The four varieties of adzuki beans, namely G24, Te Xiao Li No. 1, Gui Nong No. 1, and Qian Xiao Hei (Figure 1), planted at the breeding base (27.32°N, 105.29°E, 1,640 m altitude), were supplied by He You Xun, Bijie Institute of Agricultural Science, Guizhou Province, China. In the growth cycle, the average temperature was 19^°C, the precipitation was approximately 780.6 mm, and the sunshine duration was 667 h. The adzuki beans were ground into flour using a universal high-speed smashing machine (Dongguan Xinmeihua Electromechanical Equipment Co., Ltd., Dongguan City, P. R. China), and then screened using a 40-mesh sieve; the flours were stored at –20°C for further analysis.

Folin-Ciocalteu’s phenol reagent, pepsin (source: gastric mucosa of a pig), and gallic acid were purchased from Sigma-Aldrich Ltd. (St. Louis, MO, USA). 3, 5-Dinitrosalicylic acid (DNS) was purchased from Sinopharm Chemical Reagent Co., Ltd. Acetone and hydrochloric acid (36–38%) were purchased from Chongqing Chuandong Chemical Group Co., Ltd. (Chongqing, P. R. China). Sodium hydroxide was purchased from Chengdu Jinshan Chemical Reagent Co., Ltd. (Chengdu, P. R. China). Deionized water was purchased from Watsons Water Co., Ltd. (Guangzhou, P. R. China). HPLC grade acetonitrile and acetate were purchased from Aladdin company (Shanghai, P. R. China). Sodium carbonate was purchased from Yongda Chemical Reagent Co., Ltd. (Tianjin, P. R. China). Aluminum chloride hexahydrate was purchased from Kemio Chemical Co., Ltd. (Tianjin, P. R. China). Rutin, 2, 3, 4-trihydroxybenzoic acid, catechin, chlorogenic acid, p-hydroxybenzoic acid, and ferulic acid were purchased from Macklin Bioch. Co., Ltd (Shanghai, P. R. China). Alpha-amylase enzyme (type: pancreatic, source: Bacillus) was purchased from Solarbio Science & Tech. Co., Ltd. (Beijing, P. R. China). Hexane, ethyl acetate, acetic acid, petroleum ether, and methanol were purchased from Tianjin Fuyu Fine Chemical Co., Ltd. (Tianjin, P. R. China).

Phenolic extracts were obtained using the method described by Adom and Liu (30) and De La Parra et al. (31). In brief, 4 g of whole adzuki beans powder was homogenized in 15 ml of chilled acetone (80%, v/v) for 15 min. These mixtures were centrifuged (CenLee16R, Hunan Cenlee Scientific Instrument Co., Ltd., Hunan, China) at 8,000 × g for 5 min to collect the supernatants. The residues were again subjected to phenolic extraction twice following the procedures mentioned above, and the collected supernatants were evaporated at 45°C using a rotary evaporator (Shanghai Yarong Biochemical Instrument Factory) and eventually reconstituted to 10 ml with deionized water. The phenolic extracts were stored at –20^°C for further analysis, and the phenolic extraction was performed in triplicate.

Total phenolics (TP) were determined using the method described by De La Parra et al. (31). A volume of 100 μl of gallic acid standard and adzuki bean extracts were mixed with 400 μl deionized water and 100 μl of Folin-Ciocalteu phenol reagent. The mixtures were kept at room temperature for 6 min, and subsequently mixed with 1 ml of 7% Na₂CO₃ solution and 0.8 ml of deionized water. After 90 min of reaction in the dark, 200 μl of the reaction mixtures were pipetted into a 96-well microplate reader (SPECTRAMAX 190, Molecular Devices, Silicon Valley, USA) and determined at 760 nm, and the standard curves were plotted from the absorbances and concentrations of the gallic acid standards. The TP was calculated from the standard curve and expressed as mg gallic acid equivalents (GAE)/100 g DW.

The total flavonoids (TF) were estimated using the methods described by Zhishen et al. (32) and Bakar et al. (33). Five hundred microliters of adzuki bean extracts or rutin standard was allowed to react with 2.25 ml of deionized water and 0.15 ml of 5% NaNO₂ for 6 min; the mixtures were continuously reacted with 0.3 ml of 10% AlCl₃.6H₂O for 5 min. Afterward, 1 ml of 1 mol/L NaOH was vigorously mixed with the mixtures to terminate the reactions. After 15 min of reaction, 200 μl of the mixtures was pipetted into a 96-well microplate reader and determined at 510 nm. The standard curve was plotted from the absorbance and concentration of the rutin standards, and the TF was calculated from the standard curve and expressed as mg rutin equivalents (RE)/100 g DW.

The profile of the phenolic compounds of the adzuki beans was determined using HPLC (34, 35). The adzuki bean extracts were directly passed through a 0.45-μm filter film, and then the filtered extracts (20 μl) were injected into a 1260 HPLC device (1260 Infinity, Agilent Technologies, Santa Clara, CA, USA) equipped with a C-18 chromatography column (30°C) (Agilent ZORBBAX 250 mm × 4.6 mm, 5 μm). Acetonitrile (A) and 1% acetic acid (B) solutions were obtained using the gradient elution program: 0–5 min, 5–15% A; 5–35 min, 15–35% A; 35–40 min, 35–45% A; 40–50 min, 45–5% A, with a flow rate of 1 ml/min. These individual phenolic compounds were detected using an ultraviolet absorption detector at 280 nm. The individual phenolic compounds were identified and quantified according to the retention time and calibration curve of the corresponding standards. The individual phenolic compound contents were expressed as mg phenolics/100 g DW.

Four grams of adzuki beans powder was adequately mixed with 20 ml of deionized water at 25°C for 1 h. Next, the mixtures were centrifuged (8,000 × g, 4°C) for 30 min to collect the water extracts, and then the volume was recomposed to 20 ml. The above process was performed in triplicate (23).

The α-AI specific activity was determined using the method described by Yang et al. (14), with minor modifications. First, the water extracts were appropriately diluted to keep the α-amylase inhibition rate (IR) under 50%. Second, 0.25 ml of diluted water extracts were adequately mixed with 0.25 ml of α-amylase solution (1 U/ml) and 0.5 ml of PBS (pH 6.9) in a water bath (37^°C) for 10 min, and then a 5-min reaction was allowed after the addition of 0.5 ml of 1% (w/v) aqueous starch solution to the mixtures. Eventually, the mixtures were mixed with 1 ml DNS solution and subsequently heated in a boiling water bath for 10 min. After cooling to room temperature, the volume of the mixtures was reconstituted to 20 ml with deionized water and measured at 540 nm using a spectrophotometer instrument (L5S, Shanghai Yidian Analytical Instrument Co. Ltd., Shanghai, China).

The trial groups were estimated as (1) blank tube (no addition of water extracts), (2) blank control tube (addition of neither α-amylase solution nor water extracts), and (3) inhibition control tube (no addition of α-amylase solution). PBS was used to make up for the deficiency.

The inhibition rate of the α-AI in adzuki beans to α-amylase was calculated using formula (1):

Where A₁, A₂, A₃, and A₄ are the absorbances of the blank tube, blank control tube, inhibition tube, and inhibitory control tube at 540 nm wavelength, respectively. The amount of α-amylase that catalyzes the formation of 1 mol glucose/min at 37°C and pH 6.9 was defined as an activity unit (U). The amount of α-AI that catalyzes the inhibition of 1 mol glucose formation per min in an α-amylase catalytic starch hydrolysis reaction at 37°C and pH 6.9 was defined as an activity unit (U), while the α-AI specific activity (U/g) is defined as the α-AI activity unit in 1 g of adzuki bean, and was calculated using formula (2):

Where “1” represents the activity of α-amylase solution (U/ml), “n” is the water extract dilution times, “V” is the total volume of water extract (ml), and “m” is the mass of the adzuki beans powder (g).

The TS of adzuki beans was estimated by referring to the National Food Safety Standards of China GB 5009.9 (36). Adzuki bean powder was first processed in fat to remove sugar, and then it underwent continuous processes, such as gelatinization, saccharification, acid hydrolysis, and neutralization, to obtain the mixtures adopted for TS determination using the alkaline copper tartrate solution titration method.

In brief, an aliquot of bean powder comprising 50 mg of starch was mixed with 2 ml of deionized water, and then the mixture was heated for 10 min in a boiling water bath. In addition, 10 ml of HCl-KCl buffer (0.1 mol/L, pH 1.5) and 0.2 ml pepsin solution (1 mg/ml, dissolved in 0.1 mol/L of HCl-KCl buffer) were sequentially added to the gelatinized mixture, which was then agitated for 60 min in a water bath at 40°C. Furthermore, 15 ml of PBS (pH 6.9) and 5 ml of α-amylase solution (2.6 U/ml) (pH 6.9) were sequentially added to the mixture, which was then agitated for 180 min in a water bath at 37°C, and aliquots of the digested mixture (1 ml) were collected at intervals of 30 min (0, 30, 60, 90, 120, 150, and 180 min) for GI determination. The white bread was subjected to the above processes except for the gelatinization process (37). The other aliquots of the digested mixture (1 ml) were collected for 0, 20, and 120 min to determine the digestive properties of the starch (38). The adzuki beans were processed in triplicate. These mixtures were instantly heated to denature the α-amylase activity; eventually, the collected digestion mixtures were sequentially cooled and centrifuged (4°C, 8,000 × g, 10 min) before glucose content determination. The DNS colorimetry method was adopted to determine glucose content. The aliquots of the digestion mixture were equivalently mixed with DNS (200 μl). In addition, the mixtures were sequentially processed by heating in boiled water for 6 min and cooling in running water to room temperature before recomposing the volume to 2 ml using deionized water. The absorbances of the reacted mixtures were measured using a spectrophotometer at 540 nm.

The slowly digestible starch (SDS), rapidly digestible starch (RDS), and RS contents were calculated using formula (3), formula (4), and formula (5) (38, 39).

Where C₀ represents the initial glucose concentration (mg/ml) before the enzymatic hydrolysis reaction, “C₂₀” and “C₁₂₀” represent the glucose concentration (mg/ml) released within 20 and 120 min, respectively, “V” refers to the initial volume of the enzymatic hydrolysis, “0.9” is the conversion factor of glucose to starch, and “50” represents the starch content (mg) in the measured sample.

The IVGI of adzuki beans was computed as described in previous studies (23, 37). The starch hydrolysis curves were plotted, and the enzymatic hydrolysis time and starch hydrolysis rate were designated as the abscissa and ordinate, respectively.

The starch hydrolysis rate (SR%) was calculated using formula (6):

“C_t” represents the glucose concentration at a time “t” (mg/ml), “C₀” represents the glucose concentration at a time “0” (mg/ml), “32.2” represents the initial volume of the enzymatic hydrolysis reaction system (ml), “0.9” is the conversion factor of glucose to starch, and “50” represents the amount of starch in adzuki bean powder.

Formula (7) was used to calculate the enzymatic hydrolysis kinetic constant (k) and the infinite glucose content (C_∞) (mg/ml).

The area under the hydrolysis curve (AUC) was delineated using formula (8).

Where “t_f” and “t₀” represent the last time point (180 min) and the initial time point (0 min) of the enzymatic hydrolysis reaction, respectively.

The IVGI was computed using formula (9).

The calcHI-starch hydrolysis index indicates the AUC sample/AUC white bread standard.

The values were expressed as means ± standard deviation (n = 3) and tested using the IBM SPSS (one-way ANOVA plus post-hoc Tukey’s b test) software (IBM, NY, USA). P-values less than 0.05 indicated a significant difference.

The TP and TF of adzuki beans are shown in Table 1. The TP ranges from (137.11 ± 4.81) to (154.90 ± 4.45) mg GAE/100 g DW; Te Xiao Li No. 1 has the highest TP content, followed by Qian Xiao Hei, G24, and Gui Nong No. 1. The TF ranges from (279.78 ± 2.54) to (602.00 ± 10.23) mg RE/100 g DW; G24 has the highest TF content. In addition, gallic acid, 2, 3, 4-trihydroxybenzoic acid, chlorogenic acid, catechin, rutin, and p-hydroxybenzoic acid were identified in adzuki beans, as shown in Table 2 and Figure 2. The 2, 3, 4-trihydroxybenzoic acid, chlorogenic acid, and p-hydroxybenzoic acid contents range from (1.19 ± 0.13) to (7.46 ± 0.75) mg/100 g DW, (3.94 ± 0.27) to (7.15 ± 1.08) mg/100 g DW, and (5.96 ± 0.59) to (7.54 ± 0.16) mg/100 g DW, respectively. The catechin and rutin contents range from (12.63 ± 1.38) to (23.05 ± 1.60) mg/100 g DW and (2.41 ± 0.19) to (16.31 ± 2.09) mg/100 g DW, respectively. In Te Xiao Li No. 1, the rutin and catechin contents are dominant with 16.31 ± 2.09 and 23.05 ± 1.60 mg/100 g DW, respectively.

The α-AI specific activity indicates the inhibitory efficiency of α-AI to α-amylase in the starch hydrolysis reaction. The α-AI specific activities of adzuki beans range from (1.76 ± 0.04) to (3.41 ± 0.18) U/g. G24 shows the highest activity and Te Xiao Li No. 1 and Gui Nong No. 1 show minor fluctuations; both are ranked intermediate, while Qian Xiao Hei shows the lowest α-AI specific activity (Figure 3).

The starch digestibility properties can be categorized using the digestive process in which the RDS can be digested in 0–20 min, the SDS can be digested in 20–120 min, and the RS cannot be digested after 120 min (39). In adzuki beans, the proportions of RS to TS range from (69.78 ± 2.45%) to (81.03 ± 0.06%); Te Xiao Li No. 1 has the highest proportion of RS to TS, followed by Gui Nong No. 1, Qian Xiao Hei, and G24. The proportions of SDS to TS are between (11.02 ± 1.58%) and (15.57 ± 0.17%), and the proportions of RDS to TS are between (3.38 ± 0.12%) and (18.62 ± 1.81%) (Table 3).

The SR was determined using the in vitro digestion model coupled with DNS colorimetry, which was used to plot the starch hydrolysis curves (Figures 4A–D). Using the curves, the enzymatic hydrolysis kinetic constant (k) and the infinite glucose content (C_∞) were obtained, and both were eventually used to calculate the GI; the parameters and IVGI are presented in Table 4. The SR of adzuki beans at 180 min ranges from (17.02 ± 0.33%) to (26.60 ± 1.81%). Qian Xiao Hei, Gui Nong No. 1, and Te Xiao Li No. 1 present a low SR in the whole hydrolysis period. The IVGI of adzuki beans ranges from (39.00 ± 0.36) to (56.76 ± 4.21), and Te Xiao Li No. 1 presents the lowest IVGI.