Oligosaccharide and antinutrient content of whole red haricot bean fermented in salt–sugar and salt‐only solutions

Common beans (Phaseolus vulgaris L.) are nutritious and confer numerous health benefits. However, they are also high raffinose family oligosaccharides (RFOs) and antinutrients. Appreciable amounts of RFOs and antinutrients remain after soaking and cooking, causing flatulence and lowered mineral bioavailability to bean consumers. Fermentation has been shown to lower RFOs and antinutrients in bean flours and milk. However, beans are majorly consumed as whole grains. The purpose of this study was to develop a protocol for fermenting whole common beans. We fermented boiled whole red haricot beans and evaluated their effect on RFOs, tannins, and phytates. A factorial research design was used. Beans were sorted, soaked for 15 h, and boiled for 1 h. The beans were then fermented in 2% salt–sugar solution (SSF) and 2% salt‐only solution (SOF) for 120 h. Microbial growth and pH were monitored every 24 h during fermentation. After fermentation, the beans were dried, milled, and the flours subjected to biochemical analysis. Fermentation favored the growth of lactic acid bacteria (LAB), lowering the pH to 3.88 and 5.26 in SSF and SOF batches, respectively. Tannin content reduced significantly by 64.70% and 73.19% in the SSF and SOF batches, respectively. Phytates reduced by 58.88% and 68.85%, respectively. Raffinose reduced significantly by 96.40% and 95.01%, respectively, whereas stachyose reduced by 95.92% and 94.11%, respectively. The highest reduction of antinutrients and RFOs occurred between 24 and 72 h of fermentation. Higher antinutrient losses occurred in the SOF batch, whereas higher RFO losses occurred in the SSF batch.

Most people presoak and thermally process beans before consumption. However, antinutrients like tannins, phytates, and oligosaccharides are heat stable (Sozer et al., 2019). Therefore appreciable amounts of these antinutrients are left in the beans after heat treatment. Further reduction of RFOs and antinutrients is therefore desirable (Starzy nska-Janiszewska et al., 2014)

| Sample preparation
The red haricot beans were hand-sorted to remove dirt and defective grains. About 1.5 kg of the sorted red haricot beans were washed in distilled water and all floats removed. The beans were then soaked in distilled water at a ratio of 1:5 weight per volume (w/v) for 15 h at room temperature. The soaking water was discarded and beans rinsed in distilled water. About 200 g of the soaked beans was transferred to seven experiment bottles. The fermentation bottles were then topped up with 400 ml of distilled water (1:2 w/v). The bottles were put on a hot plate and the beans boiled for 60 min. The boiling water was then drained and the beans allowed to cool in the bottles before fermentation. Boiled beans from one bottle were removed from the bottle, rinsed in distilled water, and then dried at 60 C for 10 h in an oven.
The dried boiled beans were then milled with a steel mill then stored in a freezer awaiting biochemical analysis (boiled sample).

| Preparation of fermentation solutions
The concentration of fermentation solution used in this study was informed by our previous work (Kitum et al., 2018). Fermentation solution (2% salt-sugar solution) was prepared by dissolving 12 g of salt and 12 g of sugar (1:1 w/w) in three fermentation bottles containing 600 ml of distilled water each. Fermentation solution (2% salt-only solution) was prepared by dissolving 12 g of table salt in three fermentation bottles containing 600 ml of distilled water each.
The bottles with the fermentation solutions were sterilized by autoclaving at 121 C for 15 min then allowed to cool to room temperature before use.

| Fermentation of boiled beans
The sterile fermentation solution from above was transferred to the experiment bottles with cooled boiled red haricot beans. The six bottles were labeled 24 h salt-sugar solution (SSF), 72 h SSF, 120 h SSF, 24 h salt-only solution (SOF), 72 h SOF, and 120 h SOF. The fermentation bottles were then left on a sterile bench to ferment at 25 ± 2 C for 120 h. Fermentation solution samples were drawn aseptically using a pipette and sterile pipette tips at 0, 24, 72, and 120 h for pH determination and microbial enumeration.

| pH determination
Approximately 5 ml of the fermentation solution was drawn aseptically after every 24 h of fermentation for pH determination using a pH meter (HI 2211, Hanna Instruments, Japan).

| Microbial enumeration
Conventional microbiological methods were used for microbial enumeration. The 5 ml of the fermentation solution were transferred into a sterile 9-ml test tube and vortexed properly to mix. About 1 ml of the mixture was transferred to a 9-ml test tube containing quarter-strength Ringer's solution to make serial dilutions. Exactly 10 μl of the aliquots from different dilutions were transferred to Petri dishes containing de Man, Rogosa, and Sharpe (MRS) agar for lactic acid bacteria (LAB) enumeration; violet red bile agar (VRBGA) for enterobacteria enumeration; plate count agar (PCA) for total aerobic bacteria enumeration; and potato dextrose agar (PDA) for yeast and molds enumeration. They were then spread plated followed by incubation of the agar plates at respective temperatures. Each analysis was carried out in triplicate. All bacterial and fungal counts were expressed as colony-forming units per milliliter (CFU/ml).

| Sample preparation for biochemical tests
At the end of each fermentation, the fermentation solution was discarded and the fermented red haricot bean (P. vulgaris L) spread on clean trays. They were then dried in an oven at 60 C for 10 h. The dried beans were then milled and the resulting ample flour stored at 4 C in ziplock bags awaiting biochemical analysis.

| Determination of RFOs content on a dry weight basis (DWB)
Extraction of raffinose and stachyose was done using the method of Antonio et al. (2008) with adjustments. About 0.01 g of the raw and processed bean flours was weighed into microcentrifuge tubes containing 250-μl ice-cold chloroform:methanol (3:7 v/v). The contents were then vortexed in a vortex (Iwaki mixer, TM-151, Japan) for 1 min and the mixture incubated at À4 C for 2 h to stop metabolism and extract water-soluble metabolites. After incubation, 200 μl of ice-cold water was added, and the tubes were shaken for 30 min in a shaker (Ika Labortechnik, KS-250 B, Germany). The samples were then topped up to 1000 μl with cold distilled water and centrifuged in a microcentrifuge (Genevac, DNA-23050-A00, England) at 17,900 Â g, at 4 C for 10 min. The upper phase was transferred into sample vials (2.0 ml) followed by liquid chromatography (LC) ion trap mass spectrometric analysis in an LC-mass spectrophotometry (MS) machine (Genevac, DNA-23050-A00, England).

| Determination of tannin content (DWB)
Tannin content was determined using the vanillin-HCL method of Price et al. (1978). About 0.2 g of bean flours was weighed into centrifuge tubes. About 10 ml of 4% methanolic HCL was then added into each of the tubes and then shaken for 20 min on a shaker. The sample mixtures were then centrifuged at 2500 rotations per minute (rpm) for 10 min in a centrifuge (Hettich, D-78532 Tuttlingen, Germany).
The supernatants were transferred into 25-ml volumetric flasks.
About 5 ml of 1% methanolic HCL was then added to the precipitate in the centrifuge tubes and centrifuged as described above. The supernatants were then transferred into the 25-ml volumetric flasks above and topped up to the mark using 1% methanolic HCL. About 1 ml of the supernatants was then transferred into two test tubes each. To one test tube, 5 ml of freshly prepared mixed reagent (8% methanolic HCL + 4% vanillin in methanol) was added. To the other test tube, 5 ml of 4% methanolic HCL was added. A series of catechin standards were prepared and 1 ml of each of the standard concentrations transferred into a test tube and 5 ml of the mixed reagent added. The sample extracts and standards were allowed to sit for 20 min for color development. The absorbance of the sample extracts and standard solutions were read at 500 nm using a UV-vis photospectrophotometer (UV mini 1240 model, Shimadzu, Japan).

| Determination of phytate content (DWB)
Phytates were extracted using the method of Camire and Clydesdale (2006) with modification. About 0.5 g of milled bean flours was extracted. LC-MS (Genevac, DNA-23050-A00, England) analysis was done using Shimadzu Refractive Index Detector (RID 6A). The mobile phase was 0.005 N sodium acetate in distilled water at a flow rate of 0.5 μl/min.

| Statistical analysis
Each analysis was done in triplicate and the experiments conducted three times. Data were presented as means ± standard error of means (SEM) or standard deviation (SD) of three separate determinations.
Contrast ANOVA was conducted and pairwise comparison of estimated marginal means at P ≤ 0.05 using least significant difference (LSD). Statistical analysis was carried out using SPSS statistics version 23.

| Microbial growth
At the start of fermentation of boiled red haricot bean, LAB, total aerobic count (TAC), coliform, yeast, and mold were not detected in both the SSF and SOF batches. This is in agreement with the findings of Ulloa et al. (2015). However, after 24 h of fermentation, significant growth (P < 0.05) in LAB count and TAC occurred. The LAB counts were at log 10 5.6 and log 10 6.3 in the SOF and SSF batches, respectively, whereas the TAC reached log 10 3.5 and log 10 4.0, respectively.
The LAB counts increased with an increase in fermentation time in both batches. At the end of fermentation (120 h), the LAB count in the SSF batch was at log 10 8.5. However, in the SOF batch, there was no significant change (P > 0.05) in LAB count after 72 h of fermentation (log 10 7.4). With increased fermentation time, TAC continued to increase significantly (P < 0.005) in both the SOF and SSF batches. At the end of fermentation (120 h), TACs in SOF and SSF batches were at log 10 8.6 and log 10 9.4, respectively. Throughout the fermentation, no coliform, yeast, or mold was detected in both the SOF and SSF batches.

| pH
The changes in pH of the fermentation solutions during the fermentation of boiled whole red haricot bean are presented in Tables 1 and 2.
The pH of the fermentation solution at the beginning of fermentation was 6.06. A decrease in the pH was observed with increased fermentation time. In the SOF batch, the pH decrease was slow at the end of the fermentation; the pH was 5.26 (Table 2). In the SSF batch, a significant decrease (P < 0.05) to 4.8 was observed after 24 h of fermentation (Table 1). The pH continued to decrease significantly with an increase in fermentation time recording a pH of 3.88 after 120 h.

| Raffinose family oligosaccharides
The effect of fermentation of boiled whole red haricot bean on raffinose concentration is presented in Figure 1. The presoaked boiled beans had a raffinose concentration of 50.30 mg/100 g. Fermentation of the beans in SSF reduced raffinose concentration to 11.86 mg/100 g (76.42% decrease). Increased fermentation time resulted in a further decrease in the raffinose concentration. At the end of 120 h of fermentation, the raffinose concentration had decreased by 96.40% to 1.81 mg/100 g.
In the SOF batch, the raffinose concentration decreased by 48.5% after 24 h of fermentation to 25.9 mg/100 g. However, after 72 h of fermentation, the raffinose content decreased significantly by 81.96% to 4.67 mg/100 g. After 120 h of fermentation, the raffinose concentration in the SOF batch reduced by 95.01% to 2.51 mg/100 g.

| Tannins
The effect of fermentation of boiled whole red haricot bean (P. vulgaris L.) on tannin concentration is presented in Figure 3. The tannin concentration of the presoaked boiled beans was 274.77 mg/100 g. Fermentation resulted in a significant decrease (P < 0.05) in tannin concentration in both SSF and SOF batches. This is in agreement with Granito and Alvarez (2006)

| Phytates
The effect of fermentation on the phytate concentration of boiled whole red haricot bean is presented in Figure 4. The phytate concentration in presoaked boiled red haricot bean was 350.11 mg/100 g.

| DISCUSSION
Heat treatment of the beans through boiling lowered the LAB, TAC, yeasts and molds, and coliforms to undetectable levels. However, after 24 h of fermentation, LAB and TAC growth occurred in both batches. Tope (2013) reported findings similar to this by isolating LAB and bacillus species during the fermentation of cooked lima bean. This could be attributed to the fact that boiling may not have completely eliminated the LAB and TAC but reduced them to undetected levels (Haddaji et al., 2015). Haddaji et al., 2015 also reported that LAB has the ability to survive heat shocks. They produce proteins that enable them to survive under harsh conditions. Coliforms, yeasts, and molds were not detected throughout the fermentation process. This could be attributed to heat treatment and the acidic environment that inhibited their growth (Granito & Alvarez, 2006;Onwurafor et al., 2014). The decrease in pH during fermentation is attributed to the production of organic acids by the LAB (Granito & Alvarez, 2006).
The acid production was higher in the SSF batch compared with the SOF batch. This could be attributed to the presence of easily utilizable sugars in the SSF batch which promote LAB growth (Kitum et al., 2018).
Fermentation drastically reduced raffinose and stachyose concentration in beans. This decrease is attributed to the ability of the fermenting microbes to produce α-galactosidase enzymes that break down the α-1,6-glycosidic linkages (Adewumi & Odunfa, 2009). The presence of raffinose in the beans induces the production of α-galactosidase (Carevi c et al., 2016;Kumar et al., 2012). The highest losses of 76% and 79% in raffinose and stachyose, respectively, occurred after 24 h of fermentation in the SSF batch when the pH was 4.8. In the SOF batch, the highest losses of 82% and 72% in raffinose and stachyose concentration occurred after 72 h of fermentation when pH was at 5.8. This correlated with the optimal time between 12 and 72 h (Kumar et al., 2012) and pH (Carevi c et al., 2016) favorable for α-galactosidase activity. Carevi c et al. (2016) reported that acidic pH favored the activity of α-galactosidase whereas an increase towards neutral pH reduced its activity. This could be the reason why although pH in the SOF batch was higher than 5.0, raffinose and stachyose concentration continued to reduce significantly to levels comparable with SSF. The decrease in RFOs concentration could also be attributed to the exposure of the RFO sugars to microbial α-galactosidases as a result of boiling which softens and breaks the seed coat.
Presoaking of red haricot beans resulted in some losses of tannins. This is attributed to leaching out of tannins into soaking water. Fermentation of the beans lowered the tannin content significantly. This is a result of hydrolysis of polyphenolic compounds of tannin complexes (Adeniran et al., 2013). It was observed that tannin content decreased more in the SOF batch compared with the SSF batch. This could be attributed to the high pH 6.06-5.26 which favors the activity of enzyme tannase. Battestin and Macedo (2007) established that the optimal tannase activity was at pH 6.5. In the SSF batch, the pH had reduced significantly to 4.8 after 24 h of fermentation, therefore lowering tannase activity. In their findings, Battestin and Macedo (2007) reported a tannase activity of 45% at pH 3.5.
The reduction of phytic acid during fermentation could be attrib- and 70.6% loss of phytate after 24 and 72 h of LAB fermentation of kidney beans, respectively. At this time, the pH range was between 6.06 and 6 in the SOF batch and 6.06 and 4.8 in the SSF batch (Tables 1 and 2). Ejigui et al. (2005) reported that optimum activity for cereal phytase is between 5.5 and 5.3. Zamudio and Gonza (2001) also reported that LAB had the highest extracellular phytase activity at pH 5.5. Increased fermentation time, resulted in a further significant decrease in the phytic acid content of the SOF batch. This could be attributed to its high pH throughout fermentation. Increased fermentation time did not cause any significant changes in phytic acid concentration in the SSF batch. This could be attributed to the low pH of ≤4 which could have denatured the phytase enzymes. Ejigui et al. (2005) also reported that in addition to phytase enzyme activity, phytic acid of some cereals can also be eliminated through passive diffusion of water-soluble phytate. This could therefore be the reason why some losses in phytic acid were observed in the SSF batch even after pH was below 5.1.

| CONCLUSION
The boiling of red haricot beans reduced all bacteria and yeast and molds to nondetectable levels. This provided a favorable F I G U R E 4 Effect of fermentation time on phytate concentration during the fermentation of boiled whole red haricot beans in 2% saltsugar solution (SSF) and 2% salt-only solution (SOF) batches. Data presented as mg/100 g ± SEM (n = 3) environment for LAB to dominate the fermentation of the beans, therefore making the fermented red haricot bean safe to consume.
Production of organic acids by LABs during fermentation lowered the pH in both SSF and SOF batches. Low pH provides a conducive environment for microbial and endogenous enzymes to break down the raffinose, stachyose, tannins, and phytates in the beans. The highest reduction of tannins and phytates occurred during the first 24 h of fermentation. In the SSF batch, the RFOs had the highest reduction after 24 h of fermentation, whereas in the SOF, the highest reduction was after 72 h of fermentation. The highest reduction of each antinutrient coincided with the optimum pH of the enzyme responsible for its degradation. We, therefore, conclude fermentation is an effective method for reducing antinutrients and RFOs in whole red haricot bean.

This research was funded by Legume Centre for Food and Nutrition
Security (LCEFoNS) (VLIR-UOS).

CONFLICT OF INTEREST
The authors declare no conflict of interest as a result of the publication of this paper.

ETHICS STATEMENT
Ethical principles governing research were adhered to. No animal or human subjects were used in this study. The red haricot bean used in this study is a common staple in Kenya and has no reported toxicity.

AUTHOR CONTRIBUTIONS
All four authors contributed to the conceptualization of the research topic. Vivian C. Kitum carried out the experimentation and analysis of data. All four authors contributed to the interpretation of data and writing, preview, and editing of the manuscript.

DATA AVAILABILITY STATEMENT
The analyzed data used to support the findings of this study are included within the article.