Using Moczkowska's method (18), the dried PS sample (50 g) from 1,000 g of PS material was dissolved in the deionized water (1:20, w/v) and adjusted to pH 4.5 with polybutylene succinate (PBS). Approximately 0.8% of the mixed enzyme was added to the PS solution, stirred at 58°C for 1 h, inactivated for 9 min, centrifuged at 3,600 rpm/min for 8 min, and precipitated overnight with 4 times the volume of ethanol. The collected sediment was sufficiently dried to yield E-SDF.

Soluble dietary fibers were prepared based on the microwave method proposed by Kapusniak (19). The dried PS sample (50 g) from 1,000 g of PS material was dissolved in deionized water (1:25, w/v), and the pH value was adjusted to 4.5 with PBS. Protease was added to the sample and stirred at 58°C for 30 min, then placed in an ANKS-C1 microwave processor (ANKS Company, Qingdao, China) at 700 W for 5 min, and centrifuged and precipitated to obtain M-SDF. Detailed operations have been described earlier to obtain M-SDF.

According to Kim et al.'s method (20), 50 g of dried PS was weighed and mixed with 20 times the volume of distilled water at room temperature. The pH was adjusted to 4.5. Protease (0.6%, m/v) was added to the mixture, stirred at 60°C for 40 min, inactivated on a water bath at 100°C for 10 min, and then the parameters of electrical field intensity of 8 kV/cm, frequency of 1 Hz, and pulse width of 20 μs were set on an EX-1900 pulsed electric field device. Detailed operations have been described earlier to obtain PEF-SDF.

Water-holding capacity (WHC) was tested in accordance with a previous report (21). The SDF (0.5 g) was dissolved in distilled water, kept at room temperature for 24 h, and centrifuged at 4,000 rpm for 20 min. The sediment was weighed, and the WHC was calculated using Equation 1:

where W₁ refers to the weight (g) of the EP tube before centrifugation; W₀ represents the weight (g) of the EP tube without the supernatant; and W refers to sample weight.

Using Zhang et al.'s method (22), the SDF (0.5 g) was put into olive oil (1:10, m/v) and kept at 23°C for 6 h. The mixture was centrifuged at 6,000 rpm for 20 min, and then, the collected sediment was weighed to calculate oil-holding capacity (OHC) as described in Equation 2:

where W₁ refers to the weight (g) of the EP tube before centrifugation; W₀ represents the weight (g) of the EP tube without supernatant; and W refers to sample weight.

Swelling capacity (SC) was measured based on Zhang et al.'s report (23). The SDF (0.2 g) and was mixed with distilled water (5 ml) to hydrate at 4°C for 18 h. The final SDF volume was observed to calculate SC using Equation 3:

where V represents the final volume and W is the weight of SDF.

First, the emulsion was prepared as follows: The SDF (2 g) was dispersed in deionized water and homogenized. Corn oil was added to the mixture and stirred for 2 min. The emulsion was then centrifuged at 1,500 rpm for 3 min (24). Emulsifying activity (EA) was calculated using Equation 4:

where V₁ and V are the volumes of the emulsified layer and total liquid, respectively.

After heating for 30 min at 80°C, the emulsion was cooled to 23°C. Emulsion stability (ES) was calculated using Equation 5:

where V₁ and V represent the emulsified layer volume and the total liquid volume, respectively.

Using Coffman et al.'s method (25), a series of SDF suspensions were prepared with the SDF from 2% to 12% (w/v), heated at 98°C for 60 min, and then kept in an ice bath for 60 min. When the suspensions were changed to a solid state even after inversion and shaking, it was recorded as the lowest concentration of the original SDF suspensions and referred to as the least gelation concentration (LGC).

The soluble dietary fiber (1 g) was dissolved in deionized water (~25 ml), and the viscosity curve was tested using a Rheometer (Anton, MA, USA) at shear rates ranging from 0.1 to 1,000 s.

Approximately 10 mg of SDF samples were mixed using thermogravimetric analysis/differential scanning calorimetry (TGA/DSC) and analyzed for thermal properties in the temperature range of 30–300°C. The heating rate was set at 5°C/min, and the flow rate of liquid nitrogen was 50 ml/min.

Glucose adsorption capacity (GAC) was determined as described in a previous report (26). Briefly, the SDF sample (0.5 g) was mixed with 50–100 mmol/L glucose solution (50 ml) and incubated. After 6 h, the mixture was centrifuged, and then, the supernatant was collected to determine the reducing sugar content. GAC was calculated using Equation 6:

where A_i and A are the reduced sugar content without and with the SDF, g/100 g; v refers the solution volume, and m refers to SDF weight, g.

Pancreatic lipase activity inhibition (PLAI) was tested using Chau's method (27). Briefly, 0.5 g of SDF, 1 ml of pancreatic lipase, 10 ml of soybean oil, and 50 ml of PBS were stirred for 60 min and then placed in an ice bath for 10 min. The mixture was titrated with NaOH (0.1 mol/L) using a phenolphthalein indicator (10 g/L). PLAI was calculated by using Equation 7:

where V and V₁ are the volumes of NaOH consumed without and with SDF, ml; C refers to NaOH concentration, mol/L; and M represents free fatty acid molar mass.

Cholesterol adsorption capacity (CAC) was measured according to Jia et al. (13). Briefly, the yolks and deionized water were mixed at the ratio of 1:9 (v/v), and homogenized to obtain an emulsion, and the SDF sample (1 g) was added to 20 ml of the emulsion. Next, the emulsion was adjusted to pH 2.0 and 7.0, respectively, and incubated at 37°C for 0.5, 1, 1.5, or 3 h. Finally, the incubated emulsion was centrifuged at 6,200 rpm/min for 12 min. The CAC was calculated using Equation 8:

where W is the SDF weight; C_b and C_d correspond to the cholesterol concentration in the emulsion.

Nitrite ion adsorption capacity was determined according to Zhu et al.'s method (28). At pH 2.0 and 7.0, 0.1 g of the SDF sample was mixed with 5 ml of NaNO₂ solution (20 g/ml) and incubated at 37°C. After a 2-h incubation, the mixture was centrifuged, and the supernatant was moved into a tube and filled with 2.0 ml with deionized water. Finally, 2 ml of p-aminobenzene sulfonic acid solution (4 g/ml) and 1 ml/L naphthalene diamide hydrochloride solution (2 g/ml) were mixed to test NaNO₂ concentration to calculate the value of NIAC using Equation 9:

where m is NaNO₂ weight before adsorption, m₁ represents NaNO₂ weight after adsorption, and w refers to the weight of the sample.

Using the method proposed by Huang and Ma (29), SDF (0.5 g) was dissolved in HCl, mixed, and incubated at 4°C for whole night. After filtration, the residue was thoroughly washed and titrated with AgNO₃ solution. Finally, the residue was soaked and titrated with NaOH using phenolphthalein indicator. Cation-exchange capacity (CEC) was calculated using Equation (10):

where A_iand A are the reduced H⁺ content without and with SDF; v refers to the consumed NaOH volume; and m represents SDF weight, g.

Soluble dietary fiber samples were put on a specimen holder, sputter-coated with gold, and scanned with a scanning electron microscope at an accelerating voltage of 15.0 kV.

The molecular weight (Mw) of the SDF was determined using high-performance chromatography (30). The chromatographic column was a TSK-GEL column (8 mm × 300 mm), and the detector was a refractive index detector. The mobile phase was 0.05 M NaCl solution, and the eluted rate was 0.6 ml/min. The SDF sample was formulated in an aqueous solution (5 mg/ml) and filtered through a 0.22-μm filter. A standard dextran curve (dextrans of Mw of 1.27, 11.60, 80.90, 147.60, 273.00, 409.80, and 667.80 kDa) was prepared, and the molecular weight of SDFs was calculated.

The monosaccharide composition of the SDF was determined using high-performance liquid chromatography (HPLC) (31). First, 0.01 g of the SDF was added to trifluoroacetic acid (TFA) (2 M, 2 ml) and hydrolyzed at 100°C for 8 h. After hydrolyzation, TFA was dried, washed with 1 ml of methanol, and then dissolved in 1 ml of distilled water. Next, the reaction solution was derivatized with 0.5 mol/L PMP-methanol solution and 0.3 mol/L NaOH for 1 h at 72°C. After cooling, the reacted product was neutralized with 0.3 mol/L HCl and chloroform. Finally, the mixture was centrifuged at a speed of 4,800 rpm for 10 min. Approximately 1 ml of chloroform was added in the process of absorbing the supernatant, and the procedure was run three times under the same conditions. The resulting supernatant was filtered through a biofilm and injected into the HPLC with 20 μl injection volume of a 0.1 mol/L of acetonitrile and PBS (pH = 6.7) mixture with a ratio of 18:82 at 1 ml/min flow rate. Standard solutions containing Rha, Man, GlcA, GalA, Glc, Gal, Xyl, and Ara were determined, as described earlier.

As described in Table 2, the WHC of PEF-SDF (5.67 ± 0.67 g/g) was significantly higher than that of M-SDF (4.35 ± 0.19 g/g) and E-SDF (3.98 ± 0.29 g/g; p < 0.05). In contrast to WHC, PEF showed no significant advantage in SE and OHC compared to microwaves. The SC of PEF-SDF (6.96 ± 0.88 ml/g) and M-SDF (6.75 ± 0.73 ml/g) was significantly increased (p < 0.05) compared to that of E-SDF (4.57 ± 0.36 ml/g) and PS (3.09 ± 0.29 ml/g; p < 0.05). The OHC of PEF-SDF (3.89 ± 0.41 g/g) and M-SDF (3.57 ± 0.18 g/g) was significantly increased (p < 0.05) compared with that of E-SDF (2.88 ± 0.20 g/g) and PS (1.88 ± 0.19 ml/g). In contrast, a decreasing trend in LGC can be observed from 11.26 ± 0.71% (E-SDF) to 8.18 ± 0.28% (PEF-SDF; p < 0.05). The results showed that PEF-SDF has better gelation properties (32).

Based on the results in Table 2, the EA values were 66.37 ± 1.83, 72.35 ± 1.69, and 79.69 ± 2.36 ml/100 ml for E-SDF, M-SDF, and PEF-SDF, respectively, with a clear increasing trend from E-SDF to PEF-SDF (p < 0.05). Similarly, PEF-SDF showed the largest ES value, which indicates that PEF-SDF can be regarded as a better emulsifier.

The rheological behavior is shown in Figure 2. With a rise in the shear rate, a decrease in viscosity is shown, indicating a sign of shear-thinning behavior, which was speculated to be a pseudoplastic fluid (33). In addition, the initial apparent viscosity of SDFs at a concentration of 100 mg/L was observed to decrease slightly from the raw material PS with an increasing shear rate. It was reported that the apparent viscosity was associated with the molecular chain arrangement, which could be speculated that SDFs decreased the entanglement with increasing shear rate. Moreover, PS appeared to be a tight network that tended to be more stable than SDFs at a high shear rate (34, 35). However, a sharp reduction in the apparent viscosity of PS is shown as the shear rate increased, as a result of disruption of the entanglement by the application of shear to align the molecules in the direction of flow (36).

Figure 3 shows that the peak temperature during E-SDF, M-SDF, and PEF-SDF (141.6, 156.3, and 167.6°C) was higher than that of PS (133.2°C). The difference in peak temperature indicates that more energy is needed to decompose SDF; this finding was in accordance with the SDF from soybean residues (37). The higher peak temperatures of M-SDF and PEF-SDF were due to the abundance of short chains, which could explain that SDF had strong hydrogen bonds that required substantial energy to destroy its crystalline structure (38). In addition, when heated to 240°C, heat absorption and heat release are balanced, suggesting high thermal stability.

As shown in Table 3, SDFs from different extraction methods have a higher GAC value than PS (p < 0.05). Furthermore, PEF-SDF exhibited the highest GAC value, suggesting higher glucose levels (39), resulting in a reduced glycemic reaction. Taken together, the CAC values of E-SDF, M-SDF, and PEF-SDF were significantly increased compared with PS (p < 0.05), consistent with a previous report (40). In different SDFs, the CAC value of SDF belonging to the PEF-SDF was the largest (52.63 ± 2.01%) and that belonging to E-SDF was the smallest (32.94 ± 1.92%) in a simulated small-intestinal environment (pH = 7.0). In addition, the NIAC capacity of PS exhibited lower NO2- scavenging ability than the three types of SDFs (p < 0.05). A similar trend was observed for GAC and CAC, with NIAC increasing significantly from E-SDF to PEF-SDF (p < 0.05).

Table 3 shows the addition of PS or SDFs to cause an inhibitory effect on lipase activity. PS and E-SDF showed no significant difference in the inhibit pancreatic capacity (p > 0.05). In contrast, there were clear differences among E-SDF, M-SDF, and PEF-SDF (p < 0.05), with PEF-SDF having the strongest inhibitory effect.

Compared with PS, the CEC values of SDFs were clearly increased from 0.94 ± 0.02 (E-SDF) to 1.76 ± 0.14 mmol/g (PEF-SDF; p < 0.05; Table 3), which were 1.49 (E-SDF), 2.51 (W-SDF), and 2.79 (PEF-SDF) times larger than PS.

Figures 4a–d show the network structure of PS and SDFs, with significance in the morphologies of PS, E-SDF, M-SDF, and PEF-SDF. PS was compact and unevenly packed with particles of different sizes, and E-SDF exhibited a compact texture coated with a wrinkle along with cracks and holes. However, M-SDF and PEF-SDF had looser and more porous surfaces.

Monosaccharides are assessed and are listed in Table 4. The total sugar content of the PS and SDF samples was more than 95% (w/w), while the SDF samples (65.15% for E-SDF, 74.08% for M-SDF, and 77.26% for PEF-SDF) were dominated by neutral sugars compared with the PS (60.63%). Table 4 summarizes PS and SDF samples, which were mainly composed of rhamnose, glucose, arabinose, xylose, galactose, mannose, and galacturonic acid. The major monosaccharides were glacturonic acid, xylose, and galactose. Although the monosaccharide composition of the SDFs was similar to that of the PS raw material, the contents of arabinose, galactose, glucose, mannose, and xylose increased significantly.

The molecular structure was modeled by using sugar molar ratios. The relatively higher homogalacturonan (HG) in PS suggested that the PS was predominantly composed of HG as a main building block. Compared with PS and E-SDF, M-SDF and PEF-SDF contained more neutral sugars and small proportions of HG regions, indicating less linearity and more branching, suggesting that these were similar to a previous report (41). Correspondingly, a clear decrease in R1 and HG/RG-I and a significant increase in R2 highlight the prevalence of linear segments in the structure of PS and E-SDF, whereas M-SDF and PEF-SDF exhibited higher levels of branching (42). HG/RG-I (the ratio of homogalacturonan/rhamnogalacturonan) was calculated as 1.27, which shows that PS has a homogalacturonan-rich (linear) structure, while the SDF sample showed a rhamnogalacturonan structure. The lower ratio of Rha to GalA (R2) in PS and E-SDF indicated the contribution of RG-I blocks within the polysaccharide backbone, indicating that the samples contained only a small proportion of RG-I segments (43). The ratio of RG-I segments was relatively high (Ara + Gal)/Rha (R3) in PS samples, approximately indicating the larger degree of RG-1 segments with a longer average length of side chains than in SDF samples (44).

Soluble dietary fibers prepared by different extraction methods had different molecular weights. As shown in Table 5, compared with PS, enzymatic, microwave, and PEF treatments significantly decreased the molecular size. PEF-SDF had the smallest average molecular weight, and in addition, the results indicate that SDFs based on PEF treatment exhibited a narrower polydispersity.