In vitro digestibility of hydrothermally modified Bambara groundnut (Vigna subterranean L.) starch and flour

Abstract In‐vitro digestibility and functional properties of Bambara groundnut (Vigna subterranean) (BG) flour, and its native and hydrothermally modified starches were investigated. The isolated native starch (BNS) was modified by annealing at 50°C for 48 hr (BAS), and heat‐moisture treated at 100°C for 16 hr at 25% moisture level (BHS). The crude protein of BG flour was 16.88%. The amylose content of the flour and native starch was 6.10% and 27.70%, respectively. Hydrothermal modification increased the gelatinization profile of the starch but reduces its pasting properties. Swelling and solubility of the flour and starches increased with increase in temperature. X‐ray diffraction revealed BNS and BAS exhibited “Type C” crystallinity pattern while BHS was “Type A.” The BNS granule was oval, its diameter between 22 and 30 μm, with no significant change in the morphology of BAS and BHS. The BG flour had 33% resistant starch and 11.63% digestible starch. Heat‐moisture treatment enhanced the resistant starch content of the native starch significantly.


| INTRODUCTION
The number of plant species which nourish humanity is remarkably limited; less than 300 plants species out of 195,000 edible plants are used for food (Simopoulos, 1999). Approximately 17 plant species provide 90% of mankind's food supply, with cereal grains having the greatest percentage (Teixeira et al., 2016). There is need to explore and study other lesser known nutritious crops such as legumes used for food. One of such is Bambara groundnut (BG), an underutilized and lesser known legume, highly covet for its starch and protein. BG is rich in carbohydrates, proteins, and lipid (Baoua, Amadou, Baributs, & Murdock, 2014;Murevanhema & Jideani, 2013), similar to other popular legumes such as cowpea, soyabean, pea etc. (Wang & Castonguay, 2014). BG is also a good source of calcium, fiber, potassium and iron and has high essential amino acids content (De-Kock, 2004). These attributes make BG one of the nutritious food generally consumed by the populace, most especially the rural dwellers with little opportunity to obtain protein-rich food sources. BG pods are eaten as a fresh nut, boiled after drying or grounded into flour. BG flour has a variety of uses in the confectionary industry and also for medicinal purposes. BG are boiled and salted, eaten as a snack, or roasted (Murevanhema & Jideani, 2013). In Nigeria, BG is processed into bean cake, and its flour is widely used in compounding infant food supplements. Its starch is also extracted and used in the preparation of local delicacies.
The development of a value-added product from starch depends on a thorough knowledge of its structure and functional properties (Piyarat, 2008). Due to native starch inherent poor mechanical properties and high sensitivity to moisture, it can be modified by physical, enzymatic, and chemical modifications to produce functional starches with special properties. The effects of physical modification via hydrothermal modification on starch (such as annealing and heat -moisture treatment) are widely reported in the literature (Jacobs & Delcour, 1998;Kim & Huber, 2013;Ovando-Martinez, Whitney, Reuhs, Doehlert, & Simsek, 2013;Wang, Wang, Wang, & Wang, 2017).
Generally, starch consumed by humans can be categorized into three different groups' namely digestible starch, partially digested starch and resistant starch (Englyst & Cummings, 1987). Digestible starch is fully digested in the small intestine; partially digested starch is digested in the small intestine but not completely, some amount escape digestion; while resistant starch completely withstands digestion in the small intestine of a healthy human being. As a consequence of resistant starch's positive impact on health especially in obesity control (Shen, Zhang, Dong, Ren, & Chen, 2015), resistant starch has drawn considerable attention in food formulations with health benefits. Obesity and diabetes have become major public health concerns worldwide, and the number of cases is increasing exponentially every year (Higgins, 2014). Therefore, the expedient solution to this may be the development of foods with high resistant starch, that slow down the rate of digestion of glucose from ingested carbohydrate sources which helps to blunt glycemia, reduces insulin requirements and causes satiety (Miao, Jiang, Cui, Zhang, & Jin, 2015). The health benefit of resistant starch to its consumers also include decreasing the risk of colorectal cancer, lowering plasma cholesterol and triglyceride concentrations, enhancing vitamin and mineral absorption etc. (Aravind, Sissons, Fellows, Blazek, & Gilbert, 2013).
Due to its growing demand as the bean of choice in infant food formulation (Afolabi, 2012), especially among the rural poor in Nigeria and West Africa, there is need to understand BG digestibility. Previous studies on BG flour and starch had focused on its physical properties, chemical and physical modifications, breeding, genotype, applications, etc. (Adebowale & Lawal, 2002;Adebowale & Lawal, 2004;Afolabi, 2012;Eknayake, Jansz, & Nair, 1999;Kaptso et al., 2015;Murevanhema & Jideani, 2013;Ogundele, Minaar, & Emmambux, 2017;Oyeyinka, Singh, Ma, & Amonsou, 2016;Oyeyinka, Singh, Patrick, Gerrano, & Amonsou, 2015;Uarrota et al., 2013). Ademiluyi, Oboh, Boligon, and Athayde (2015) also reported the health benefit of fermented BG in diabetes treatment. However, there is a paucity of information on BG digestibility, the objective of this study, therefore, is to evaluate the in vitro digestibility and functional properties of BG flour and its starches. It is hoped that data generated from this study will enhance the starch's industrial application most especially in the food and pharmaceutical industries.

| Starch isolation
BG (Vigna subterranean) was purchased at Bodija market, Ibadan, Oyo state, Nigeria, and the bean was screened to eliminate defective seeds. BG bean was milled to produce the Bambara groundnut flour (BGF) used in the study. Its starch was isolated as described by Afolabi (2012).

| Hydrothermal modification
The native starch of BG starch (BNS) was hydrothermally modified by heat-moisture treatment at 100°C for 16 hr at 25% moisture level (BHS), while annealing was carried out at 50°C for 48 hr (BAS) following the method of Adebowale, Afolabi, and Olu-Owolabi (2005).
The purity of the isolated starch was evaluated by determining their protein, fat, and ash content.

| Swelling power and starch solubility
The effect of temperature and pH on solubility and swelling power of the starch samples were investigated following the methods of Afolabi et al. (2012).

| Pasting properties
A Rapid Visco-Analyser (RVA Tecmaster, Perten instrument, Sweden) was used to determine the pasting properties of the starch using AACC (2003) method.

| Thermal properties
The gelatinization parameter of the native and modified BG starch was studied with differential scanning calorimeter (DSC) using the procedure of Afolabi et al. (2012).

| X-ray diffraction
The crystallinity of the native and modified starch was determined with a Rigaku D-Max-2200 X-ray diffractometer (Rigaku Denki Co. Tokyo, Japan). The scanning region of the diffraction angle was from 3 to 40°, with target voltage 40 KV, target current, 100 mA, and aging time 5 min. The relative crystallinity of the starches was determined as enumerated by Afolabi (2012).

| Granule morphology
Granule morphology of the starch was studied by scanning electron microscope, SEM (Hitachi TM-1000 Table-top Scanning electron microscope) at 500 magnification.

| Preparation of α -amylase, and glucoamylase
Alpha-amylase and glucoamylase were produced on a solid state fermentation medium containing rice bran, soyabean flour and cassava starch mixed in ratio (10:3:1 w/w) according to Akpan and Adelaja (2004). The mixture in 250 ml Erlenmeyer flask was moistened with sterile distilled water to 60% moisture content. The medium pH was adjusted to 5.0 with 0.1 mol/L HCl, and sterilized at 121°C for 15 min.
The sterilized medium was inoculated with spores of Aspergillus niger for α -amylase production, while spores of Rhizopus oligoshporus were used for glucoamylase synthesis. Both media were incubated at 30°C for 72 hr.
Crude enzyme extracts were recovered by mixing moldy bran with 0.2 mol/L acetate buffer (pH 6.0 for α-amylase, pH 4.5 for glucoamylase) in the ratio 1:4 (w/v) in conical flasks. The mixtures were then shaken on an orbital shaker at 150 rpm at 28°C for 1 hr. The extracts were then filtered using muslin cloth. The filtrates were partially purified using 70% ammonium sulfate and kept at 4°C for further use.

| Resistant starch determination
The in vitro determination of resistant starch content of the BG flour and starches were analyzed by previously described methods (Champ, Martin, Noah, & Gratas, 1999;Englyst, Wiggins, & Cummings, 1982) with some modification. The sample (100 mg starch or flour) was mixed with sodium acetate buffer containing α-amylase, and incubated at 37°C for 16 hr. Absolute ethanol (40 ml) was added to the mixture, equilibrated for 1 hr and centrifuged (Centrifuge 5702R Eppendorf AG 22331 Hamburg, Germany) at 3,913 g for 30 min. The residue was washed twice with 80% ethanol and dried at 60°C. Water (1.56 ml) and 4.0 mol/L KOH (1.5 ml) was added to the dried residue and mixed for 30 min at room temperature. To 1.5 ml of the dispersion; 12 ml of water, 0.65 ml of 2 mol/L acetic acid, and 0.1 ml of amyloglucosidase was added and shaken with the aid of a thermostated shaker (Uniscope SM101 shaking water bath, Surgifriend Medicals, England) for 90 min at 65°C.
The glucose content of the slurry was determined with glucose oxidase assay kit (Cypress Diagnostic, HB009; G-544, Belgium), measuring the absorbance wavelength (PG instruments, T60-U UVvisible spectrophotometer, USA) at 505 nm. The resistant starch content was calculated as mg of glucose × 0.9.

| Total and digestible starch determination
The method of Goni, Alonso, and Saura-Calixto (1997) was employed in determining the total starch content of the BG flour and starches.
The sample (50 mg) was dispersed in 6.0 ml of 2 mol/L KOH, and incubated for 30 min at room temperature. The solubilized starch was hydrolyzed by adding 60 μl of amyloglucosidase, then incubated at 60°C for 45 min in a thermostated shaker (Uniscope SM101 shaking water bath, Surgifriend Medicals, England), and subsequently centrifuged (Centrifuge 5702R Eppendorf AG 22331 Hamburg, Germany) at 3,170 g, 15 min. Glucose oxidase-peroxidase kit (Cypress Diagnostic, HB009; G-544, Belgium) was used to measure the glucose content of the supernatant, and the total starch content was calculated as mg of glucose × 0.9.
The digestible starch was determined by calculating the difference between total starch and resistant starch of the sample on dry weight basis.

| Statistical Analysis
All determinations were carried out in triplicate and result reported as the mean ± standard deviation. The swelling and solubility profile were subjected to one-way analysis of variance (ANOVA) using SPSS statistical software (version 20) to investigate the effect of pH and temperature on starch samples. The Shapiro-Wilk test of normality and the Levene's tests of homogenous variance were carried out to assess the assumptions of ANOVA in order to validate the results.

| Starch composition
The starch yield of BG on flour basis was 41% (Table 1) T A B L E 1 Proximate composition of bambara groundnut: flour (BGF), native starch (BNS), annealed starch (BAS), and heat-moisture treated starch (BHS) (Adebowale et al., 2002) reported for BG. The discrepancy in the yield is probably due to an improved method of isolation of the starch. The BG starch yield is within the 18%-49% range reported for different pulses (Hoover, Hughes, Chung, & Liu, 2010;Mensah, 2011). The moisture content of 9.15% for BGF is at par with the 9.50% and 9.70% reported for BGF by Adebowale and Lawal (2002) and Enwere and Hung (1996), respectively. The moisture content of the native starch (BNS) is 14.11%, while hydrothermal modification reduces moisture content, probably because hydrothermal modification limit the amount of water-retainable by starch (Zavareze & Guerra Dias, 2011).
The 16.88% protein content of BGF (Table 1) is similar to the 16.60% reported by Enwere and Hung (1996), but higher than the 15.48% reported by Piyarat (2008) (2008) and Enwere and Hung (1996), respectively. The amylose content of the native starch, BNS was 27.7%. Oyeyinka et al. (2015) reported varied amylose contents (20-35%) among five genotypes of BG starches. High amylose content in starch has been reported to inhibit swelling during gelatinization and also increase the viscosity (Tester & Morrison, 1990).

| Swelling and solubility
The effect of temperature on swelling power revealed that as the temperature increased, the swelling power of the flour and starch increased ( Figure 1). The increase in swelling power of the starches as the temperature increased is consistent with other reports on leguminous starches (Afolabi, 2012). Although all the starches swelled as the temperature increased, hydrothermal modification induced a reduction in the swelling power of the BG starches, with BHS having the lowest values. The reduction in the swelling power of BHS could be attributed to the increase in crystallinity and strengthening of intermolecular bonds due to heat-moisture treatment (Singh, Chang, Lin, Singh, & Singh, 2011), which leads to a restriction in the swelling of the starch granules. Due to the inhibitory effect of amylose to swelling, starch with low total amylose contents (being less rigid) swell freely when heated (Singh, Kaur, & McCarthy, 2007). Also, the reduc- to that reported for red sorghum by Adebowale et al. (2005). The solubility of the starches in the alkaline medium (8-12) was higher than that in the acidic medium (pH 2-6). The increased solubility in the alkaline medium could be attributed to the enhanced water affinity of the starch at the alkaline pH, and partial gelatinization which usually occurs at this pH (Lawal & Adebowale, 2005

| Pasting and thermal properties
The pasting and thermal properties of the BG starches are presented in  The increase in pasting temperature can be attributed to changes in structure, the increase of crystallinity, and the transition of the partial amorphous region to crystallinity after hydrothermal modification (Zavareze & Guerra Dias, 2011).
The gelatinization profile of the starches was enhanced by hydrothermal modification (Table 2)  Results are means ± standard deviation of triplicate determinations. Means followed by different superscript in the same row are significantly different (p < .05).
F I G U R E 5 X-ray diffractogram of native (BNS), annealed (BAS), and heat moisture treated (BHS) bambara groundnut starch
BAS had the highest ΔH value (9.56 J/g), onset and peak temperature (66.10 and 78.45°C, respectively) but with the lowest gelatinization range of 27.60°C. Other reports also showed that annealing of starch leads to increase in ΔH and gelatinization tem- , with a decrease in gelatinization temperature range (T c -T o ) (Jayakody & Hoover, 2008 (Gunaratne & Hoover, 2002) or as a result of partial gelatinization of amylose and amylopectin molecules that are less stable during heating.

| Starch crystallinity
The native starch, BNS with strong peaks at 5.85 Å and 5.16 Å, medium peak at 3.85 Å and weak peak at 3.37 Å ( Figure 5) exhibit "Type C" pattern which is characteristic of legume. This crystalline pattern was attributed to a mixture of "A" and "B" polymorphs (Afolabi, 2012).
The native starch crystallinity pattern changed from "Type C" to "Type A" after heat-moisture treatment (BHS) with a weak peak at 8.05 Å, medium peak at 3.88 Å and a strong peak at 5.87 Å and 5.17 Å. The presence of "Type A" crystallinity pattern in BHS is similar to reports on potato (Vermeylen, Goderis, & Delcour, 2006) and yam starch (Gunaratne & Hoover, 2002). According to Zavareze and Guerra Dias (2011), the effect of heat -moisture treatment on crystallinity depends on the source of the starch and treatment conditions. Annealing did not alter the crystallinity of the starch, BAS exhibit the "Type C" diffractogram with a weak peak at 7.95 Å and 3.40 Å, medium peak at 3.87 Å and a strong peak at 5.17 Å. This indicates that the expected change in orientation of starch crystallites, crystallite perfection, and formation of amylose crystallites may have been in low magnitude after annealing (Jacobs & Delcour, 1998).

| Granule morphology
The native BG starch granule was oval, with few mild rupture, and diameter is between 22 and 30 μm (Figure 6), this is within the range of 10-45 μm reported by Jane, Kasemsuwan, Leas, Zobel, and Robyt (1994) for pulses with oval granule morphology. The presence of the mild rupture could be attributed to the isolation method of the starch . Annealing and heat-moisture treatment of the starch did not have any significant effect on the granule size, surface morphology, and granule size distribution; this observation is in tandem with reports on starches like oat, wheat, barley, lentil, finger millet, and potatoes Jacobs & Delcour, 1998;Wang et al., 2017).
The scanning electron micrograph of the native (BNS), annealed (BAS), and heat moisture treated (BHS) bambara groundnut starch at 500X magnification
The 33% resistant starch of BG flour was lower than the 35.00% reported for red kidney (Eyaru, Shrestha, & Arcot, 2009), but higher than the 16.43% reported for chickpea (Garcia-Alonso, Goni, & Saura-Calixto, 1998), 25.40% for lentil (Bednar et al., 2001), and 11.03% for faba bean flours (Ambigaipalan et al., 2011). However, the 44.64%  (Table 1). This poor correlation between the resistant starch and amylose content was also reported for starches isolated from different botanical sources (Vasanthan & Bhatty, 1998;Walter, da Silva, & Denardin, 2005;Zhang et al., 2007). Hence, the formation of resistant starches may be attributed to the molecular association between starch components, the degree of crystallinity, and starch gelatinization properties (Zhang et al., 2007). The increase in the resistant starch content of BHS may also be due to some interactions formed during the treatment which may have survived after gelatinization, thereby partly restricting accessibility of starch chains to the hydrolyzing enzymes (Hung et al., 2016). This is corroborated by the enthalpy change which decreased after heat moisture treatment of the starch ( Table 2).
The digestible starch content of the BG starch increased in the following order: BAS > BNS > BHS (Figure 7). The above observation is consistent with the starch's degree of crystallinity and enthalpy of gelatinization (ΔH) which also increased in the following order BAS > BNS > BHS (Table 2)

In-vitro digestibility (%)
Total starch Resistant starch Digesitive starch and diabetes, due to its high resistant starch content. Heat-moisture treatment also reduced the amylose content of the starch with a consequential decrease in its swelling power, this is a desirable property in the application of the modified starch in food and allied industries.