**3. Results**

#### *3.1. Encapsulation E*ffi*ciency*

Encapsulation e fficiency of the nanoparticles is shown in Figure 1. The encapsulation e fficiencies of all samples were greater than 70%, and the SIF concentration at 2.4 mg/mL had the highest encapsulation efficiency (89%). The encapsulation e fficiency decreased as SIF content increased, suggesting that a portion of SIF were not embedded into the polymerized goa<sup>t</sup> milk whey protein matrix at higher SIF concentration. Similar findings were reported by Patel et al. [21] who found that as curcumin concentration increased, encapsulation e fficiency of zein-curcumin decreased.

**Figure 1.** Effect of the soy isoflavone (SIF) contents on encapsulation e fficiency of polymerized goa<sup>t</sup> milk whey protein (PGWP)-SIF. Di fferent subscript letters indicate a significant di fference (*P* < 0.05). Error bars represent standard deviation of the means.

#### *3.2. Particle Size and Zeta Potential*

Particle size of PGWP and PGWP-SIF are shown in Figure 2A. All PGWP-SIF samples showed larger particle sizes than that of PGWP. This can be attributed to the fact that the PGWP had more hydrophobic groups, and SIF were entrapped in or absorbed on protein to form compact nanoparticles [22]. The particle size of nanoparticles was dependent on the concentrations of SIF in PGWP-SIF. At low SIF concentrations (at 2.1 and 2.4 mg/mL), there was no significant difference (*P* > 0.05) in particle size, which may be because the majority of SIF were entered into the hydrophobic core of protein. However, at higher SIF concentrations (the SIF concentration at 2.7 and 3.0 mg/mL), a significant (*P* < 0.05) increase in particle size was found, which suggested that more SIF were absorbed at the surface of PGWP until the surface was saturated. The phenomenon was similar to previous results reported by Rodríguez et al. [23], where at low green tea polyphenols contents, the particle size was maintained that of the β-lactoglobulin, and at large green tea polyphenols contents, the particle size was increased with the increasing of green tea polyphenol concentration.

**Figure 2.** Effects of the SIF contents on particle size (**A**) and zeta potential (**B**) of PGWP-SIF. Different subscript letters indicate a significant difference (*P* < 0.05). Error bars represent standard deviation of the means.

Zeta potential is related to the stability of the systems. Values of zeta potential below –30 mV indicates a stable solution, which may be due to strong electrostatic repulsion [24]. Values of zeta potential in PGWP and PGWP-SIF are shown in Figure 2B. The PGWP-SIF showed lower value of zeta potential than PGWP (*P* < 0.05), suggesting that SIF adhered to PGWP and reduced the negative charge of PGWP. A similar tendency was reported by Von Staszewski et al. [25], who observed that the addition of green tea polyphenols resulted in a decrease in zeta potential values of the β-lactoglobulin-green tea polyphenols complexes. Increasing SIF content from 2.1 mg/mL to 3.0 mg/mL showed an increase of negative charge among PGWP-SIF samples, which indicated that the large number of the embedded SIF entrapped in the core had not affected the surface charge of protein [11].

#### *3.3. Rheological Properties*

The rheological properties of PGWP and PGWP-SIF are shown in Figure 3. All samples showed shear-thinning behavior (Figure 3A). This may be due to that the interaction between particles were decreased at high shear rates, causing a decrease in the size of dispersions, and thus leading to a decrease in the viscosity [26]. All PGWP-SIF showed significantly higher (*P* < 0.05) viscosity at 200 s<sup>−</sup><sup>1</sup> compared with the PGWP (Figure 3B). The increase in viscosity of the nanoparticles can be attributed to the entrapment of SIF into the PGWP networks. Values of viscosity at 200 s<sup>−</sup><sup>1</sup> for PGWP-SIF were increased (*P* < 0.05) as SIF content increased from 2.1 mg/mL to 3.0 mg/mL.

**Figure 3.** Effects of the SIF contents on flow behavior (**A**) and viscosity at 200 s<sup>−</sup><sup>1</sup> (**B**) of PGWP-SIF. Different subscript letters indicate a significant difference (*P* < 0.05). Error bars represent standard deviation of the means.

#### *3.4. FT-IR Spectra*

FT-IR analysis was used to determine the structural changes of PGWP when interacted with SIF. The amide I band 1600–1700 cm<sup>−</sup><sup>1</sup> (C=O stretch), amide II bands 1500–1600 cm<sup>−</sup><sup>1</sup> (C-N stretching combined with N-H bending modes), and amide A band at 3300 cm<sup>−</sup><sup>1</sup> (N-H stretching and hydrogen bonds) were used to explore the changes of the secondary structure of goa<sup>t</sup> milk whey protein in PGWP-SIF [27]. FT-IR spectra of all samples were shown in Figure 4A. FT-IR spectra of the SIF showed characteristic peaks at 1623.77, 1516.31, and 3354.15 cm<sup>−</sup>1. The PGWP showed bands at 1655.03, 1541.30, and 3303.87 cm<sup>−</sup>1, which may belong to amide I, amide II, and amide A. After interacting with SIF, a hypsochromic shift occurred for the absorption band of PGWP-SIF. The peak in PGWP at 1655.03 cm<sup>−</sup><sup>1</sup> was shifted to 1655.15 cm<sup>−</sup><sup>1</sup> for PGWP-SIF, indicating that the addition of SIF affected the formation of the PGWP-SIF through the interaction related to C=O between PGWP and SIF. The peak at 1541.30 cm<sup>−</sup><sup>1</sup> (PGWP) shifted to 1541.95 cm<sup>−</sup><sup>1</sup> (PGWP-SIF) and may be due to the interactions between PGWP and SIF through C-N and N-H groups. The peak at 3303.87 cm<sup>−</sup><sup>1</sup> (PGWP) shifted to 3308.22 cm<sup>−</sup><sup>1</sup> (PGWP-SIF) and indicated a formation of hydrogen bonds between PGWP and SIF, suggesting that phenolic hydroxyl groups were involved in the non-covalent interaction between PGWP and SIF.

**Figure 4.** FT-IR spectra of SIF, PGWP, and PGWP-SIF (**A**) and the contents of secondary structures of PGWP and PGWP-SIF (**B**). Different subscript letters indicate a significant difference (*P* < 0.05). Error bars represent standard deviation of the means.

Contents of the secondary structure of proteins were calculated by using Amide I band. The results are shown in Figure 4B. When compared with PGWP, content of α-helix and β-sheet in PGWP-SIF decreased from 26.20 ± 0.07% to 23.39 ± 0.42%, and 39.32 ± 0.06% to 36.36 ± 0.25%, respectively. Results indicated that secondary structures of goa<sup>t</sup> milk whey protein were transformed from ordered to disordered when combined with SIF.

#### *3.5. Fluorescence Spectra*

#### 3.5.1. Inherent Fluorescence

Fluorescence intensity of PGWP and PGWP-SIF are shown in Figure 5. The PGWP-SIF showed lower fluorescence intensity than PGWP, indicating that SIF quenched the intrinsic fluorescence of PGWP. The fluorescence intensity decreased as SIF contents increased. Additionally, in the presence of SIF, the maximum peak position slightly shifted to a larger wavelength from 335 nm to 339 nm, which indicated that more fluorophores were exposed to the solvent, and thus implying a change in the polarity of tryptophan residues. The results suggested that some hydrophobic residues were buried during the interaction of SIF with PGWP, leading to the changes of structure in PGWP, and eventually improving the hydrophilicity of the medium.

**Figure 5.** Fluorescence emission spectra of PGWP-SIF systems in 10 mM PB pH 7.4. a–e, PGWP, PGWP-SIF-A, PGWP-SIF-B, PGWP-SIF-C, PGWP-SIF-D. The Stern–Volmer plots for the quenching of PGWP by SIF at di fferent temperatures.

#### 3.5.2. Stern–Volmer Analysis of Quenching Data

To further understand the interactions between PGWP and SIF, the Stem-Volmer equation was used to analyze the data at temperatures of 298, 303, and 308 K, as follows in Equation (2):.

$$\text{F}\_0\text{F} = 1 + \text{Ksv} \text{ [L]} = 1 + \text{Kq} \text{ }\tau\_0 \text{ [L]} \tag{2}$$

where F0 and F are fluorescence intensities of PGWP and PGWP-SIF, respectively; [L] is the concentration of SIF; Ksv is the Stern–Volmer quenching constant; Kq is the biological macromolecules quenching constant; and τ0 is the average lifetime of the biomolecule without quencher (<sup>τ</sup>0 = 10−<sup>8</sup> s) [28].

It was reported that Kq larger than 2.0 × 10<sup>10</sup> L/mol/s indicated static quenching interaction, while q smaller than 2.0 × 10<sup>10</sup> L/mol/s indicated dynamic quenching interaction. Additionally, for static quenching, higher temperature indicated lower quenching constant, while dynamic quenching had the opposite trend [29]. From Table 1, it can be seen that values of Kq were greater than 2.0 × 10<sup>10</sup> L/mol/s, and the Ksv decreased at higher temperatures, which suggested that the interaction was static quenching.

**Table 1.** Relevant parameters of the Stern–Volmer quenching constant (Ksv), the biological macromolecules quenching constant (Kq), the binding constant (Ka), and the number of binding sites (n) were calculated from Stern–Volmer and double log plots at different temperatures. Thermodynamic parameters of enthalpy (ΔH), entropy (ΔS) and free energy (ΔG) changed based on the van't Hoff equation.


For static quenching, the binding constant (Ka) and the number of binding sites (n) conformed to the Equation (3) [30]. The slope and the intercept values of a plot of log [(F0 − F)/F] versus log [L] give n and Ka values, respectively.

$$\log\left( (\mathbf{F}\_0 - \mathbf{F}) \mathbf{\hat{F}} \right) = \log \mathbf{K} \mathbf{a} + \mathbf{n} \log \left[ \mathbf{L} \right] \tag{3}$$

Values of n and Ka parameters are shown in Table 1. Values of n were close to 1, indicating that PGWP had only one binding site for SIF. On the other hand, values of Ka increased as temperature increased, indicating endothermic binding reaction. The result indicated that the stability of PGWP-SIF increased as temperature increased. Our results were similar with findings of Jia et al. [19], who reported that the interaction of β-lactoglobulin with epigallocatechin-3-gallate was endothermic and the stability increased as temperature increased.

#### 3.5.3. Thermodynamic Parameters

The four main interaction forces between polyphenols and proteins are hydrogen bonds, electrostatic forces, hydrophobic forces, and van der Waals forces [31]. The main interaction forces can be obtained by calculating the thermodynamic parameters using van't Hoff Equations (4) and (5), as follows:

$$
\ln \text{Ka} = -\Delta \text{H/RT} + \Delta \text{S/R} \tag{4}
$$

$$
\Delta \mathbf{G} = \Delta \mathbf{H} - \mathbf{T} \Delta \mathbf{S} \tag{5}
$$

where ΔH is the enthalpy change, ΔS is the entropy change, ΔG is free energy change, R is the gas constant (8.314 J/mol/K), T is the reaction temperature, and Ka is the binding constant at the temperatures of 298, 303, and 308 K. ΔH, ΔS, and ΔG could be acquired by Equations (3) and (4) [32].

Ross and Subramanian [33] reported that ΔS and ΔH > 0 indicated hydrophobic forces, ΔS and ΔH < 0 indicated van der Waals forces, and ΔH < 0 and ΔS > 0 indicated electrostatic forces. From Table 1, it can be seen that ΔG < 0, indicating that the interaction between PGWP and SIF was spontaneous [34]. Both ΔH and ΔS values > 0, indicating that hydrophobic interaction was involved in the interaction between PGWP and SIF. This was similar to the findings of Xu et al. [35], who reported that the interaction between β-lactoglobulin and theaflavin/chlorogenic acid/delphinidin-3-O-glucoside was mainly maintained by hydrophobic forces.

#### 3.5.4. Synchronous Fluorescence Spectra

Synchronous fluorescence analysis was used to study effects of SIF on structure of PGWP and obtain information about tyrosine residues or tryptophan residues at synchronous spectrum performed with Δλ at 15 or 60 nm, respectively [19,35]. Some doubts about the effectiveness of the method were recently reported by Bobone et al. [36]. From Figure 6, it can be seen that as SIF content increased, the fluorescence intensity decreased in both fluorescence spectra, which indicated that the binding of SIF with PGWP exposed more chromophores into the solvent and led to a decrease in the fluorescence intensity.

**Figure 6.** Synchronous fluorescence spectra of PGWP-SIF systems, a–e, PGWP, PGWP-SIF-A, PGWP-SIF-B, PGWP-SIF-C, PGWP-SIF-D. (**A**: Δλ = 15 nm, **B**: Δλ = 60 nm).

#### *3.6. Di*ff*erential Scanning Calorimetry (DSC)*

DSC analysis was used to investigate the thermal property of nanoparticles. DSC curves of all samples were shown in Figure 7. The pure SIF showed four endothermic peaks at approximately 85.17, 136.85, 178.67, and 209.83 ◦C. The first three peaks may be attributed to three molecular forms in SIF, including genistein, daidzein, and glycitein [4]. The PGWP exhibited two endothermic peaks at 87.67 ◦C and 155.17 ◦C, and the PGWP-SIF showed two endothermic peaks at 105.37 ◦C and 157.04 ◦C. The result indicated that the thermal stability of the PGWP-SIF improved. Moreover, the characteristic peaks of SIF disappeared in PGWP-SIF, which may be because SIF were encapsulated into PGWP microspheres. Yang et al. [27] also reported that the disappearance of endothermic peaks of pyrogallic acid in the nanoparticle was due to the covalent interactions between pumpkin seed protein isolate and pyrogallic acid.
