Effects of ultrasonic treatment on the structural, functional properties and beany flavor of soy protein isolate: Comparison with traditional thermal treatment

Graphical abstract

Highlights

• •The comprehensive effects of ultrasonic treatment and thermal treatment on soy protein isolate (SPI) structural and functional protein have been compared.
• •Ultrasonic treatment improved the functional properties of SPI better than thermal treatment.
• •The SPI system of ultrasonic treatment reduced hexanal, (E)-2-hexenal, 1-Pentanol and 1-Nonanol beany flavor compound content, and there was no significant difference from the thermal treatment.
• •The correlation between the conformational changes of protein and the functional properties, volatile flavor compounds composition of SPI was analysed.

Abstract

This research explored the influences of ultrasonic and thermal treatments on the structure, functional properties, and beany flavor of soy protein isolate (SPI). In comparison with traditional thermal treatment, ultrasonic treatment effectively induced protein structural unfolding and exposure of hydrophobic groups, which reduced relative content of α-helix, increased relative content of β-turn, β-sheet and random coil, and improved the solubility, emulsifying and foaming properties of SPI. Both treatments significantly decreased the species and contents of flavor compounds, such as hexanal, (E)-2-nonenal, (Z)-2-heptenal and (E)-2-hexenal in SPI. The relative content of hexanal in the major beany flavor compound decreased from 11.69% to 6.13% and 5.99% at 350 W ultrasonic power and 150 s thermal treatment procedure, respectively. After ultrasonic treatment, structural changes in SPI were significantly correlated with functional properties but showed a weak correlation with flavor. Conversely, the opposite trend was observed for thermal treatment. Thus, using ultrasonic treatment to induce and stabilise the denatured state of proteins is feasible to improve the functional properties and beany flavor of SPI.

1. Introduction

SPI is a plant-based protein with high nutritional value that frequently finds use in food processing as a functional food ingredient because of its physiological activity and various functional properties. However, an unpleasant beany flavor in SPI severely constrains its use in food processing [1]. The beany flavor is primarily produced by lipoxygenases (LOX) that catalyze the degradation of polyunsaturated fatty acids to generate hydroperoxyl derivatives, which are further degraded to alcohols, aldehydes, ketones and other volatile compounds with varying odor thresholds [2]. The protein itself has no flavor. However, once formed, the beany flavor tends to tightly bond with protein through non-covalent bond interactions. The strength of the interaction between the two depends on the structure of protein and the characteristic functional groups of the flavor compounds. Thus, any factor that affects the structure of protein can alter the flavor binding ability of protein [3], [4], [5]. Traditional thermal treatment can availably inhibit the activity of LOX and remove the beany flavor in SPI [6]. However, excessive treatment can cause the break or re-bond of disulfide and hydrogen bonds, inducing protein denaturation, degradation and aggregation, causing irreversible variations of protein structure, ultimately reducing functional properties such as solubility and emulsification [7].

As a non-thermal treatment technology, ultrasonic treatment is widely applied in protein modification due to its advantages of environmental friendliness and sustainability. Generally, ultrasonic treatment creates acoustic cavitation and air bubbles in liquid, and this oscillation and collapse of acoustic cavitation bubbles can generate a variety of effects, such as high-speed liquid jets and shock waves [8]. These physical effects can promote protein unfolding, which enhances protein − water interactions, increases functional properties of protein and affects the beany flavor of soy foods [9], [10]. Furthermore, the activity of LOX may be inhibited by cavitation effects and highly reactive free radicals produced by ultrasonic treatment, which reduces the beany flavor compounds, such as aldehydes and alcohols [11]. The study found that changes in soybean protein conformation induced by ultrasound altered the flavor tofu gel, and the free radicals from ultrasonic treatment might enhance oxidation reactions and affect flavor variations [12].

Consequently, this research explored the influences of ultrasonic and thermal treatments on the structure, functional properties and beany flavor for SPI by obtaining experimental data from Fourier transform infrared (FTIR) spectroscopy, intrinsic fluorescence spectroscopy, electronic nose (E-nose) and headspace solid-phase microextraction gas chromatography − mass spectroscopy (HS − SPME − GC − MS). This study can facilitate the development of ultrasonic technology to improve the functional properties and reduce beany flavor of SPI.

2. Materials and methods

2.1. Materials

Defatted soybean flakes were supplied by Shandong Scents Grains & Oil Co., Ltd. (Shandong, China). Sodium dodecyl sulphate (SDS) and dithiothreitol (DTT) were obtained from Biotopped Technology Co., Ltd. (Beijing, China). 1-Anilino-8-naphthalene sulphonate (ANS) was obtained from Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China).

2.2. Preparation of SPI by ultrasonic and thermal treatments

According to the previous extraction method, SPI extraction was modified slightly [13]. NaOH (2 mol/L) was used to adjust the defatted soybean meal distributed in distilled water (1:10, w/v) to pH 7.2. The suspension was centrifuged at 4,000 rpm for 10 min after being magnetically stirred at 120 rpm for 30 min, and using HCl (2 mol/L) to adjust to pH 4.5. The precipitate was collected after centrifugation at 4,000 rpm for 15 min, then washed three times with distilled water. Finally, the NaOH solution was used to adjust to pH 8.0. The protein solution was used for subsequent treatment.

Separate ultrasonic and thermal treatments were applied to the protein extract solution. The ultrasonic treatment used an ultrasonic cell disruptor (LC-JY96-IIN, Shanghai Lichen Bangxi Technology Co., Ltd., Shanghai, China) with the following parameters: ultrasonic frequency of 20 kHz, pulse on and off time of 2 s. Treatment was performed with different ultrasonic powers (200, 250, 350, 450 and 500 W) for 7 min. Thermal treatment was performed in an oil bath at 120 °C (60, 90, 120, 150 and 180 s). Then, all samples underwent degassing treatment at 0.1 MPa for 10 min. The final SPI samples were obtained after spray drying (>90% protein, using the Dumas combustion method to determine), where the inlet air and outlet air temperatures of spray drying were 170 ± 5 °C and 90 ± 5 °C, respectively. The feed rate was8.0 mL/min. After ultrasonic and thermal treatments, SPI samples were named U-200, U-250, U-350, U-450, U-500, T-60, T-90, T-120, T-150 and T-180, respectively. Under the same conditions, untreated SPI was used as a control.

2.3. SDS-polyacrylamide gel electrophoresis (SDS-PAGE)

A separation gel (12%) and stacking gel (5%) were used for SDS-PAGE [14]. The protein solution (0.5 mg/mL) was boiled for 5 min after mixing with 4 × Laemmli sample buffer with and without DTT. Add a 10-µL aliquot of the sample solution to each gel well. And electrophoresis (Mini-Protean Tetra Vertical Electrophoresis Cell, Bio-Rad Laboratories Co., Ltd., Hercules, CA, USA) was performed using a gel voltage of 80 mV for the stacking gel and 120 mV for the separating gel. After electrophoresis, the Coomassie Brilliant Blue R-250 was stained. 25% methanol and 10% acetic acid were decolourised. Afterwards, a gel imaging system (Gel Doc EZ, Bio-Rad Laboratories Co., Ltd., Hercules, CA, USA) was scanned to analyse the image.

2.4. FTIR spectroscopy

An FTIR spectrometer (Nicolet is50, Thermo Fisher Scientific Inc., Waltham, MA, USA) was applied to measure the secondary structures of SPI samples. The SPI powders were compressed into pellets after being blended with potassium bromide (KBr) at 1:100 (w/w). The 400–––4000 cm⁻¹ wavenumber range was scanned for the samples. With a resolution of 4 cm⁻¹, the average number of scans was 32.

2.5. Intrinsic fluorescence spectroscopy

A fluorescence spectrophotometer (RF-6000, Hitachi Co., Ltd., Tokyo, Japan) was used to obtain the fluorescence spectra of SPI samples [15]. Briefly, the SPI concentration was prepared at 0.1 mg/mL. The scanning conditions were set as follows: excitation wavelength 280 nm, emission wavelength range 300–––500 nm, 120 nm/min for scanning speed. The emission and excitation slits were set to 5 nm.

2.6. Particle size and zeta-potential

The method of Ma et al. was used for the measurement of the particle size and zeta-potential [16]. The SPI concentrations were prepared at 0.5 mg/mL. A granularity analyser (Zetasizer Nano ZS90, Malvern Instruments Ltd., Malvern, UK) was utilized to measure the particle size and zeta-potential of SPI.

2.7. Surface hydrophobicity (H₀)

The H₀ of SPI samples was determined according to the method of Wang et al. [17]. The SPI solution was diluted (0.005–0.2%, w/v) with distilled water before loading 20 μL ANS (8 mmol/L) to each 4 mL of diluted solution. The fluorescence intensity was determined using a fluorescence spectrometer (RF-6000, Hitachi Co., Ltd.) at wavelengths of 484 nm (emission) and 365 nm (excitation).

2.8. Solubility

The solubility of SPI samples was measured by Rahman et al. with minor adjustments. Briefly, the SPI concentration was prepared at 10 mg/mL [18]. Afterwards, the SPI solution was separated by centrifugation at 10,000 rpm for 15 min to obtain the supernatant. The equation was applied to calculate solubility: Solubility (%) = (supernatant protein content/initial protein isolate protein content) × 100. The Biuret method was used to determine the protein content of the supernatant, and that of the isolate was measured using the combustion method, using 6.25 as a correction factor.

2.9. Emulsifying properties

The method of Ren et al. was used for the measurement of the emulsifying activity index (EAI) and emulsifying stability index (ESI) with slight modifications [19]. Briefly, soybean oil (5 mL) was loaded into the SPI solution (0.01 g/mL, 20 mL) and homogenised for 2 min at 10,000 rpm using a high-speed homogeniser. Next, immediately (0 min) or 30 min after homogenisation, the collected emulsion (50 μL) was diluted (1:200, v/v) in 0.1% SDS solution. At 500 nm, the spectrophotometer measured the absorbance of the diluted emulsion. (UV-2600, Shimadzu Co., Ltd., Kyoto, Japan). EAI (Eq. (1)) and ESI (Eq. (2)) were determined by the following equations:

$$\mathit{EAI}{m}^2/g=\frac{2\times 2.303\times A_0\times DF}{c\times \theta \times 10,000}$$ (1)

$$\mathit{ESI}\left(\min \right)=\frac{A_0}{A_0-A_{30}}\times 30$$ (2)

where c represents the protein content (0.01 g/mL), DF represents the dilution factor (2 0 0), and A₀ and A₃₀ represent the absorbance values at 0 and 30 min, respectively, θ represents the proportion of oil formed the emulsion (0.2).

2.10. Foaming properties

The method of Cui et al. was used for the measurement of foaming capacity (FC) and foaming stability (FS) [20]. Briefly, distilled water (20 mL) was added to SPI sample (0.5 g) and stirred to dissolve sufficiently. The SPI solution was then homogenised for 3 min at 10,000 rpm before being moved to a graduated cylinder (50 mL). The foam volumes immediately (0 min) (V₀) and 30 min (V₃₀) following the transfer were recorded. FC (Eq. (3) and FS (Eq. (4) were calculated as follows:

$$\mathrm{F}\mathrm{C}(\%)=\frac{V_0}{V}\times 100$$ (3)

$$\mathrm{F}\mathrm{S}(\%)=\frac{V_0-V_{30}}{V_0}\times 100$$ (4)

2.11. E-nose analysis

The volatile compounds in the SPI samples were detected using the E-nose (PEN3, AIRSENSE Co., Ltd., Schwerin, Germany) [21]. Specifically, the SPI sample (2.0 g) was transferred to a vial in order to equilibrate at room temperature for 30 min to enrich the flavor compounds. The air cleaning time was set for 100 s before the sample detection, and the E-nose probe was inserted and remained 60 s. The instrument’s sensors were sensitive to the following types of substances: W1C: aromatic, W3C: ammonia and aromatic, W5C: short-chain alkane aromatic components, W1S: methyl, W2S: alcohols, aldehydes, ketones and ethers, W3S: long-chain alkanes, W5S: nitrogen oxides, W6S: hydrogen, W1W: inorganic sulphides, W2W: organic sulphides.

2.12. Volatile beany flavor compounds analysis

Volatile beany flavor compound profiles of SPI samples were detected utilizing an HS − SPME − GC − MS apparatus (Nexis GC-2030/QP2020NX, Shimadzu Co., Ltd., Kyoto, Japan) [22]. First, 0.34 g the SPI sample was put into a headspace bottle (20 mL), and the equilibration was conducted at 60 °C for 10 min. A 50/30 µm DVB/CAR/PDMS fibre needle (SAAA-SPME-HOLDER-S-TD, ANPEL Laboratory Technologies (Shanghai) Inc., Shanghai, China) was used for the extraction of the volatile compounds. The fibre needle was inserted into the bottle at 60 °C for 40 min and desorbed at 250 °C for 5 min. Volatile compounds were analysed with GC − MS equipped with a DB-5MS capillary column (30 m × 0.25 mm × 0.25 μm, Agilent Technologies Inc., Santa Clara, CA, USA). The initial oven temperature was kept at 40 °C for 3 min before ramping at 10 °C/min to 60 °C, 3 °C/min to 150 °C and lastly 20 °C/min to 250 °C, which was maintained for 5 min. Helium carrier gas flowed at a rate of 1 mL/min. GC − MS data were determined by the National Institute of Standards and Technology (NIST 14L) standard library and quantified by peak areas.

2.13. Statistical analysis

The results of the experiment were presented as mean ± standard deviation (SD), with each experiment being repeated in triplicate. A one-way analysis of variance (ANOVA) was performed using variance test analysis (IBM SPSS 21.0 software, IBM Corp., Chicago, IL, USA), and a statistically significant difference was deemed according to Tukey’s test. The graphs and Pearson correlation analysis data were designed using Origin 2021 software (Origin, Inc., Massachusetts, USA). SIMCA 14.1 software (Umetrics Corp., Umea, Sweden) was applied for principal component analysis (PCA).

3. Results and discussion

3.1. SDS-PAGE analysis

Alterations to the subunit structure of SPI after ultrasonic and thermal treatments were analysed by SDS-PAGE. Glycinin (11S) and β-conglycinin (7S) are mostly made up of SPI, where the 7S fraction includes the α (68 kDa), α′(72 kDa) and β (52 kDa) subunits and the 11S fraction consists of acidic subunit A (35 kDa) and basic subunit B (20 kDa) by a single disulfide bridge [23]. According to Fig. 1A1, under non-reducing conditions, the band intensity corresponding to the α′ and α subunits of 7S was weakened when the ultrasonic power reached 350 W. Cavitation, shear and turbulence effects generated by ultrasonic treatment caused the subunits of protein to depolymerize into smaller units, weakening the band strength of the α′ and α subunits of 7S [24], [25]. Moreover, the band corresponding to acidic subunit A of 11S almost disappeared, and that of basic subunit B of 11S was dramatically enhanced. This indicated that the hydrophobic aggregation of the B subunit was promoted again when the ultrasonic power was ≥ 350 W, thereby enhancing the intensity of the B subunit band [26]. According to Fig. 1A2, the S-S bonds of the protein molecules were destroyed by the addition of DTT, and the many bands corresponding to unknown high-molecular-weight aggregates or oligomers that might be discovered under non-reducing conditions almost disappeared. According to Fig. 1B1, the band intensity of the α, α′ and β subunits in 7S and that of the A and B subunits in 11S of SPI weakened after thermal treatment under non-reducing conditions compared to the control. This might be attributed to the enhanced protein thermal denaturation generated by the continuous high-temperature treatment [27]. According to the reduction profiles (Fig. 1B2), the SPI samples reduced entirely to smaller subunits or monomers, and the subunit bands of SPI after thermal treatment exhibited similar trends to those after ultrasonic treatment. This suggests that although the ultrasonic and thermal treatments might not alter the primary structure of SPI, they could induce a dissociation − aggregation reaction of protein, resulting in the production of soluble or insoluble aggregates. These aggregates were mainly composed of subunits cross-linked by S − S bonds.

Fig. 1.

SDS-PAGE electrophoretic profiles of SPI by ultrasonic (A1, A2) and thermal (B1, B2) treatments.

3.2. Effects on SPI structural properties of ultrasonic and thermal treatments

3.2.1. FTIR analysis

FTIR spectroscopy evaluated the chemical bonds and secondary structures of SPI samples. According to Fig. 2A1 and B1, after ultrasonic and thermal treatments, characteristic bands corresponding to the backbone structures of the control protein were observed the FTIR spectra of SPI samples. The variation of typical absorption peaks at 1635.35 cm⁻¹ (amide I), 1527.03 cm⁻¹ (amide II) and 1236.39 cm⁻¹ (amide III) were corresponded to C O stretching, N − H bending and C − N stretching vibrations, respectively [28]. Amide bands I and II of SPI after ultrasonic and thermal treatments displayed higher absorption peak intensity than that of the control, suggesting that the treatments increased the exposure of non-polar groups (such as hydrophobic groups) or changed the distribution of functional groups and the protein chain structure, thereby the secondary structure of the protein was modified [29]. The following was the correspondence between secondary structure and each sub-peak in the amide I region (1700–1600 cm⁻¹): α-helix (1650–1660 cm⁻¹), β-turn (1660–1700 cm⁻¹), β-sheet (1600–1640 cm⁻¹), random coil (1640–1650 cm⁻¹). The results of calculations for the relative contents of secondary structure in SPI were shown in Fig. 2A2 and B2. Compared to the control, the relative content of the α-helix, β-turn and β-sheet structures of SPI samples exhibited an early increase and subsequent decrease after ultrasonic treatment, nevertheless the random coil structure displayed the opposite trend. This result suggests that medium-power (300–450 W) ultrasonic treatment disrupts the intermolecular interactions of the protein and perturbs the intramolecular hydrogen bonds, resulting in increased disorder in the structures of samples. However, when the ultrasonic power was 500 W, the protein molecule refolded and aggregated, which is consistent with the study of Wang et al [30]. With increasing thermal treatment time, the relative contents of the α-helix and β-turn of SPI samples first increased and then slightly declined. The relative content of β-sheet increased, whereas that of the random coil declined. Thermal treatments cause protein structural unfolding from the natural state, and protein structure refold and aggregate with the increase in treatment time, resulting in the α-helix, β-turn, β-sheet and random coil constantly interconverting [31], [32]. However, when the thermal treatment time was ≥ 120 s, the hydrophobic regions of protein molecule were disrupted, which unbalanced the inter-molecular forces and eventually promoted aggregation of protein, resulting in the increment in the degree of rigidity of protein.

Fig. 2.

FTIR spectra and secondary structure content of SPI by ultrasonic (A1, A2) and thermal (B1, B2) treatments.

3.2.2. Intrinsic fluorescence spectroscopy analysis

The aromatic amino acids interact with the surrounding groups to produce different fluorescent emission frequencies, when the structure of protein molecules is altered [33]. According to Fig. 3A, when the ultrasonic power was ≤ 450 W, the SPI fluorescence intensity was enhanced to a different degree. This data implies that the rapid shear and molecular collision effects of the ultrasonic treatment influence the protein conformation by increasing the exposure of buried chromophores, which enhance the fluorescence intensity of SPI [34]. Notably, the intensity of the fluorescence decreased as the ultrasonic power increased to 500 W, but it was still higher than that of the control. High-power ultrasonic may cause protein aggregation, leading to the re-embedding of protein chromophores. Compared to the control (Fig. 3B), thermal-treated SPI samples indicated an initial enhancement in fluorescence intensity, followed by a decline. In general, the hydrophilic groups of native SPI were outside, the hydrophobic groups were inside, and most of the chromophores were buried in the molecule. This result indicates that the thermal treatment can induce the unfolding of protein molecular and the exposure of more chromophores in the interior of the molecules to the exterior outside, resulting in the enhancing of the fluorescence intensity for SPI. However, extended thermal treatment might cause the unfolded structure of the protein to reassemble and the protein chromophores to re-embed, further reducing the fluorescence intensity.

Fig. 3.

Fluorescence spectra of SPI by ultrasonic (A) and thermal (B) treatments.

3.2.3. Particle size and zeta-potential analysis

Protein aggregation is described by particle sizes and zeta-potential data for molecular protein–protein interactions [35]. Therefore, the particle size and zeta-potential were analysed in order to reveal structural properties of SPI. According to Fig. 4A, in comparison to the control, ultrasonic-treated SPI samples displayed an initial decline in particle size, followed by an increase, and the average particle size of sample U-500 reached 223.23 nm. It may be attributed that the higher-power ultrasonic treatment induced the aggregation of protein molecules and increased particle size [36]. The effect of ultrasonic treatments on the zeta-potential of SPI seemed to be quite weak, however, the absolute zeta-potential value of sample U-500 showed an obvious decrease. This may have been connected to the protein aggregation induced by heat or rapid turbulence generated by severe sonication [34]. According to Fig. 4B, compared to the control, thermal-treated SPI displayed an initial increase in the particle size, followed by a decline. This may have been connected to the thermal treatment process causing the generation of large-molecular-weight protein soluble or insoluble aggregations. Ju et al. made similar observations [37]. The absolute zeta-potential value of samples T-60 and T-90 after thermal treatment were 25.4 and 27.87 mV, respectively, and thus higher than that of the control. This result suggests that short-term thermal treatment may significantly influence the distribution of surface charges of SPI and enhance the stability of system.

Fig. 4.

Particle size and zeta-potential of SPI by ultrasonic (A) and thermal (B) treatments. Significant differences between samples are indicated by different lowercase letters (p < 0.05).

3.2.4. H₀ analysis

The H₀ affects protein conformations through non-covalent bond interactions, thus affecting stability and functionality [38]. According to Fig. 5A, the H₀ of ultrasonic-treated SPI samples increased with ultrasonic power compared to the control. This suggests that ultrasonic treatment causes partial molecular unfolding of the protein and the exposing hydrophobic groups that were previously buried in protein molecules. However, the H₀ of sample U-500 treated with ultrasonic treatment decreased. Excessive ultrasonic treatment led to partial denaturation or aggregation of protein, burying hydrophobic groups, thus reducing the pathway for ANS binding [39], [40]. According to Fig. 5B, the H₀ of SPI samples that underwent thermal treatment (except for sample T-180) was noticeably higher than that of the control, which was in line with the fluorescence spectroscopy analysis. Another explanation was that hydrophobic groups embedded in the inner region of protein became gradually exposing to the surrounding medium. However, extended thermal treatment might cause protein reaggregation and the decrement in the stability of the system.

Fig. 5.

H₀ of SPI by ultrasonic (A) and thermal (B) treatments. Significant differences between samples are indicated by different lowercase letters (p < 0.05).

3.3. Effects on SPI functional properties of ultrasonic and thermal treatments

Solubility is essential for using proteins as effective functional components in food. According to Fig. 6A1, the solubility of SPI samples treated with ultrasound increased significantly compared with the control, from 88.78% to 97.81% (p < 0.05). This may be because appropriate ultrasonic treatment can generate the shear force and cavitation that can loosen the molecular structure of SPI, thus enhancing the protein-water interactions [41], [10]. According to Fig. 6B1, sample T-150 had the highest solubility of 93.41% in comparison with the control, then the solubility of SPI decreased. The main reason may be the redistribution of protein bonding energy after prolonged thermal treatment. The H₀ was enhanced, the -SH groups became oxidised to form S-S, and the protein aggregated thermally or formed new cross-links, resulting in protein denaturation and decreased solubility [42].

Fig. 6.

Effects on SPI functional properties of ultrasonic and thermal treatments. Solubility (A1 and B1), emulsifying properties (A2 and B2), foaming properties (A3 and B3).

One of the crucial functional properties of proteins is their emulsifying properties. Generally, they are described in terms of the EAI and ESI. The EAI was significantly increased (p < 0.05) in U-350 and U-450 compared to control (Fig. 6A2). The reason may be that ultrasonic treatments induce the exposure of more hydrophobic amino acid of SPI, allowing more hydrophobic groups to enter the oil phase and enhance the EAI [43], [44]. This was consistent with the absorption peak intensity changes in the amide bands I and II regions of the FTIR spectra and the surface hydrophobicity data. According to Fig. 6B2, when compared to the control, the EAI of samples T-60, T-90 and T-120 increased significantly (p < 0.05) after thermal treatment. This was because the protein structural unfolding during the thermal treatment increased the surface activity of SPI [45]. Interestingly, the EAI of SPI samples increased with ultrasonic and thermal treatments but the ESI decreased due to the reaggregation of proteins following the unfolding of SPI in an ultrasonic or thermal environment, resulting in the destabilisation of the emulsion [45], [46].

FC and FS characterise the foaming properties of SPI. According to Fig. 6A3, the FC and FS of SPI samples were enhanced first, then stabilised as the ultrasonic power was increased in comparison to the control. This might be because ultrasonic treatment could unfold the tight structure of protein, resulting in more interior groups being exposed to the surrounding medium. This loose structure facilitates the absorption and diffusion at the oil–water interface of the protein molecule as well as formed a viscoelastic film, thus increasing the FC and FS of SPI [9]. According to Fig. 6B3, the SPI T-90, T-120 and T-150 samples showed improved FC and FS compared to the control. The appropriate thermal treatment can affect the changes in protein conformation, increasing the H₀. This may result in increased volumes and stability of foam. Yang et al. [47] also reported similar phenomena.

3.4. Effect of ultrasonic and thermal treatments on the beany flavor profile of SPI

3.4.1. E-nose analysis

The comprehensive odour fingerprint of SPI samples exposed to ultrasonic and thermal treatments was characterised by an E-nose sensor. According to Fig. 7A1 and B1, compared to the control, the response values of sensors W1S, W2S and W6S showed a reduction trend, indicating that the abundances of alcohols, aldehydes, methyl compounds and ketones in SPI samples were reduced by ultrasonic and thermal treatment. According to Fig. 7A2 and B2, the E-nose data were used to perform the PCA, and the total cumulative variance contribution rate of principal component 1 (PC1) and principal component 2 (PC2) was 91.8% (ultrasonic treatment) and 91.1% (thermal treatment), respectively. The viability of the method was demonstrated by a cumulative PC > 85%. Thus, the main information characteristics of SPI samples can be reflected by PC1 and PC2. PC1 variance contribution was greater than PC2 variance contribution, demonstrating that the sample variation increased with the distance of the PC1 axis. There was a more significant change in the data points of SPI samples in the direction of PC1 and PC2 than in the control after ultrasonic and thermal treatments, which can be attributed to the difference between the volatile components in SPI samples. The class distribution positions of ultrasonic-treated samples U-250, U-350 and U-500 were close to each other, and the beany flavor compounds of samples T-120 and T-180 were not significantly different after thermal treatments, indicating that they were similar in flavor.

Fig. 7.

Radar chart of senor responses (A1 and B1) and PCA of E-nose data (A2 and B2) of SPI by ultrasonic and thermal treatments.

3.4.2. Analysis of the volatile beany flavor compounds

HS-SPME-GC–MS was applied to analyse changes in volatile components of SPI samples. Table S1 shows that the 86 volatile compounds detected in SPI samples mainly included aldehydes, alcohols, ketones, alkanes and other classifications. The control SPI had relatively high contents of hexanal, (Z)-2-heptenal, (E)-2-hexenal, 1-octanol, 1-pentanol, 1-nonanol and 1-heptanol. After ultrasonic treatment, the contents of hexanal, nonanal, decanal and 1-pentanol in SPI were significantly decreased, and (Z)-2-heptenal, (E)-2-nonenal, and (E)-2-hexenal have not been detected (Fig. 8A). Hexanal with an unpleasant grassy flavor was the most typical volatile beany flavor, which had an extremely low threshold and could be sensed even at low concentrations [32]. The content of beany flavor compounds in SPI treated with an ultrasonic power of 200 W increased, indicating that low-intensity ultrasonic induction can alter the conformations of protein and the exposure of the interior hydrophobic regions to increase the binding affinity towards beany flavor compounds [2]. Ultrasonic treatment significantly reduce the content of these flavor compounds as the power increases, indicating that ultrasonic cavitation induces changes in SPI conformation, thereby disrupting the interactions of protein with flavor compounds. This was consistent with the study by Mu et al. on the effect of ultrasonic treatment on the flavor of soybean milk [11]. Among them the hexanal decreased from 11.69% in the control to 6.13% in sample U-350, there was no significant difference from the thermal treatment of sample T-150 (Fig. 8B). However, during thermal treatment the nonanal content steadily increased with prolonged thermal treatment time, and the nonanal had a strong fatty and rose-like aroma, which had an effect on the beany flavor of SPI [31]. This further confirmed that ultrasonic and thermal treatment in a certain range were similarly effective in removing volatile beany flavor components in SPI.

Fig. 8.

Changes of relative quantities of the main volatile beany flavor compounds of SPI ultrasonic (A) and thermal (B) treatments.

3.5. Correlation analysis

The correlations between the structure, functional properties and flavor properties of SPI after ultrasonic and thermal treatments were further investigated using Pearson correlation analysis. The α-helix, β-turn, β-sheet, and random coil represent the protein secondary structures, and the changes in content are usually related to the protein functional properties [10]. According to Fig. 9A and B, after ultrasonic treatment, the α-helix, β-turn and β-sheet (highly ordered protein structures were important), the solubility, EAI, FC and FS of SPI had a significant negative correlation. After thermal treatment, the α-helix, β-turn, β-sheet, EAI, FC, and FS of SPI had a significant positive correlation. These results suggested that the shear forces of ultrasonic cavitation probably disrupted the protein–protein interactions and altered its spatial structure, thereby increasing the disorder in the protein structure and improving the functional properties of SPI. However, prolonged thermal treatment encouraged protein aggregation (Fig. 1B1), which improved the EAI, FC and FS and decreased the solubility and ESI of SPI. Therefore, using ultrasonic treatment to induce and stabilise the denatured state of proteins appears feasible to improve the functional properties of SPI. Compared to the ultrasonic treatment, after thermal treatment, the α-helix, β-turn, and β-sheet were highly negatively associated with the abundances of alcohols, aldehydes and ketones (i.e., the W2S response) (>0.7), whereas the relative content of the random coil had a positive correlation (>0.7). This may be because the prolonged thermal treatment was accompanied by a series of variations that destroyed the molecular structure of the protein, resulting in greater protein denaturation [4]. Overall, due to the unique mechanism of action of ultrasonic treatment, selecting this treatment method to increase the functional properties and beany flavor of SPI was feasible.

Fig. 9.

Correlating structural, functional properties and beany flavor of SPI after ultrasonic (A) and thermal (B) treatments. In the Figure visualises the correlation between ellipse size and color, with blue indicating negative correlation and red indicating positive correlation.

4. Conclusion

This present study compared the structure, functional and flavor changes of SPI during various ultrasonic and thermal treatments. Compared with traditional thermal treatment, medium-power ultrasonic treatment caused structural unfolding and conformational changes of SPI molecules, resulting in reducing the α-helix relative contents, increasing the β-turn, β-sheet, and random coil relative contents, and enhancing the fluorescence intensity and H₀. These variations in structural characteristics further improved the functional properties of SPI, including solubility, EAI, FC and FS. E-nose and HS-SPME-GC–MS analysis demonstrated that the species and contents of flavor compounds in SPI decreased after thermal and ultrasonic treatments, however, the contents of hexanal and 1-pentanol significantly reduced. The (E)-2-nonenal, (Z)-2-heptenal, (E)-2-hexenal, and 1-nonanol have not been detected. Correlation analysis indicated that structural changes in the protein affected its functional properties and the volatile beany flavor compound composition of SPI. It can be observed that ultrasonic and thermal treatments have the potential to influence the redistribution of secondary and tertiary protein structures, consequently impacting their functional properties. In particular, the functional properties of SPI might be efficiently improved by ultrasonic treatment, and the effect of removing the beany flavor was similar to that of thermal treatment. Compared with traditional thermal treatment, ultrasonic treatment, as an innovative food processing technology, has been shown to be effective in improving the functional properties and beany flavor of SPI.

CRediT authorship contribution statement

Yue Kong: Data curation, Writing – original draft. Lina Sun: Writing – review & editing, Resources. Zenan Wu: Software. Yanhui Li: Validation. Zimeng Kang: Formal analysis. Fengying Xie: Conceptualization, Supervision, Funding acquisition, Project administration. Dianyu Yu: Supervision.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: