Bacteria (LAB) and Yeast Survivability during Cold Storage
LAB (lactic acid bacteria) and yeast concentration after fermentation, and viability over storage period in the beverage formulations (A-sample with 5% (w/w) lupin seeds content, B-sample with 10% (w/w) lupin seeds content, C-sample with 15% (w/w) lupin seeds content) are presented in Table 1. At any time, the bacterial and yeast counts were maintained in the samples over the recommended for kefir level >10 7 CFU/g and >10 4 CFU/g, for bacteria and yeast, respectively. However, during storage, some significant fluctuations (p < 0.05) of both LAB and yeast counts were observed, which may be linked with different nutrients availability in the samples. High survivability of kefir microflora in plant-based beverages was also reported [11,18,19]. Jimenéz-Martínez et al. observed high LAB survivability in lupin milk [16], whereas Schlegel et al. noted comparable survivability of different Lactobacillus strains in fermented lupin protein isolate [43]. However, values reported by Zaworska et al. for fermented lupin seeds [40], are lower than reported in present study. On day 21 significant differences between bacterial (p < 0.05) as well as fungal (p < 0.05) counts were observed. The fact that microorganisms in the samples remained highly concentrated might be due to low (6 • C) storage temperature, as it widely known that refrigerated conditions are one of the pivotal points maintaining LAB and yeast viability in fermented beverages, increasing their shelf-life [1,11,12].  [27]. The observed decrease of pH is also in line with findings of other authors for plant matrices fermented with kefir grains such as hazelnut milk [18], and flaxseed oil cake [11]. However, slight deacidification was noticed during cold storage time (p < 0.05). The acidity of fermented products is commonly maintained or decreased during storage, a fact that is linked with microbial activity, which is a continuous fermentation process in which LAB and yeast assimilate carbohydrates [11,44]. It was reported that some of the common problems in legume-based fermented foods is acidity, which during the fermentation of kefir is of great importance, because of inhibitory effects against spoilage and pathogenic microorganisms [41,42]. The pH level was observed to be below 4.7, which is considered essential maximum for microbial stability [16]. In the present study we hypothesized that kefir microflora would use some structural sugars. Indeed, as presented in Figure 2 there was a significant increase in reducing sugars content (RSC) after fermentation in comparison to the non-fermented samples (p < 0.05). After fermentation, the highest RSC was noticed for sample C (23.24 ± 0.74 mg/mL). The increased RSC may be attributed with enzymatic hydrolysis of oligo-, and polysaccharides from lupin seeds. Plant matrices are rich in non-starch polysaccharides, i.e., arabinoxylans, β-glucans, cellulose, and lignin, and often exist in composite structures with other small molecular weight compounds, e.g., phenolics, flavonoids, and minerals. During fermentation LAB and yeast produce different types of hydrolases, which partially or fully degrade the oligosaccharides into simple sugars [40]. The enzymatic activity is responsible for biopolymers degradation, leading to cell wall degradation (softening). LAB make the food easily digestible, decreasing the level of high-chain carbohydrates and some indigestible poly-and oligosaccharides. This is of great importance from the sensorial, physicochemical and dietary point of view, because carbohydrates digestibility is related to many human health issues [40,45]. The hydrolysis of oligosaccharides and production of simple sugars was reported by Camacho et al., who fermented lupin seeds with L. acidophilus B-1910 and L. fermentum B-585 [41]. Also, Romero-Espinoza et al. observed that probiotic bacteria and yeasts partially degraded oligosaccharides of fermented L. mutabilis [46]. However, in present study a decreased trend of RSC was generally observed in the following days of cold storage (p < 0.05). This behavior is linked with the utilization of produced sugars, required for maintaining of microbial metabolic activity, and was reported in other studies [11]. As presented in Figure 3, total free amino acids level (TFAAL) of all the samples significantly increased as a result of fermentation and an increase trend was observed during cold storage, which indicates directly progressive proteolysis (p < 0.05). On day 21 TFAAL in a range 10.79 ± 0.09 mg Gly/mL (sample A)-27.81 ± 0.03 mg Gly/mL (sample C) was noticed. It is known, that upon fermentation microbial proteases are released and degrade to a certain extend the proteins in a composite food matrix [45]. Many biochemical transformations which alter the ratio of nutritional compounds occur during the fermentation process, and these changes affect the characteristics of processed foods such as digestibility and amino acid profile [25,47,48]. The increased TFAAL level is in line with findings of Bartkiene et al., who showed that lupin protein digestibility and amino acids profile can be improved using lactofermentation [26], and also with conclusions of Kasprowicz-Potocka et al., who reported that fermentation of lupin seeds using yeast increased the concentration of essential amino acids [39].   Means with different lowercase are significantly different at p < 0.05. A-sample with 5% (w/w) lupin seeds content, B-sample with 10% (w/w) lupin seeds content, C-sample with 15% (w/w) lupin seeds content.

The Changes of Total Polyphenolics, Flavonoids, and Ascorbic Acid Contents
The changes of total polyphenolics content (TPC), total flavonoids content (TFC) and ascorbic acid content (AAC) are summarized in Table 2. As expected, the content of bioactive compounds was modified by the fermentation process. A significant increase of TPC (almost 2-fold) in comparison to un-fermented samples was noticed for all the samples on day 21 (p < 0.05). Similarly, a significant increase of TFC was found in all samples (p < 0.05). The highest TPC (35.86 ± 0.03 mg GAE/mL) and TFC (33.89 ± 0.08 QE/mL) was observed in sample C on day 21. Siger et al. found that L. angustifolius seeds are a rich source of polyphenolics and flavonoids with high antioxidant potential [49]. Moreover, these results correspond well with those obtained by Bartkiene et al. who reported higher TPC and TFC content in fermented lupin [25]. Generally, the major phenolic compounds identified in lupin species belong to subclass flavones, phenolic acids, and isoflavones [32,[49][50][51][52]. The main identified flavones in the group are aglycone and/or glycosides of luteolin, apigenin, and diosmetin, while the principal contribution in the isoflavone group is from genistein and its derivatives. In phenolic acids, protocatechuic acid (dihydroxybenzoic acid) and p-hydroxybenzoic acid are the major representative components. Although flavones are found in higher quantities, many of the isoflavones present in lupin species get more importance because of their nonsteroidal phytoestrogenic activity in mammals [32,49]. The effect of fermentation on the TPC and TFC (attributed to delinking of some phenolic compounds that were bounded to proteins and cell wall carbohydrates) and antioxidant activity of plant matrices has been reported in numerous studies [11,12,14,25]. It is of particular importance, because phenolic compounds need to be in a soluble form to enter the human blood circulation system and bring about their antioxidant properties [45]. Generally, fermentation did not affect ascorbic acid level, only on day 21 in sample C, a significantly higher AAC was found (p < 0.05). A similar effect was reported for flaxseed oil cake kefir-like beverages [11].

The Changes of Reducing Power and Radical Scavenging Activties
The conducted research proved that the reducing power, DPPH, ABTS +· , · OH and O 2 − radicals scavenging activities were effectively increased as a result of fermentation (p < 0.05), what is summarized in Table 3. It is noteworthy, that fermented beverages did not lose their antioxidative properties as a consequence of cold storage, but, in fact, exhibited increased antioxidative activity. The highest DPPH (94.27 ± 0.06%), and ABTS +· (92.37 ± 0.11%) inhibition was noticed for sample C on day 21, whereas for this sample the highest ·OH scavenging activity was found on day 14. However, the highest O 2 − scavenging activity (95.47 ± 0.08%) was noticed for sample B on day 21. Lupin is reported to possess the antioxidant activity, and of all the bioactive compounds present in lupin seeds, phenolic compounds are primarily responsible for the antioxidant capacity of the seeds [29,32,49]. Bartkiene et al. also reported higher antioxidant activity of fermented lupin seeds wholemeal and protein isolates when compared with non-fermented ones [25]. Thambiraj et al. found that polysaccharide and mono-sugar fractions from L. angustifolius seeds have strong ABTS +· scavenging activity [53]. Liu et al. observed a greater reducing power of milk-kefir and soymilk-kefir than that of respective milks from which they were made [54]. Similarly, peanut milk kefir extract displayed stronger antioxidant properties than peanut milk alone, suggesting a fermentation impact of kefir grain on peanut milk's efficacy [6]. Generally, the increase of antioxidant activity of the fermented products is linked with microbial activity, production and liberation of certain bioactive compounds which demonstrate reducing power and react with free radicals to stabilize and terminate radical chain reactions [11,14]. Presence of excessive free radicals in a biological system can lead to the process of DNA (deoxyribonucleic acid) damage and hence can cause serious diseases [53]. Oxidative stress is associated with the development of so-called "civilization diseases", such as cancer, stroke, myocardial infarction, inflammation, as well as the degenerative processes associated with aging [55]. Hence, radical scavenging activity is an essential process by which oxidative free radicals can be removed and DNA damage can be prevented, and consumption of foods rich in antioxidants plays an essential role in the prevention of these diseases [53]. Thus, it is reasonable to conclude that fermentation of lupin with kefir grains can result in new compounds with health-modulating potential. A-sample with 5% (w/w) lupin seeds content, B-sample with 10% (w/w) lupin seeds content, C-sample with 15% (w/w) lupin seeds content. Values are means ± standard deviation of triplicate determinations. Means with different lowercase in the same column are significantly different at p < 0.05. Means with different uppercase in the same raw are significantly different at p < 0.05. Table 4 presents the color parameters of both fermented and non-fermented samples. As can be seen, fermentation significantly increased lightness (L*) values (p < 0.05). A similar increase of L* as a result of fermentation was observed for flaxseed oil cake kefir-like beverage. It was observed that yellowness (b*) values of the samples decreased (p < 0.05). An increase of a* value (redness) of sample A was found, whereas values of samples B and C decreased (p < 0.05). A significant fluctuations of color parameters of fermented samples were observed during the storage time (p < 0.05). Those changes can be attributed to pH variations, and oxidation of some pigments presented in the raw material [11]. On the contrary, Jimenéz-Martínez et al. observed lower a* and higher b* values for lupin-based plant milk [16].

Viscosity and Textural Changes
At the beginning, the viscosity of the samples showed significant differences because of lupin seeds concentration as presented in Table 5 (p < 0.05). It was observed that after fermentation and up to day 4 the viscosity of the samples increased (p < 0.05). Similarly, the viscosity of flaxseed oil cake kefir-like beverages increased, which was linked with production of polysaccharide kefiran [11]. However, since day 7 a decrease of the samples' viscosity was found (p < 0.05). Likewise, the firmness values of the beverages showed similar trend. This may be linked with polysaccharides hydrolysis, leading to cell wall degradation (indicated by formation of RSC), and, finally softening of the beverages' matrix. Additionally, proteins play a significant role in textural characteristics formation due to proteins gelation properties. In this context, the changes in the texture and viscosity of the samples can be also explained by the constant breakdown of the proteins bonds as a result of proteolysis and formation of TFAAL [14,37].

Samples Preparation and Fermentation
The process of lupin-based beverages preparation consisted of few steps. Firstly, lupin seeds were mixed with distilled water (w/w) to obtain three variants: A (5%), B (10%), C (15%), and incubated in refrigerator (6 • C) for 24 h. Then, the aqueous phase was replaced with fresh sterile distilled water, and the mixtures were heated at 90 • C for 20 min with constant stirring (250 rpm). Then the mixtures were cooled down to room temperature. The samples were homogenized for 5 min with a homogenizer (SilentCrusherM, Heldolph, Germany) at 12,000 rpm. After homogenization the mixtures were dispensed into containers and were pasteurized by heating for 30 min at 60 • C, then cooled down and stored in a refrigerator day before the fermentation. Kefir-like beverages were produced by mixing 500 mL of a particular variant (pre-heated to 25 • C) with 5 g of kefir grains (containing 1.6 × 10 7 ± 0.40 CFU/g of LAB and 1.5 × 10 7 ± 0.11 CFU/g of yeast) and fermented at 25 • C in 100 mL closed sterile plastic containers for 24 h. After processing, beverages were cooled down and stored at 6 • C for 21 days. The analyses were performed after 1, 4, 7, 14, and 21 days of storage. The non-fermented reference samples were treated the same way but without kefir grains addition, which served for comparison.

Microbiological Analyses and pH Determination
During the overall storage, samples (10 g) were collected and diluted with 90 mL of sterile physiological saline (0.9% NaCl), and serial dilutions were prepared [11]. Lactic acid bacteria counts were determined on MRS medium (Merck, Darmstad, Germany) after incubation at 37 • C under anaerobic conditions for 72 h, whereas yeast counts were determined on Sabouraud Agar supplemented with chloramphenicol at 25 • C for 72 h. The enumeration of microorganisms was performed in triplicate and the viable cell counts were expressed as CFU/g of the samples. The pH of non-fermented and fermented samples were measured directly at 25 • C using a pH-meter (CP-411, Elmetron, Zabrze, Poland).

Supernatants Preparation
To obtain clear fluids for analyses the samples were prepared as described elsewhere [11]. Briefly, the samples were transferred into 1.5-mL Eppendorf tubes and centrifuged at 14,000 rpm/min for 10 min at 20 • C (Centrifuge 5418 Eppendorf, Warsaw, Poland). The supernatants of the particular type of sample were mixed and filtered through 0.22-µm nylon membrane filters (Sigma-Aldrich, Darmstadt, Germany). The obtained clear fluids were used for further analyses.

Determination of Total Polyphenolic Content (TPC), Total Flavonoid Content (TFC), Reducing Sugars Content (RCS), Total Free Amino Acids Level (TFAAL), and Ascorbic Acid Content (AAC)
The total polyphenolics content of each supernatant was determined by Folin-Ciocalteu method [11]. The supernatants (100 µL) were mixed with 6 mL of distilled water and 0.5 mL of Folin-Ciocalteu's reagent. After 3 min, 1.5 mL of saturated Na 2 CO 3 solution was added and the mixture was incubated for 30 min in darkness at 40 • C. The absorbance of the mixture was measured at 765 nm (UV-Vis Thermo Scientific Evolution 220 spectrophotometer). The concentration of TPC was calculated as mg of gallic acid equivalents (GAE) per mL of sample (mg GAE/mL).
The total flavonoids content (TFC) of each sample was determined by mixing 250 µL of supernatant with 1 mL of distilled water and 75 µL of 5% NaNO 2 solution. After 5 min, 75 µL of 10% AlCl 3 solution was added, and the mixture was allowed to stand for 6 min before the addition of 250 µL of 1 M NaOH. The total volume mixture was made up to 3 mL with distilled water, and then the absorbance was measured at 510 nm (UV-Vis Thermo Scientific Evolution 220 spectrophotometer). Quercetin was used for a calibration curve, and the results were expressed as mg of quercetin equivalents (QE) per mL of the sample (mg QE/mL) [11].
The reducing sugars content (RSC) was determined by DNS (3,5-dinitrosalicylic acid) method. A total of 10 g of DNS was dissolved in 200 mL of distilled water by continuous stirring, then slowly 16 g of NaOH (dissolved first in 150 mL of distilled H 2 O) was added. The mixture was incubated at 50 • C with stirring to obtain a clear solution. Then 403 g of potassium sodium tartrate tetrahydrate was added in small portions. The mixture was filtered using a paper filter and the volume was made up to 1000 mL with distilled water. One milliliter of supernatant was mixed with 1 mL of 0.05 m acetate buffer (pH 4.8), and 3 mL of DNS reagent was added, then vigorously shaken. The mixtures were incubated in boiled water for 5 min then cooled at room temperature. The absorbance values were then recorded at 540 nm (UV-Vis Thermo Scientific Evolution 220 spectrophotometer). Glucose in acetate buffer was used for a calibration curve [11].
Total free amino acids level (TFAAL) was determined as described elsewhere with a slight modification [56]. A quantity of 1 mL of the supernatants were mixed with 2 mL of a Cd-ninhydrin reagent (0.8 g ninhydrin was dissolved in a mixture of 80 mL ethanol and 10 mL glacial acetic acid, followed by the addition of 1 g CdCl 2 dissolved in 1 mL of distilled water). The mixtures were vortexed and heated at 84 • C for 5 min and cooled in ice-water, and the absorbance was determined at 507 nm. The results were expressed as milligram Gly per mL of the sample by reference to a standard curve which was first prepared using glycine at various concentrations.
The Tillmans titration method involving a reduction of 2,6-dichlorophenolindophenol was used to determine the ascorbic acid content [57]. Two milliliters of supernatant were mixed with 2 mL of oxalic acid solution (2%) and vigorously shaken. The solution was quickly titrated with 2,6-dichlorophenolindophenol until pink color held for 30 s. The content of ascorbic acid was expressed as milligrams per mL of the sample.

Determination of Reducing Power and Radical Scavenging Activity
To determine the reducing power, the supernatants (500 µL) were placed in a tube, to which 1.25 mL of phosphate buffer solution (0.2 M, pH 6.6), as well as 1.25 mL of 1% potassium ferricyanide solution were added. After incubation at 50 • C for 20 min, 1.25 mL of trichloroacetic acid solution were added to the tube. Next, 1.25 mL of supernatant obtained by centrifugation at 3000 rpm for 10 min was diluted with 1.25 mL of deionized water. Finally, 0.25 mL of 0.1% ferric chloride solution was added to complete the assay. The absorbance was determined at 700 nm which represented the reducing power [14]. DPPH, ABTS +· , · OH, and O 2 − radicals scavenging activities were determined according to the procedures as described in previous study [14]. In brief, the DPPH radical scavenging activity was determined by mixing 1 mL of the supernatants with 1 mL of 0.01 mM DPPH methanolic solution.
The absorbance was measured at 517 nm. Three mL of ABTS +· solution were mixed with 50 µL of the supernatants and the absorbance was measured at 734 nm. To determine · OH radical scavenging activity, 1 mL of supernatants and 1.5 mL of ortophenantroline solution (0.005 mmol/L) were mixed with 2 mL of phosphate buffer (pH 7.4, 0.05 mol/L). Then 1 mL of FeSO 4 solution (0.0075 mol/L) was added and then mixed with 1 mL of H 2 O 2 (0.1%), and finally supplemented with distilled water to a total volume of 10 mL. The reaction solution were kept at 37 • C for 1 h in darkness, then the absorbance was measured at 536 nm. To determine O 2 radical scavenging activity, 3 mL of 50 mmol/L (pH 8.2) Tris-HCl buffer were mixed with 1 mL of the supernatants. These mixtures were mixed with a pyrogallol solution (0.3 mL, 7 mmol/L, preheated to 25 • C) and allowed to react for exactly 4 min, then 1 mL od 10 mmol/L of HCl was added to terminate the reaction, and absorbance was measured at 318 nm.

Texture Profile Analysis and Viscosity Measurements
Texture profiles were performed at room temperature using a Zwick/Roell 2,5 Z equipment (Zwick/Roell, Ulm, Germany), equipped with a cylindrical probe (diameter 40 mm). The samples were analyzed directly, penetration rate into the samples was 10 mm/s, and the penetration depth was 25 mm. From the results of the force-time curves, the firmness and hardness were calculated. The viscosity measurements were performed in a rheometer (AR G2, TA Instruments Ltd., New Castle, DE, USA). The samples were analyzed at 20 • C using a stainless steel cone plate having a diameter of 62 mm. Steady-state flow measurements were carried out at a shear rate 50 s −1 and the viscosity values were obtained from the TA Rheology Advantage Data Analysis equipment software V 5.4.7. (TA Instruments, New Castle, DE, USA).

Statistical Analysis
All data were expressed as mean ± standard deviation (SD). Statistical significance was determined using an analysis of variance (two-way ANOVA) followed by NIR Fisher test. The values were considered as significantly different when p < 0.05. All analyses were performed with Statistica version 10 (StatSoft Polska, Kraków, Poland).

Conclusions
The findings of this work indicated that blue lupin seeds can be fermented with kefir grains, resulting in a novel functional beverage type, with high beneficial microflora (lactic acid bacteria and yeast) viability. The development of kefir-fermented product is possible, allowing the consumption of beneficial microorganisms by consumers where non-dairy alternatives are desired. The developed products are a rich source of bioactive compounds (such as polyphenolics, flavonoids, amino acids) with significant radical scavenging activities. It should be noted that consumption of the beverages may be potentially beneficial to the human organism, especially when considering diseases associated with oxidative stress. However, it is still needed to conduct more research focusing on the in vivo analysis of the benefits of consuming these products and the impact on the functioning of the body.
Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.