Effects of ultrasonic pretreatment on the molecular conformation, gel texture, and rheological properties of oat-whey dual-protein gel induced by lactic acid bacteria fermentation

Abstract

The development of novel fermented nutritional foods based on plant-animal protein combinations remains challenging due to the limited gel- forming ability and structural regulation of dual-protein systems. This study investigated the effects of ultrasound pretreatment (0–600 W) of oat protein-whey protein dispersions before lactic acid bacteria (LAB) fermentation on the gel and structural properties of the resulting dual-protein gels (OWLG). Ultrasound pretreatment regulated protein aggregation and gel network formation in a power-dependent manner. The 300 W pretreatment group exhibited the smallest particle size (137.30 ± 1.46 nm), the shortest gelation time (5.0 ± 0.06 h), and improved texture, water-holding capacity, and rheological behavior. Structural analyses revealed that 300 W ultrasonic pretreatment promoted protein rearrangement, increasing the β-sheet content to 48.36 ± 0.16% and resulting in the most compact and stable gel network. This study provides a promising strategy for regulating cereal-animal protein fermented gels and developing functional fermented nutritional foods.

1. Introduction

Fermented dairy products such as yogurt, pudding, and cheese have become globally popular snack foods due to their distinctive flavors and rich nutritional value. In these matrix systems, protein is at the core of structural assembly; beyond providing essential nutrients, proteins serve as fundamental building blocks that govern gel network formation, textural evolution, and water-holding capacity (Yang et al., 2025). Therefore, improving protein functionality, particularly gel-forming ability, has become a critical strategy for developing novel fermented foods with enhanced quality and nutritional value. With the continuous growth of the global population and increasing demand for sustainable protein sources, the development of plant–animal protein composite systems has become an important research direction in the food industry (Rao et al., 2025). Combining plant proteins with animal proteins can effectively improve nutritional balance, functional properties, and consumer acceptance.

Oat serves as a low-cost, widely cultivated plant-based protein source, offering not only higher protein content than most grains but also excellent digestibility and high biological value (Jennifer et al., 2024). Additionally, the abundant β-glucan content in oat has been shown to positively influence blood cholesterol levels (Ulla et al., 2024). Meanwhile, it offers distinct advantages over legume proteins as a gluten-free, low-allergenicity source, making it an ideal protein source for processing traditional dairy alternatives (Huang et al., 2024). Nevertheless, from the perspective of nutrition, when applied alone, Oat protein (OP) exhibits an imbalanced amino acid profile, making it difficult to meet comprehensive nutritional requirements. To enhance food nutrition and consumer acceptance, combining it with animal protein to develop nutritionally balanced dual-protein fermented gel products is feasible. Whey protein (WP) as a representative animal protein, exhibits excellent solubility, gelling properties, and high nutritional value. In recent years, it has become an ideal partner in plant protein blending systems. Current research primarily focuses on exploring the gelling properties and nutritional value of legume protein–whey protein systems (Wang et al., 2026; Xin et al., 2021), whereas the structural characteristics, intermolecular interactions, and gelation behavior of cereal protein–whey protein systems, especially OP-WP systems, remain insufficiently explored. Therefore, developing and regulating the protein structure of the OP-WP composite gel system is of great significance for clarifying the gelation mechanisms of cereal-dairy protein systems and promoting the high value application of oat protein in fermented foods.

Lactic acid bacteria (LAB) serve as effective acidic coagulants in dairy processing. During fermentation, the gradual release of organic acids lowers the pH toward the protein isoelectric point, reducing electrostatic repulsion among heat-treated protein chains and inducing an ordered three-dimensional gel network (Han et al., 2024). Beyond structural assembly, starter cultures and probiotic strains influence peptide profiles, overall composition, sensory traits, and gut barrier support in fermented matrices (Gabrilyants et al., 2025; Iqbal et al., 2025). The peptides and amino acids hydrolyzed during this process are more readily absorbed by the human body (Luo et al., 2024). In addition, the protein gel induced by LAB fermentation has shown great potential for product flavor, promoting the degradation of allergens, and supporting the repair of the intestinal barrier (Han et al., 2023; Na et al., 2025). Compared to other acidic coagulants, the slower acidification rate of LAB extends the duration of protein interactions, allowing proteins to rearrange more effectively and thereby improving the overall quality of the gel (Chang et al., 2024; Lv et al., 2024; Yang & Li, 2021). Although LAB exhibits significant advantages in individual protein gel systems, research on the preparation process, property regulation, and enhancement mechanisms of the dual-protein gels it induces remains limited, indicating a gap in current studies. Concurrently, researchers have discovered that structural differences between proteins may result in weaker interprotein interactions, leading to poorer texture in the prepared dual-protein gels (Taco, 2019). Therefore, exploring the application of LAB in dual-protein gel foods and its enhancement mechanisms shows great potential.

Ultrasonic pretreatment is an environmentally friendly and highly efficient modification technique that has found extensive application across multiple sectors of the food industry. Through acoustic cavitation, ultrasound generates localized shear forces, turbulence, and high-energy microenvironments, which can disrupt large protein aggregates, promote molecular unfolding, expose buried functional groups, and facilitate new intermolecular interactions (Chen et al., 2020). Research has found that ultrasonic pretreatment can effectively improve the textural properties of protein gels and promote the formation of a tightly ordered gel network (Bai et al., 2025; Zhang, Liu, et al., 2025). However, excessive ultrasonic pretreatment may cause over-disruption of protein structures and induce irregular re-aggregation, suggesting that optimization of ultrasonic conditions is essential for achieving desirable gel properties (Liu et al., 2025). Although ultrasound-assisted modification has been widely investigated in single-protein gels and some plant–animal protein systems (Alisha & Gail, 2025; Zhang et al., 2022; Zhang et al., 2024; Zhao et al., 2024), its influence on cereal protein–whey protein gels induced by LAB fermentation remains unclear. In particular, the relationship between ultrasonic regulation of protein molecular conformation, intermolecular interactions, fermentation behavior, and final gel properties has not been systematically elucidated. Therefore, this study aims to investigate the effects of different ultrasonic power pretreatments on the physicochemical, textural, and rheological properties of OP-WP gel systems prepared by lactic acid bacteria induction. Through Fourier transform infrared spectroscopy (FTIR) analysis, fluorescence spectroscopy, scanning electron microscopy (SEM), and molecular docking techniques, this study will characterize the gel structure, elucidate the gel formation mechanism, and reveal the relationship between protein structural transformation and improvements in gel quality. This study provides theoretical support and reference for the application of novel plant protein-whey protein blends in enhancing the quality of fermented dairy products.

2. Materials and methods

2.1. Materials

Oat protein (OP, protein content ≥80%) powder was obtained from Shanxi Jushengyuan Biotechnology Co., Ltd. (Shanxi, China). Whey protein concentrate (WP, protein content ≥80%) powder was purchased from Germany's WHEYCO Netherlands VOF Food Co., Ltd. (Hamburg, Germany). Sucrose was purchased from Guangxi Laibin Dongtang Phoenix Co., Ltd. (Guangxi, China). Lactic acid bacteria (LAB, MV040) was obtained from Chr. Hansen Trading Co., Ltd. (Beijing, China). All other reagents used were of analytical grade and were purchased from Tianjin Reference Chemistry Reagent Co., Ltd. (Tianjin, China).

2.2. Preparation of ultrasonically pretreated OP-WP protein solution samples

Following preliminary laboratory screening of the gel properties and digestion properties of the dual-protein gel with different OP-WP ratios, the optimal OP-WP ratio was determined to be 1:4. At this ratio, dual-protein gel exhibited favorable gel properties and high protein digestibility. Consequently, the following experiments were conducted using this ratio. The OP-WP mixture was prepared with deionized water at a protein concentration of 12% (w/v) and stirred on a magnetic stirrer for 2 h. Using an ultrasonic cell disruptor (LC-CB-1000E, Li-Chen; Shanghai, China) to treat the composite protein solutions at 100, 200, 300, 400, 500, and 600 W (pulse mode: 2 s on, 2 s off; ultrasonic frequency: 20 kHz) in an ice-water bath for 20 min. The untreated solution served as the control group. After ultrasonication, the composite protein solutions were hydrated overnight at 4 °C in a refrigerator.

2.3. Preparation of OP-WP gel induced by ultrasonic pretreatment of lactic acid bacteria

The ultrasonically treated OP-WP composite protein solution was supplemented with 2.0% (w/v) sucrose to ensure an adequate carbon source for lactic acid bacteria fermentation (Guo et al., 2025). After thorough mixing on a magnetic stirrer, the solution was sterilized in a 95 °C water bath for 15 min. Following cooling to 43 °C, the LAB preparation was added and mixed uniformly. The mixture was then placed in the 43 °C thermostatic incubator (Yiheng Scientific Instruments; Shanghai, China) for 8 h to obtain lactic acid bacteria-induced OP-WP gel samples (OWLG). Subsequently, all samples were stored at 4 °C, and a portion was freeze-dried using a freeze dryer (Alpha 1–2 LDplus, Christ, Germany), then ground and passed through an 80-mesh sieve for later use.

2.4. Molecular docking simulation

The structural information of OP could not be obtained directly from the Protein Data Bank (PDB). Using AlphaFold (https://alphafold.com/), the three-dimensional structure was predicted based on its primary 12 S amino acid sequence (derived from Avena sativa in the UniProt database https://www.uniprot.org/, accession number Q38779). The three-dimensional structural data for WP was obtained from the RCSB PDB database (https://www.rcsb.org/, ID: 3NPO) (Dong et al., 2023). Using GROMACS software, removed water from the three-dimensional structures of OP and WP. Employ the protein preparation module in Maestro 14.0 to predict and correct the protonation states of each residue in both proteins at pH = 4.2 (gelation endpoint) (Laurence et al., 2007). Using 3NPO as the receptor protein and Q38779 as the ligand protein, molecular docking was performed on the corrected protein structures via the GRAMM server. The lowest-energy conformation was selected and visualized using PyMOL 3 software.

2.5. Particle size and polydispersity index (PDI)

The average particle size and PDI value of the gel samples were determined using a method adapted from Zhao et al. (2023). The lyophilized protein gel powder was reconstituted in deionized water to the concentration of 1 mg/mL. The supernatant was filtered through the 0.45 μm membrane and analyzed using a particle size analyzer (Nano ZS, Malvern Instruments, UK) to determine the average particle size and PDI value.

2.6. Sulfhydryl and disulfide bond content

The determination of sulfhydryl and disulfide bond content was performed using a slightly modified method of Qiu et al. (2025). Preparation of Tris-Gly buffer (Buffer A: 0.086 M Tris, 0.09 M Gly, 0.04 M EDTA, pH 8.0) and Tris-Gly-Urea buffer (Buffer B: 0.086 M Tris, 0.09 M Gly, 0.04 M EDTA, 8 M urea, pH 8.0), along with Ellman reagent (4 mg DTNB/mL Buffer A). For each of the 15 mg lyophilized gel samples, dissolved in 5 mL of Solution A or Solution B, add 50 μL of Elman reagent. Thoroughly mixed, then reacted in the dark for 1 h. Centrifuge at 8000 ×g for 10 min at 4 °C. At 412 nm, determined the sample absorbance to calculate the free sulfhydryl content (FSH) and total sulfhydryl content (TSH). Free sulfhydryl, total sulfhydryl, and disulfide bond (S—S) contents were calculated using Eqs. (1), (2), respectively.

$$\mathrm{FSH}\text{content}/\mathrm{TSH}\text{content}\left(\frac{\text{μmol}}{\mathrm{g}}\right)=\frac{{10}^6\times {\mathrm{A}}_{412\mathrm{nm}}\times \mathrm{D}}{1.36\times {10}^4\times \mathrm{C}}$$ (1)

where A_412nm is the absorbance at 412 nm. D is the dilution ratio. C is the original protein concentration (g/mL).

$$\mathrm{S}-\mathrm{S}\text{content}\left(\frac{\text{μmol}}{\mathrm{g}}\right)=\frac{\mathrm{TSH}\text{content}-\mathrm{FSH}\text{content}}{2}$$ (2)

2.7. Surface hydrophobicity (H₀)

Following the method of Li et al. (2025) with minor modifications, the surface hydrophobicity of gel samples was determined using a fluorescent probe technique. Samples were prepared in PBS at concentrations of 0.125, 0.25, 0.5, and 1.0 mg/mL. Each 4 mL sample solution was mixed with 20 μL of 8 mM ANS solution and reacted at room temperature in the dark for 20 min. Fluorescence intensity was recorded using a fluorescence spectrometer (F-4000, HITACHI, Japan) at an excitation wavelength of 390 nm, emission wavelength of 470 nm, and slit width of 5 nm. The slope of the fluorescence intensity versus protein concentration curve was recorded as the surface hydrophobicity of the sample.

2.8. pH and gel time

Following the method of Sun et al. (2022), the pH and gel formation time of the gel samples were determined. Starting from the initiation of fermentation, samples were collected every 2 h and stirred clockwise until fluid. The pH of samples was measured at 0–8 h and 24 h post-fermentation, and the gel formation time when the composite protein gel pH reached the isoelectric point of 4.2 was recorded.

2.9. Texture properties

Following the method of Zeng et al. (2025) with minor modifications, the TA New Plus textural analyzer (ISENSO, USA) was calibrated to a height of 30 mm prior to testing, followed by sensor calibration. The calibrated instrument was then used to perform TPA testing on OWLG samples treated with different ultrasonic power levels. Parameters were set as follows: trigger force 5 ×g, deformation 30%, pre-test speed 2.0 mm/s, test speed 1.0 mm/s, post-test return speed 2.0 mm/s.

2.10. Water-holding capacity (WHC)

Taking 2.0 g of gel into the 50 mL centrifuge tube, centrifuged at 8000 ×g for 10 min. Recorded the gel mass before and after centrifugation, and calculated the gel water-holding capacity using formula (3) (Peng et al., 2025).

$$\mathrm{WHC}\left(\%\right)=\frac{{\mathrm{W}}_1}{\mathrm{W}}\times 100$$ (3)

where W₁ and W are the weights of the protein gels before and after centrifugation, respectively.

2.11. Low-frequency NMR relaxation (LFNR)

Water distribution in OWLG was determined using the low-field nuclear magnetic resonance analyzer (NM120, Niumag, Shanghai, China) (Wang, Yin, Wang and Zhai, 2025). At room temperature, appropriate samples were placed into NMR tubes for sampling. Parameters were set as follows: transverse relaxation time at a resonance frequency of 18 MHz, sampling frequency of 200 kHz, echo frequency of 0.3 ms, with a total of 18,000 repetitions (T₂). Analysis employed the CPMG sequence, and inversion processing was performed using MRI software. Relative water content under different states was obtained through integration.

2.12. Rheological properties

All rheological property tests were evaluated using the rotational rheometer (MCR102, Anton Paar, Germany) equipped with parallel plates (diameter 50 mm, gap 1 mm), unless otherwise specified. Taking 1.0 g of OWLG sample, the sample was placed on the base parallel plate, ensuring the surface was as flat as possible. The test was conducted at 25 °C, and measurements commenced after 5 min rest period to release stress.

2.12.1. Viscosity measurement

Under the temperature of 25 °C and fixed strain of 0.5%, the viscosity variation with shear rate was recorded within the range of 1–100 s-1(An et al., 2024). The flow behavior was fitted using the power-law model Eq. (4).

$$\sigma =\mathrm{{\rm K}}\times {\gamma }^{\mathrm{n}}$$ (4)

where σ, K, and γ represent shear stress (Pa), viscosity coefficient (Pa·s), and shear rate (1/s), respectively, and n is the fluid exponent.

2.12.2. Frequency sweep

Frequency-sweep testing of the viscoelastic properties of gel samples was conducted. The storage modulus (G′) and loss modulus (G″) were measured at the fixed strain of 0.5% within the frequency range of 0.01–10 Hz (Ali et al., 2023).

2.12.3. Strain scanning

Utilizing large-amplitude oscillatory shear (LAOS) testing with the angular frequency fixed at 6.283 rad/s, the strain scan range was set from 0.01 to 1000%, and the frequency was 1 Hz. The variation of G′ and G″ with the scanned strain was recorded (Huang et al., 2025).

2.12.4. Creep recovery

Creep recovery testing involved applying 5 Pa stress within the LVR for 5 min, followed by immediately releasing the stress to allow the gel sample to recover for 5 min (Han et al., 2024). The recovery curve was calculated using Eq. (5).

$$\mathrm{R}\left(\%\right)=\frac{{\gamma }_{\mathrm{{\rm T}}}-{\gamma }_{\mathrm{{\rm P}}}}{{\gamma }_{\mathrm{{\rm T}}}}\times 100$$ (5)

where ${\gamma }_{\mathrm{{\rm T}}}$ is the total strain and ${\gamma }_{\mathrm{{\rm P}}}$ is the recovered strain.

2.13. Fourier infrared spectra

Following the method of Hou et al. (2024), the secondary structure of the freeze-dried protein gel powder samples was determined using an infrared spectrometer (PerkinElmer, Waltham, MA, USA). Fourier transform infrared (FTIR) spectra were acquired with 64 scans at a resolution of 4 cm⁻¹ over the frequency range of 4000-400 cm⁻¹. The relative content of each secondary structure was determined based on peak area.

2.14. Scanning electron microscopy

The microstructure of OWLG was characterized using the S3400 scanning electron microscope (Hitachi, Ltd., Tokyo, Japan). First, fresh composite protein gel samples were cut into 2 × 5 mm strips with a thickness of 1 mm and placed in a 2.5% glutaraldehyde solution for fixation at 4 °C for 5 h. Samples were rinsed three times with 0.1 M phosphate-buffered saline (pH 6.8) for 10 min each. Dehydration was performed sequentially with 50%, 70%, and 90% ethanol for 10 min each, followed by three dehydration steps with 100% ethanol for 10 min each. Subsequently, replaced once with 100% ethanol/tert-butanol (1:1) and twice with pure tert-butanol, each for 15 min. The treated samples were freeze-dried, sputter-coated with gold, and observed at 5 kV voltage to examine the cross-section (Ren et al., 2025).

2.15. Statistical analysis

Each experimental group was replicated three times, with results expressed as “mean±standard deviation”. The data were analyzed using one-way analysis of variance (ANOVA), and Duncan's multiple range test was used to assess significance. All statistical analyses were performed using SPSS 23 software (IBM, Chicago, IL, USA). Differences were considered statistically significant at P < 0.05. Principal component analysis and correlation analysis were performed using Origin 2021 software (OriginLab, Northampton, MA, USA), and statistical graphs were generated.

3. Results and discussions

3.1. Molecular docking

This study initially employed molecular docking techniques to predict potential interactions between OP and WP. Based on the findings, OP and WP molecules were blended and subjected to a series of processing steps to prepare a dual-protein gel (OWLG), thereby establishing a foundation for investigating the effects of ultrasonic pretreatment on its properties. The 12S globin was the primary component of OP. This study predicted the OP structure based on this component and performed docking with WP. The docked model was analyzed using the PDBePISA online protein interaction website (https://www.ebi.ac.uk/msd-srv/prot_int/cgi-bin/piserver) and visualized with PyMOL software. The OWLG preparation process and its molecular docking prediction model were illustrated in Fig. 1. The OP-WP docking model revealed a binding energy of −709.11 kcal/mol, indicating excellent protein-protein interaction and promoting aggregation (Yan et al., 2020). Combined with magnified 3D views, multiple intermolecular forces were observed between OP and WP. The yellow dashed lines in Fig. 1B indicated hydrogen bond interactions, while the pink dashed lines denoted salt bridge interactions. Analysis revealed that Q38779 and 3NPO formed seven hydrogen bonds between amino acid residues such as GLY43 and THR439, ARG148, SER150, and GLU509, forming seven hydrogen bonds. These interprotein hydrogen bonds constituted the fundamental framework of OWLG. Salt bridges were observed between GLU134 on the 3NPO protein and LYS527 on the Q38779 protein, ARG148 and GLU509, and ASP33 and ARG504. Molecular docking models indicated that the relatively low binding energy between OP and WP suggests that the interaction contributes to maintaining gel stability (Wang et al., 2025). Nevertheless, molecular docking inherently represents a static, non-covalent pairwise simulation under idealized conditions, which cannot fully capture the dynamic conformational unfolding, covalent cross-linking, and multi-protein aggregation driven by subsequent ultrasonic, and LAB fermentation. Therefore, to further investigated the specific effects of ultrasonic pretreatment on the OP-WP interaction, and to verify the potential presence of disulfide bonds, as well as other non-covalent interactions in the system (which were not fully predicted by molecular docking), subsequent experiments were conducted in this study.

Fig. 1.

OWLG preparation flowchart and molecular docking prediction model.

3.2. Particle size and PDI

To investigate the effect of different ultrasonication conditions on the aggregation degree of OWLG, the average particle size and PDI value were measured. The average particle size represented the size of the aggregate particles, providing a direct reflection of the protein aggregation state, while the PDI value indicated the uniformity of the particle size distribution within the protein aggregates. Furthermore, studies have shown that protein systems with smaller particle sizes and PDI values may exhibit higher gel strength (Han et al., 2024). The experimental results in Fig. 2A indicated that the average particle size of OWLG first decreased and then increased with increasing ultrasonic power. Compared with the control group, the PDI value of the gel sample after ultrasonic pretreatment was smaller. The smaller particle size and lower PDI value of the aggregates suggested that the structure of the gel sample was more stable at this point. As ultrasonic power increased from 0 W to 300 W, the average particle size of OWLG decreased from 221.33 ± 0.98 nm to 137.30 ± 1.46 nm, with the PDI value decreasing by 19.52%. This reduction was likely attributed to the cavitation and turbulent effect of ultrasonic pretreatment accelerating the depolymerization of unstable macromolecular aggregates (Xin et al., 2023). However, further increases in ultrasonic power did not result in a continued decrease in the average particle size of OWLG. Excessive disruption might lead to over-exposure of active groups, causing protein molecules to re-aggregate (Liu et al., 2025). The study by Zhang et al. (2022) on ultrasonic pretreatment of acid-induced gels from soybean protein isolate also reached similar conclusions. This indicated that under 300 W ultrasonic pretreatment, OWLG aggregates exhibited smaller particle sizes and more stable structural architecture.

Fig. 2.

Effect of different ultrasonic power on the average particle size, PDI value (A), FSH and S—S (B), H0 (C), pH change during OWLG formation and gel formation time (D) of OWLG.

3.3. Sulfhydryl content and H₀

Disulfide bonds and hydrophobic interactions played crucial roles in protein cross-linking and aggregate formation, thereby influencing gel mechanical strengt (Hu et al., 2013; Nicoleta et al., 2012). Fig. 2B showed changes in FSH and S—S content in OWLG under different ultrasonic power treatments. The study found that ultrasonic pretreatment significantly reduced the FSH content in OWLG, though the difference between different ultrasonic power levels was not significant. The S—S bond content initially increased and then decreased with increasing ultrasonic power. Under ultrasonic power of 300 W, the S—S content in the composite protein system reached a maximum of 1.01 ± 0.02 μmol/g. During ultrasonic pretreatment, sulfhydryl groups within protein molecules were exposed. The resulting free groups, combined with H₂O₂ produced during LAB fermentation, accelerated the oxidation of free sulfhydryl groups. Under thermal denaturation conditions, this process promoted the conversion of FSH into S—S bonds (Hu et al., 2013). When ultrasonic power exceeded 300 W, excessive disruptive forces caused existing S—S bonds within the system to break. Simultaneously, it accelerated molecular aggregation, and the formation of macromolecular aggregates re-encapsulated sulfhydryl groups internally, thereby reducing the formation of new S—S bonds. Furthermore, the presence of β-lactoglobulin and α-lactalbumin which constitute the major proportion of whey protein in the system places FSH in a unique position. This was also a key reason why FSH showed no significant effect under ultrasonic power (Jayani et al., 2010). Zhang et al. (2023) also confirmed in their study on ultrasonically assisted soybean protein isolate-fish paste protein gel that ultrasonic pretreatment reduced the sulfhydryl content in the gel system (Zhang et al., 2017) and promoted the formation of S—S bonds. However, Liu et al. (Zhang et al., 2024) reported opposite findings, observing a significant increase in FSH levels in soybean-egg white protein gel under ultrasonication. This discrepancy in results might stem from differences in the protein systems investigated, further corroborating that the unique structure of OP-WP influences sulfhydryl content.

Fig. 2C showed the change in H₀ of the OWLG system under different ultrasonic powers. H₀ exhibited a significant decreasing trend with increasing ultrasonic power. As the ultrasonic power increased from 0 to 300 W, H₀ decreased from 142.6 ± 2.27 to 58.81 ± 0.42. The OWLG exhibited the highest S—S content and lowest H₀ at 300 W. At this power level, ultrasound promoted the dissociation of OP and WP molecules, exposing more active sites and facilitating the occurrence of protein hydrophobic interactions. Concurrently, S—S bond formation promoted protein cross-linking and re-aggregation, trapping hydrophobic binding sites within the gel matrix, leading to a reduction in surface hydrophobicity (Xin et al., 2023). At higher ultrasonic power levels, the tightly bound gel system was disrupted. The weakening of S—S bond interactions re-exposed internal hydrophobic residues, enhancing surface hydrophobicity (Chu et al., 2024; Liu et al., 2021). Thus, both S—S content and H₀ changes illustrated that moderate ultrasonic pretreatment positively influenced the cross-linking of OP and WP as well as the formation of compact aggregates in OWLG.

3.4. pH and gel time

Fig. 2D showed the pH changes during fermentation and the gel formation time at the isoelectric point for OWLG samples treated with different ultrasound powers. All samples had an initial pH around 6.5, which decreased with prolonged fermentation time and eventually stabilized near 4.2. Within the first 2 h of fermentation, the pH change curves for all samples were largely consistent. Between 2 and 6 h, pH rapidly decreased to near the isoelectric point, accompanied by significant differences (P < 0.05). After 8 h, the curves leveled off, reaching the fermentation endpoint. Compared to the control group, ultrasonic pretreatment accelerated the decrease in OWLG pH. Meanwhile, in Fig. 2D, the column chart indicated that ultrasonic pretreatment could significantly shorten the formation time of OWLG gel. Pacheco et al. (Flaviana et al., 2023) reached similar conclusions in their study on ultrasonication-assisted fermentation of buffalo yogurt. The cavitation effect produced by ultrasonic pretreatment disrupted the cellular pore structure and pre-unraveled the spherical chains during heat treatment. This loosening of the structure exposed buried reaction sites and promoted the degradation of the OP-WP system by LAB during subsequent processing (Zhang, Niu, et al., 2025). This process fragmented OP-WP aggregates into smaller monomers, accelerated lactic acid production, rapidly increased H⁺ concentration within the system, and consequently shortened OWLG formation time (Sun et al., 2022). With the increase in the ultrasound power, the formation rate of OWLG first accelerated and then slowed down. At the ultrasound power of 300 W, the time required for OWLG formation was the shortest, at 5.0 ± 0.06 h. LAB was weak in hydrolyzing proteins on its own. After ultrasonic pretreatment, the depolymerization of the OP-WP system enhanced the decomposition efficiency of LAB, hydrolyzing the high molecular weight protein aggregates into monomers or smaller peptide segments and amino acids. This improved the nitrogen source environment required for LAB fermentation, thereby accelerating the fermentation process and enabling the system pH to reach the isoelectric point faster (Wang et al., 2021). When ultrasonic power exceeded 300 W, the gel formation time of OWLG slightly increased. This might be attributed to the re-aggregation of the OP-WP system under excessive sonication, causing the active sites bound to LAB to become re-encapsulated within the system (Han et al., 2024), thereby slowing down the fermentation rate of LAB. In summary, appropriate ultrasonic pretreatment could accelerate OWLG formation and enhance fermentation efficiency.

3.5. Texture properties

Texture Profile Analysis was a commonly used method for evaluating gel texture, reflecting the sensory experience when consuming food. The effects of different ultrasonic power levels on OWLG texture properties are illustrated in Fig. 3. The hardness, elasticity, and chewiness of OWLG significantly increased as ultrasonic power rose from 0 to 300 W, while cohesiveness showed no significant change. Under 300 W ultrasonic pretreatment, OWLG exhibited maximum hardness (21.39 ± 0.25 g), elasticity (0.66 ± 0.01), and chewiness (6.98 ± 0.32). Ultrasonic energy depolymerized macromolecular aggregates, enhanced protein solubility, and strengthened effective interactions between protein molecules and water molecules, thereby improving gel hardness and elasticity. Concurrently, the formation of S—S bonds facilitated tighter cross-linking networks among small protein molecules, ultimately constructing a dense gel structure (Zhang, Zhu, et al., 2025). However, following 400 W ultrasonic pretreatment, all textural evaluation metrics of OWLG exhibited a downward trend, with its mechanical properties weakened. This might be attributed to excessive exposure of hydrophobic groups under prolonged ultrasonication, leading to the formation of the hydration layer on the protein surface. Consequently, the compact intermolecular structure between OP and WP molecules became disordered and loosened, causing protein particles to aggregate and form insoluble aggregates (Sui et al., 2017) (Fig. 2A). The research result by Cui et al. (2020) also confirmed that moderate ultrasonic pretreatment could enhance the improvement of protein gel texture properties. In conclusion, the changes in OWLG texture properties were closely related to the effects of ultrasonic induction on protein molecular conformation and aggregation behavior. Moderate ultrasonic pretreatment could effectively improve the texture properties of the dual-protein gel.

Fig. 3.

Effect of different ultrasonic powers on the texture properties of OWLG.

3.6. Water-holding capacity (WHC)

WHC was an important indicator for evaluating the mechanical properties of gels, reflecting the degree of interaction between proteins and water molecules in OWLG (Liu et al., 2023). As shown in Fig. 4A, the WHC of OWLG was significantly enhanced after ultrasonic pretreatment. Ultrasonication accelerated the gelation of protein molecules while promoting interactions between OP and WP molecules, facilitating the formation of a dense network and enhancing gel hardness (Fig. 3). Consequently, more water molecules were retained within the gel (An et al., 2024). As ultrasonic power increased, the WHC of OWLG rose from 36.74% (0 W) to 54.85% (400 W). The cavitation effect induced by ultrasound promoted protein unfolding, depolymerization of protein aggregates, and reduction in particle size, thereby increasing the contact surface area between protein molecules and water. Concurrently, the decrease in H₀ (Fig. 2C) also indicated the formation of tighter bonds between protein molecules and water molecules. Beyond 400 W, WHC exhibited a decreasing trend. Protein particle aggregation led to the encapsulation of hydrophobic groups, disrupting the continuous gel network and causing water loss (Xin et al., 2025). Notably, although 400 W yielded the numerical maximum in WHC, no statistically significant difference was observed between the 300 W and 400 W groups (P > 0.05). Considering that the 300 W treatment group exhibited superior texture properties, faster gel formation time, and comparable water-holding capacity, the improvement in gel quality was more balanced.

Fig. 4.

Effect of different ultrasonic powers on the WHC (A), water distribution situation (B), relaxation time (C) and peak area of each water (D) of OWLG.

3.7. Water distribution situation

LF-NMR was applied to measure the T₂ relaxation time of water molecule motion, serving as a non-destructive detection technique to characterize water distribution patterns in gel samples (Qian et al., 2023). The T₂ relaxation time depended on the mobility of water within the sample, with shorter T₂ values indicating weaker water mobility. Shorter T₂ values corresponded to lower water activity and fluidity, stronger binding forces, and more intense interactions between water and other components in the sample. Fig. 4B observed that water molecules within OWLG primarily fall into three categories, T₂₁ (0.57–16.30 ms) representing strongly bound water tightly associated with macromolecules, T₂₂ (32.75–305.39 ms) denoting moderately bound water fixed within protein interstices, and T₂₃ (351.12–1629.75 ms), which readily lost from the gel interior. All samples showed no significant changes in the T₂₁and T₂₃peaks, while ultrasonic pretreatment caused significant alterations in the position and area of the T₂₂ peak in OWLG. As ultrasonic power increased, changes in the water relaxation times and peak areas were shown in Fig. 4B and C. Ultrasonic pretreatment significantly shortened the relaxation time of T₂₂ and increased the peak area ratio PT₂₂ of T₂₂. At 300 W ultrasonic power, T₂ exhibited the shortest relaxation time, with the maximum peak area ratio PT₂₂ reaching 79.06% for T₂₂ and the minimum PT₂₃ reaching 1.16% for T23. This indicated that OWLG at this point demonstrated stronger binding affinity for water molecules (Zhao et al., 2024). The migration of water toward shorter relaxation times was attributed to increased gel attraction toward water molecules. Ultrasonication induced denaturation of OP and WP, exposing more water-binding sites and promoting disulfide bond formation (Fig. 2B). This enhanced protein cross-linking and immobilized greater amounts of water within the gel matrix (Ma et al., 2022). However, ultrasound intensity exceeding 300 W led to prolonged T₂₁ and T₂₂ relaxation times, indicating that ultrasound-induced protein overaggregation formed a loose, coarse gel network. This resulted in reduced gel hardness and impaired water retention capacity of the OWLG gel, while increasing water molecule mobility (Bai et al., 2025). Simultaneously, excessive sonication was accompanied by a decrease in PT₂₂ and an increase in PT₂₃, indicating that the free water content within the system rose, making water molecules more prone to loss. This trend was consistent with the changes in gel WHC (Fig. 4A).

3.8. Rheological properties

The rheological properties of gel products played a key role in determining the texture and mouthfeel. Through a series of rheological tests (viscoelastic curves, frequency scans, LAOS, and creep recovery tests), the gel network structure, viscoelasticity, and gel strength were investigated to establish relationships with OWLG structure and gel properties.

3.8.1. Viscosity measurement

Apparent viscosity was an important indicator for measuring gel flow behavior, reflecting the ability of the gel to resist shear forces (Xu et al., 2023). Fig. 5A demonstrated the apparent viscosity of all samples. All OWLG samples exhibited a decreasing viscosity trend with increasing shear rate, displaying typical shear-thinning behavior. The apparent viscosity decreased significantly between 1/s ∼ 10 s⁻¹ exhibiting characteristics of a non-Newtonian fluid, and then stabilized beyond 10 s⁻¹. This might be due to high shear forces disrupting the intertwining between OP and WP, thereby weakening the forces that maintain the stability of the gel network (Li et al., 2025). When shear forces reached a certain threshold, intermolecular rearrangement achieved equilibrium, and the curve tended toward stability (Zhao et al., 2012). Research has found that the initial viscosity of OWLG significantly increased after ultrasonic pretreatment, rising from 37.05 Pa·s (0 W) to 111.95 Pa·s (300 W). This might be related to changes in protein particle size within the OWLG system and the promotion of LAB fermentation. The ultrasonic reduction in OWLG particle size increases soluble substances within the system, thereby expanding the contact area between LAB and proteins in the mixture. This would contribute to protein aggregation under acidic conditions, increasing flow resistance within the system and enhancing gel viscosity (Tan et al., 2021).

Fig. 5.

Effect of various ultrasound powers on the rheological properties of OWLG (Fig. A-D show viscosity curves, frequency scan curves, strain scan curves and critical strain, creep recovery curves, respectively).

3.8.2. Frequency sweep

To further explored the viscoelastic properties of OWLG, frequency sweep tests were conducted. Fig. 5B showed that the variations in G′ and G″ for all samples exhibited frequency dependence, and the gel phase G′ consistently exceeded G″, which was typical of viscoelastic materials (Shen et al., 2017; Wang et al., 2020). Compared to the control group, ultrasonic pretreatment significantly improved the viscoelastic properties of OWLG. Under ultrasonic action, macromolecular particles in the system were dispersed, and the exposure of sulfhydryl and hydrophobic groups promoted the tight binding of OP and WP (Huang et al., 2025). Similar results were previously reported by Iuri and Eric (2025). The 400 W ultrasonication group exhibited the highest G′ and G″ values. However, excessively high ultrasonic power disrupted the dense gel structure, causing a decline in viscoelasticity. This pattern was consistent with the trends observed in texture analysis regarding changes in gel hardness and elasticity.

3.8.3. Strain scanning

Strain scanning characterized the mechanical properties and deformation tolerance limit of the gel through the Linear viscoelastic region (LVR) and critical strain (γc). As observed in Fig. 5C, during the low-strain phase, both G′ and G″ curves of OWLG tended to stabilize, with G′ values exceeding G″ values, reflecting the elastic-dominant characteristics of the gel (Saeed et al., 2018). As strain continued to increase, the G′ curve rapidly declined while the G″ curve decreased more gradually. The two curves converged at the critical strain point, marking the transition of the gel from the LVR into the nonlinear viscoelastic region (NLVR) (Thomas & Natalie, 2019). When strain exceeded the critical threshold, the gel network fractured due to excessive stress, causing G″ to surpass G′. This transition manifested as enhanced viscosity and diminished elasticity. Following ultrasonic pretreatment, the LVR narrowed and the critical strain decreased. This shift could result from cavitation forces promoting denser OWLG structures, which increased gel hardness and brittleness while reducing flexibility, thereby accelerating fracture (Yu et al., 2024). Although ultrasonic pretreatment significantly enhanced the viscoelasticity of OWLG, the resulting higher hardness and brittleness weakened its deformation tolerance limit.

3.8.4. Creep recovery

The creep recovery test was employed to evaluate the deformation recovery performance of OWLG. As the applied stress duration increased, the OWLG creep recovery curve initially exhibited a rapid upward trend. Upon stress removal, the strain decreased and stabilized during the recovery phase. Fig. 5D showed that ultrasonic pretreatment significantly reduced OWLG strain and accelerated recovery rates after stress removal, indicating that ultrasonic pretreatment effectively enhanced the resistance of OWLG to external forces and its deformation recovery capability (Han et al., 2024). The 300 W ultrasonic pretreatment yielded the lowest strain value and highest recovery rate (61.34 ± 1.14%), indicating a tightly structured rigid gel at this condition (Brito-Oliveira et al., 2022). This result was confirmed by the variation pattern of G′ values during frequency scanning. Furthermore, it was demonstrated that moderate ultrasonication could enhance the viscoelasticity, mechanical strength, and deformation recovery capacity of the gel by promoting protein molecular cross-linking.

In summary, rheological evaluations demonstrated that moderate ultrasonic pretreatment markedly enhanced the viscoelasticity, gel strength, and structural recovery of the OWLG matrix. Specifically, the 300 W pretreatment group exhibited superior comprehensive rheological performance. Although the 400 W group displayed slightly higher G′, high ultrasonic intensity induced excessive matrix rigidity and brittleness. This structural alteration was evidenced by a narrower linear viscoelastic region (LVR) and a reduced critical strain, rendering the gel network vulnerable to strain-induced breakdown. In contrast, 300 W pretreatment achieved an optimal balance between mechanical strength and strain tolerance, resulting in the lowest creep deformation and the highest strain recovery rate.

3.9. Secondary structure

In order to investigate the relationship between ultrasonic pretreatment and the structural and gel properties of OWLG, and to further elucidate the underlying mechanism, this study performed deconvolution and fitting of the amide I region of the secondary structure of OWLG using FTIR, calculating the content of each secondary structure. Fig. 6A showed no new absorption peaks in the gel samples after ultrasonic pretreatment, indicating that the treatment did not destroy or form new protein functional groups. However, the absorption peak intensities exhibited changes. The prominent absorption peak at 3000–3500 cm⁻¹ is attributed to O—H stretching vibrations, primarily reflecting the intermolecular hydrogen-bonding environment of the system (Liu et al., 2026). As the ultrasonic power increased, the intensity of the O—H stretching vibration first increased and then decreased, with the strongest absorption peak observed at 300 W. This enhancement effect might stem from the ultrasonic cavitation effect, which increased the exposure of hydrophilic groups and promoted the formation of hydrogen bonds between OP and WP molecules (Xin et al., 2021). Nevertheless, this absorption enhancement diminished at ultrasonic power above 300 W. Excessive ultrasonic energy disrupted pre-established intermolecular hydrogen bonds between macromolecules, aligning with the findings of Chen, Han, et al. (2025). Moreover, the secondary structure components of proteins mainly reside in the amide I band (1700–1600 cm⁻¹). Fig. 6B showed that ultrasonication increased the β-sheet and β-turn contents of OWLG. Simultaneously, the α-helix content decreased from 17.48 ± 0.24% (0 W) to 13.94 ± 0.22% (300 W), accompanied by the significant reduction in random-coil content. The ordered structures characteristic of α-helix and β-sheet contributed to enhanced gel rigidity and stability (Chang et al., 2016). However, the conversion of α-helix to β-sheet might result from intramolecular hydrogen bond breakage during ultrasonic disruption, followed by intermolecular hydrogen bond formation within OP-WP (Xin et al., 2023). The loose molecular chains of random coils were the main factor determining gel deformation. Their reduced content indicated enhanced network density in OWLG, resulting in increased gel hardness and decreased flexibility, consistent with texture analysis (Fig. 3A) and strain scanning (Fig. 5C) results. When ultrasonic power reached a certain threshold, further enhancement accelerated protein-protein collisions, causing proteins to reaggregate into large random-coil particles. This increased random-coil content, heightened OWLG flexibility, and resulted in a more loosely structured gel.

Fig. 6.

Effects of different ultrasound powers on the FTIR (A) and secondary structure content (B) of OWLG.

3.10. Microstructure

The microstructure of OWLG was characterized using scanning electron microscopy at magnifications of 1000×, 2000×, and 5000×. Micrographs of OWLG at different magnifications were shown in Fig. 7. All OWLG samples exhibited a continuous three-dimensional network framework at 1000× magnification. Under the induction of LAB, OP and WP molecules aggregated and cross-linked, transforming the flow-state mixture into a non-flowing gel state (Chen, Li, et al., 2025). At higher magnifications, the pore distribution of OWLG was observed. Compared to ultrasonically treated OWLG, the untreated group exhibited numerous irregular pores on the gel surface, along with dendritic protrusions at crosslinking points. This rough network structure was probably caused by the aggregation of irregular macromolecular particles (Fig. 2A). The OWLG gel surfaces treated at 200 W and 300 W exhibited relatively smaller and more uniformly distributed pores, with the irregular protrusions disappearing and the surface becoming smoother. This resulted from increased S—S and β-sheet content promoting cross-linking between OP and WP molecules, improving gel texture and viscoelasticity, and accelerating the formation of a regular, ordered gel network (Wang, You, et al., 2025). As ultrasonic power further increased, the gel network structure again became loose and irregular, with pore diameters gradually enlarging, resulting in a loose OWLG texture and reduced deformation resistance (Fig. 5D). This occurs because under moderate ultrasonic conditions, large OWLG particle aggregates depolymerize and rearrange to form an ordered structure, whereas excessive disruption causes depolymerized molecules to re-aggregate into large-area irregular structures (Ren et al., 2024). Therefore, moderate ultrasonic pretreatment of OWLG could facilitate the formation of a stable gel framework and enhance its compactness.

Fig. 7.

Microstructures of OWLG treated with different ultrasound powers c at magnifications of 1000×, 2000×, and 5000×, the blue and red circles indicate crosslinking points and pore locations. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3.11. Correlation and principal component analysis

This study performed correlation and principal component analysis on the experimental results to explore the relationship between ultrasonic pretreatment and the conformation, fermentation, and gel properties of OWLG, in order to elucidate the mechanism of the effect of ultrasonic pretreatment on OWLG. In Fig. 8A, redder and flatter ellipses indicated strong positive correlations (p ≤ 0.001), while bluer and flatter ellipses indicated strong negative correlations. The particle size of OWLG aggregates, S—S content, β-sheet content, and gel elasticity exhibited significant negative correlations. Ultrasonic depolymerization reduced protein particle size, enhancing S—S cross-linking between OP and WP to form an ordered gel framework, ultimately improving gel elasticity. Furthermore, β-sheet content in OWLG secondary structure showed significant negative correlations with gel H₀ and gel formation time, while exhibiting positive correlations with gel hardness, elasticity, and water retention. This indicated that OWLG secondary structure plays an absolute role in gel quality. The denser gel structure could encapsulate more hydrophobic binding sites, trapping more water molecules internally and enhancing gel mechanical strength. The weak correlations among other properties were probably attributable to complex intermolecular forces within OWLG and variations in the tightness of the gel's three-dimensional network structure.

Fig. 8.

Pearson correlation analysis (A) and PCA analysis (B) of physicochemical, conformational, and gel properties of OWLG.

The PCA bioplot in Fig. 8B indicated that the first principal component accounted for 78.7% of the variance, with the top two principal components cumulatively explaining 88.3% of the variance. The aggregate particle size, β-sheet content, and S—S bond content were identified as the core factors driving the differences in gel properties, followed by α-helix, random-coil content, and FSH content. All sample groups exhibited spherical distributions across different quadrants, with the moderate ultrasonication group (300 W) significantly diverging from the control group, while samples from lower or higher ultrasonication conditions showed more clustered distributions relative to the control. The OWLG aggregate particle size showed positive correlations with α-helix, random-coil, H₀, PDI, FSH, gel formation time, and critical stress, while exhibiting inverse trends with S—S content, gel mechanical properties, and water retention capacity. The disruptive effect of ultrasonication on aggregates promoted tight protein-protein interactions, accelerated gel formation, reduced flexible structural content, increased ordered rigid structural content, and improved the coarse, loose gel structure, thereby enhancing the quality characteristics of the gel.

4. Conclusion

This study combined OP and WP in specific proportions and prepared OWLG through ultrasonic-assisted LAB induction. This approach enriched the existing variety of dual-protein gels, while also filling a gap in exploring the interaction mechanism between cereal proteins and WP. Experimental results indicated that moderate ultrasonic pretreatment disrupted irregular protein aggregation, reduced aggregate particle size and PDI, fully exposed internal sulfhydryl residues, and promoted S—S bond formation via radical oxidation, thereby enhancing intermolecular cross-linking. Concurrently, exposed hydrophobic groups further strengthened intermolecular hydrophobic interactions. This synergistic effect accelerated gel network formation, shortening fermentation time. Concurrently, the depolymerization effect led to ordered rearrangement of protein molecules. The decrease in α-helix and random-coil content indicated that relatively loose intramolecular structures were replaced by denser ones, enhancing the mechanical strength and viscoelasticity of the gel. However, the study also observed that excessive ultrasonic power during OWLG disruption diminished gel texture improvements, degraded gel quality, and reduced deformation resistance due to increased gel hardness and loss of flexible structures. While this characteristic limited its application in conventional fermented gel foods, it provided a differentiated technical approach for developing specialized foods for dysphagia patients. Pearson correlation analysis and PCA further confirmed that aggregate particle size, β-sheet content, and S—S bond content were the primary factors driving OWLG quality changes. The judicious application of this technology could significantly shorten production cycles for fermented gel foods like yogurt and cheese, optimize product texture, and enhance stability during storage and transportation. Future research should further investigate the optimization of ultrasound processing parameters and fermentation conditions to establish a more comprehensive processing strategy for OWLG. In addition, the nutritional transformation, gastrointestinal digestion behavior, and bioactive compound release mechanisms of OWLG should be systematically explored, with particular attention to protein digestibility, peptide release, and flavor compound formation during fermentation. Moreover, OWLG shows potential as a functional gel matrix for the encapsulation and targeted delivery of bioactive peptides and lipophilic nutrients. Further studies on the synergistic effects of ultrasonication with other modification strategies, as well as the scalability of continuous ultrasonic processing and its applications in plant-based yogurt, fat replacers, and dysphagia-oriented foods, will provide valuable insights for the development of functional fermented food products.

CRediT authorship contribution statement

Shiyu Li: Writing – original draft, Methodology, Data curation. Yuyang Huang: Formal analysis. Mingshou Lv: Software. Bingyu Sun: Project administration, Formal analysis. Ying Zhu: Visualization. Chen Wang: Methodology. Yong Zhou: Validation. Min Qu: Investigation. Xiuqing Zhu: Supervision, Data curation. Yanguo Shi: Formal analysis, Conceptualization. Linlin Liu: Writing – review & editing, Project administration, Conceptualization.

Ethics statement

This article does not contain any studies involving human participants or animals performed by any of the authors.

Funding

This work was supported by the Natural Science Foundation of Heilongjiang Province of China (PL2025C039). The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

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.

Data availability

Data will be made available on request.