pH-shifted walnut protein-bamboo shoot soluble dietary fiber complexes for structuring glucono-δ-lactone-acidified emulsion gels

Abstract

This study developed high-oil (70% v/v) cold-set emulsion gels using pH-shifted walnut protein isolate (PWPI) and bamboo shoot soluble dietary fiber (SDF). pH-shifting significantly improved the solubility of walnut protein from 20.84% to 77.15% and promoted the formation of smaller, more stable protein particles. After SDF incorporation, the resulting PWPI-SDF complexes showed improved colloidal and interfacial properties for stabilizing oil-in-water emulsions. At 0.5% (w/v) SDF, the emulsions exhibited a more uniform droplet distribution and stronger viscoelastic behavior. GDL (Glucono-δ-lactone) was then used to induce cold-set gelation of the emulsions. Among the tested formulations, the gels prepared with 1.5% GDL showed higher β-sheet content, enhanced hydrogen-bonding interactions, improved elasticity, and better textural properties. LF-NMR and MRI further indicated greater water immobilization and a more uniform gel structure, whereas excessive acidification at 2.0% GDL was associated with partial network disruption and water syneresis. Overall, these results indicate that the combined use of pH-shifting and SDF incorporation improved the stability of high-oil emulsions and facilitated the formation of cold-set emulsion gels, providing a useful approach for the valorization of bamboo shoot by-products in plant-based food structuring.

1. Introduction

Emulsion gels have emerged as candidates for fat mimetics and functional ingredient delivery systems due to their dual characteristics of emulsion functionality and structural integrity of gels (Li et al., 2023; Wang et al., 2026). Driven by the demand for healthier, plant-based and clean-label products, naturally derived proteins and polysaccharides are increasingly utilized as synergistic structurants. While proteins provide viscoelastic interfacial films to prevent coalescence (O'Sullivan et al., 2016), polysaccharides enhance the system's stability through continuous phase thickening and specific molecular interactions, such as electrostatic forces, hydrogen bonding, and hydrophobic associations (Asyrul-Izhar et al., 2023). Synergistic protein-polysaccharide complexes are highly effective in achieving superior rheological performance and droplet immobilization capacity.

Increasing the use of plant proteins offers both nutritional and environmental advantages by providing essential amino acids and reducing dependence on animal-derived sources (Malila et al., 2024). Walnut protein isolate (WPI) is a high-quality plant protein source with a balanced amino acid profile, high digestibility, and good biological value (Cai et al., 2018; Wen et al., 2023). However, compared with plant proteins such as soy and pea proteins that have been more widely used in emulsion and gel systems, walnut protein remains underutilized in structured food applications. At the same time, its high glutelin content and pronounced hydrophobicity also result in poor solubility and limited interfacial functionality in the native state, which restrict its direct application in food processing (Akharume et al., 2021). pH-shifting treatment has been recognized as a simple and effective strategy to modulate protein conformation by transiently exposing proteins to extreme pH, thereby unfolding their tertiary structures and improving solubility, surface charge, and interfacial activity after neutralization (Yang et al., 2025). While pH-shifting improves solubility, the unfolded protein chains are susceptible to excessive aggregation. Herein, bamboo shoot SDF was introduced in this study as a structural modulator (Spiess et al., 2025; Xiong et al., 2023). Bamboo shoot processing yields large amounts of underutilized SDF, offering significant potential for valorization (Zhang et al., 2024). SDF derived from bamboo shoots, primarily composed of monosaccharides such as arabinose, galactose, glucose, and uronic acids, has good water-retention ability, oil-binding capacity, and adsorption properties, and may interact with proteins through hydrogen bonding, electrostatic interactions, and steric effects (Wang et al., 2025). However, for the fabrication of desired soft solids, cold-set gelation is often preferred over thermal methods to preserve nutrient integrity and reduce energy consumption. Cold-set gels are typically fabricated via a two-step approach, starting with the formation of structurally activated protein, followed by post-cooling gelation induced by acidification, ionic bridging, or enzymatic crosslinking (Dickinson, 2012). GDL is an effective acidulant for this process, as its gradual hydrolysis facilitates a homogeneous pH reduction, enabling controlled protein-polysaccharide co-assembly into cohesive gel networks (Li et al., 2020).

Despite the potential of WPI-polysaccharide complexes, systematic insights into the synergistic mechanisms between pH-shifted walnut protein and bamboo shoot SDF in stabilizing high-oil (70% v/v) emulsions and their subsequent acid-induced gelation remain limited. Therefore, this study aims to elucidate the roles of PWPI-SDF complexes in stabilizing emulsions and constructing structurally robust cold-set emulsion gels. This work provides insights into the high-value utilization of bamboo shoot by-products and the development of novel plant-based soft solids.

2. Materials and methods

2.1. Raw materials and reagents

Bamboo shoot by-products (Chimonobambusa quadrangularis), used as the source for soluble dietary fiber, were provided by Hongchishui Group Limited (Guizhou, China). Walnut protein isolate (WPI) with a protein content of ≥98% was acquired from Xi'an Lvteng Biotechnology Company (Xi'an, China). Standard monosaccharides were obtained from Sigma-Aldrich (St. Louis, MO, USA), including glucose, galactose, arabinose, rhamnose, fucose, xylose, mannose, fructose, ribose, as well as glucuronic, galacturonic, mannuronic, and guluronic acids. All other chemicals used were of analytical purity, and solutions were prepared using deionized water.

2.2. Extraction of SDF from bamboo shoots

SDF was derived from bamboo shoot by-products using an alkaline hydrogen peroxide method combined with ultrasonic assistance with slight modifications (Zhang et al., 2020). Bamboo shoot by-products were washed, cut, dried at 45 °C, milled, and passed through a 100-mesh sieve to obtain powder. The powder was mixed with 2% (w/v) hydrogen peroxide solution at a 1:20 ratio, after which the pH was adjusted to 11 with 1 M NaOH. The suspension was maintained at 60 °C in a water bath for 2 h and subsequently sonicated for 40 min at the same temperature. After neutralization and centrifugation (8000 rpm, 10 min), the supernatant was concentrated to one-third of its initial volume by rotary evaporation, and four volumes of 95% ethanol were added to promote polysaccharide precipitation at 4 °C overnight. The ethanol was then removed, and the precipitate was collected and freeze-dried to obtain the SDF.

2.3. Characterization of SDF

2.3.1. Chemical composition of SDF

The carbazole‑sulfuric acid method was applied to determine the uronic acid content of the SDF, and the absorbance was measured at 530 nm with galacturonic acid serving as the calibration standard (Wang et al., 2021). The total sugar was quantified via the phenol‑sulfuric acid method, with absorbance measured at 490 nm against glucose after incubation in darkness (Fang et al., 2022). Protein content in the SDF was determined by means of a BCA protein assay kit (Solarbio, Beijing, China), strictly adhering to the manufacturer's protocol.

2.3.2. Monosaccharide composition

The monosaccharide profile of SDF was determined based on the high-performance anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD, ICS5000+, Thermo Scientific, USA) on a CarboPac PA20 column (3 × 150 mm), adapted from the method by Wu et al. (2024). Approximately 5 mg of SDF was subjected to hydrolysis in 2 M trifluoroacetic acid at 121 °C for 2 h, dried under nitrogen, washed with methanol, and re-dissolved in deionized water. Chromatographic separation was carried out using a stepwise gradient consisting of H₂O, 0.1 M NaOH, and a mixture of 0.1 M NaOH/0.2 M NaAc at a flow rate of 0.5 mL/min. Standard monosaccharides (Man-UA, Glc-UA, Gul-UA, Gal-UA, Rib, Fru, Man, Xyl, Glc, Gal, Ara, Rha, Fuc) were utilized as external standards.

2.3.3. Molecular weight distribution

The molecular weight distribution of SDF was analyzed by SEC on an LC-20AT HPLC system (Shimadzu, Japan) equipped with an SB-806 HQ column (Shodex). Separation was performed using 0.2 mol/L NaCl containing 0.03% NaN₃ as the mobile phase at 0.5 mL/min and 25 °C. Molecular weight was estimated using a dextran calibration curve (Gu et al., 2023).

2.4. Preparation of PWPI-SDF complexes

A 5% (w/v) WPI solution was dissolved in deionized water and adjusted to pH 12.0 with 3 M NaOH, stirred for 2 h at 25 °C, and subsequently neutralized to 7.0 (PWPI) to induce structural unfolding based on previous studies on pH-shifting modification of plant proteins (Dai et al., 2023). On this basis, SDF was incorporated into the PWPI solution at final SDF levels of 0.25, 0.5, 0.75, and 1.0% (w/v). The mixtures were thoroughly mixed for an additional 2 h and then kept overnight at 4 °C to form PWPI-SDF complexes. Accordingly, the experimental groups were designated as WPI (native protein), PWPI (pH-shifted protein, without SDF), and PWPI-S0.25, PWPI-S0.5, PWPI-S0.75, PWPI-S1 (pH-shifted protein with different SDF levels).

2.5. Properties of complexes

2.5.1. Solubility

A 1 mg/mL dispersion was centrifuged at 10,000 rpm for 15 min, after which the supernatant was collected for soluble-protein determination. Protein concentration (Ws) was quantified using a BCA assay kit (Solarbio, Beijing, China). Solubility was expressed as the percentage of Ws relative to the total protein content (Wt).

2.5.2. Particle size and ζ-potential

Freshly prepared composite solutions were diluted to 1 mg/mL with deionized water at room temperature, and the particle size and ζ-potential of the samples were determined using a zeta analyzer (DelsaNano C, Beckman Coulter, USA).

2.5.3. Free sulfhydryl content

Free sulfhydryl (-SH) content was measured using Ellman's method with 5,5′-dithiobis(2-nitrobenzoic acid) (DTNB), according to a previously reported method (Zhao et al., 2023). 30 mg samples were fully dispersed in 3 mL Tris-Gly-SDS buffer (0.086 M Tris, 0.09 M Glycine, 4 mM SDS, pH 8.0) and maintained at room temperature for 30 min with intermittent vortexing. After reaction with 30 μL DTNB for 30 min in the dark, the mixtures were centrifuged at 8000 rpm for 20 min. Absorbance of the supernatant was measured at 412 nm against a buffer blank. The free sulfhydryl content was calculated as follows:

$$-\mathrm{SH}\left(\text{μmol}\right)=\frac{73.53\times {\mathrm{A}}_{412}\times \mathrm{D}}{\mathrm{C}}$$ (1)

where A₄₁₂ is the absorbance at 412 nm, D is the dilution factor, and C is the protein concentration of the test solution (mg/mL).

2.5.4. Measurement of surface hydrophobicity (H₀)

Surface hydrophobicity (H₀) was determined using ANS as a fluorescent probe, based on a modified method by Ye et al. (2025). Protein solutions were prepared in PBS phosphate buffer solution (PBS, pH 7.0) to concentrations of 0.2–1.0 mg/mL. 2 mL of each solution was mixed with 20 μL of 8.0 mmol/L ANS solution (prepared in PBS) and incubated in the dark for 15 min. Fluorescence intensity was measured at 390 nm excitation and 470 nm emission wavelengths. The initial slope of the regression line between fluorescence intensity and protein concentration was taken as the surface hydrophobicity index.

2.5.5. Intrinsic fluorescence spectroscopy

Fluorescence spectra of walnut protein and its complexes with different ratios of SDF were analyzed according to the method of Ge et al. (2025). Samples were diluted to 0.1 mg/mL with distilled water and analyzed on fluorescence spectrophotometer (F98, Shanghai Lingguang Co., Ltd. Shanghai) at an excitation wavelength of 280 nm and scanned from 300 to 500 nm with a 5 nm slit width.

2.5.6. FTIR

The FTIR spectra of WPI, PWPI, SDF, and PWPI-SDF complexes were generated using a Spectrum Two spectrometer (PerkinElmer, USA) equipped with a diamond ATR accessory. Freeze-dried powder samples were placed directly on the ATR crystal. Background spectra of air were collected before each measurement. Spectra were collected in the range of 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹ and 32 scans accumulated.

2.5.7. Emulsifying properties

The emulsifying activity index (EAI) was evaluated from the absorbance at 500 nm (A₀) of freshly prepared emulsions, which were diluted 100-fold with 0.1% (w/v) SDS solution immediately after homogenization. Emulsifying stability index (ESI) measurements were obtained by allowing the diluted emulsions to stand for 10 min, after which the absorbance at 10 min (A₁₀) was recorded under identical conditions. The emulsifying properties were calculated as follows:

$$\mathrm{EAI}\left({\mathrm{m}}^2/\mathrm{g}\right)=\frac{2\times 2.303\times {\mathrm{A}}_0\times \mathrm{N}}{\mathrm{C}\times \left(1-\mathrm{\phi }\right)\times {10}^4}$$ (2)

$$\mathrm{ESI}\left({\mathrm{m}}^2/\mathrm{g}\right)=\frac{{\mathrm{A}}_0}{{\mathrm{A}}_0-{\mathrm{A}}_{10}}\times 10$$ (3)

where A₀ and A₁₀ are the absorbance values of diluted emulsions at 0 and 10 min, respectively; C is the protein concentration (g/mL) in the aqueous phase; N represents the dilution factor; φ is the oil volume fraction (Han et al., 2024).

2.6. Preparation of emulsion and emulsion gel

Camellia oil was selected as the oil phase because it is rich in oleic acid, has relatively high oxidative stability and a mild flavor (Gao et al., 2024). It is therefore suitable for constructing plant-based emulsions and subsequent emulsion gel systems. Emulsions were prepared by homogenizing PWPI-SDF composite solutions with camellia oil on an IKA T18 high-speed disperser (15,000 rpm, 3 min). Samples with 0.5% (w/v) SDF and oil fractions of 30, 40, 50, 60, and 70% (v/v) were denoted as PE-O30, PE-O40, PE-O50, PE-O60, and PE-O70, respectively. Samples with a 70% oil fraction and SDF concentrations of 0–1.0% (w/v) were denoted as PE-S0, PE-S0.25, PE-S0.5, PE-S0.75, and PE-S1. PE-O70 and PE-S0.5 represent the same formulation (70% oil, 0.5% SDF). A summary of the emulsion and emulsion gel designations is provided in Supplementary Table S1.

Based on the emulsion characterization results, PE-S0.5 was selected for subsequent gel preparation. Cold-set emulsion gels were prepared from PE-S0.5 emulsions by adding GDL powder directly to the emulsion system, followed by immediate homogenization (15,000 rpm, 3 min). The emulsions containing 0, 0.5, 1.0, 1.5, and 2.0% GDL (w/v, based on the total emulsion volume) were then sealed and incubated at 25 °C for 12 h to allow acidification and gelation. The resulting samples were designated as EG-G0, EG-G0.5, EG-G1, EG-G1.5, and EG-G2, respectively. In addition, samples without SDF were prepared as controls and named EG-S0-G0 and EG-S0-G1.5.

2.7. Emulsion characteristics

2.7.1. Appearance and storage stability of emulsion

Emulsion samples were sealed in sample bottles and stored at 4 °C for 0, 7, and 14 days. During storage, visual changes including phase separation, serum release, and overall emulsion uniformity were recorded.

2.7.2. Fluorescence microscopy observation

Emulsion microstructures were observed using a fluorescence microscope (BX50, Olympus, Japan) equipped with a mercury lamp and filter sets for Nile red and FITC. 1 mL of emulsion was appropriately diluted; oil droplets were stained with Nile red (10 μg/mL in isopropanol) and proteins were stained with FITC (10 μg/mL in isopropanol). After mixing, samples were kept in the dark for 30 min. Images were captured using a 10× objective lens.

2.7.3. Rheology measurement of emulsion

The rheological properties of emulsions were determined using a rheometer (HR-10, TA Instruments, USA) according to the method of Wang et al. (2023). Approximately 2 mL of freshly prepared emulsion was loaded onto the 40 mm plate with 1 mm gap at 25 °C. Flow curves were obtained by measuring apparent viscosity over a shear rate range of 0.1–100 s⁻¹. Frequency sweep tests were performed over 0.1–100 rad/s at a constant strain of 1% to obtain the storage modulus (G′) and loss modulus (G″).

2.8. Emulsion gel characteristics

2.8.1. FTIR

For FTIR analysis, volatile cyclohexane was used as a substitute oil phase only during sample preparation. This treatment was intended to minimize spectral interference from retained lipids and residual water, following a similar sample-preparation strategy reported previously (Tian et al., 2026). After gelation, the samples were rapidly frozen in liquid nitrogen and then freeze-dried in a lyophilizer. The FTIR spectra were recorded as described in Section 2.5.6 at a resolution of 4 cm⁻¹. The amide I region (1600–1700 cm⁻¹) was baseline-corrected and deconvoluted using PeakFit V4.12, and the protein secondary structure was analyzed by Gaussian second-derivative curve fitting. All other structural characterizations and functional evaluations were conducted using the camellia oil-based system.

2.8.2. Rheological measurement

The gels were cut into pieces of uniform thickness prior to testing; frequency sweeps were performed as described in Section 2.7.3. Before frequency sweep measurements, strain sweep tests were performed to confirm that 1% strain was within the linear viscoelastic region (LVR) of the emulsion gel samples.

2.8.3. Water holding capacity and cooking loss

Approximately 5 g of emulsion gel was transferred to a 50 mL centrifuge tube and centrifuged at 8000 rpm for 15 min using a refrigerated centrifuge (RJ-TGL-1850R, Wuxi, China). The supernatant was decanted, and the residue was gently blotted to remove surface moisture before weighing. WHC was calculated as:

$$\mathrm{WHC}\left(\%\right)=\frac{{\mathrm{M}}_1-{\mathrm{M}}_2}{{\mathrm{M}}_1}$$ (4)

where M₁ is the mass of the sample before centrifugation, and M₂ is the mass of the sample after centrifugation.

Cooking loss was determined as follows. Briefly, 5 g of emulsion gel was weighed into a glass beaker, sealed with parafilm, and heated at 90 °C for 30 min. After cooling to room temperature, the sample was removed, gently blotted dry on the surface and re-weighed.

$$\text{Cooking loss}\left(\%\right)=\frac{{\mathrm{M}}_1-{\mathrm{M}}_2}{{\mathrm{M}}_1}$$ (5)

where M₁ and M₂ denote the sample weights before and after heating.

2.8.4. Textural property analysis (TPA)

Texture profile analysis (TPA) of the emulsion gels was conducted according to Han et al. (2025), using a TA-TOUCH texture analyzer (Bosin, Shanghai, China) equipped with a cylindrical TA/0.5 probe. Each sample was molded into cylindrical pieces (20 mm × 7 mm) of uniform size and kept at 25 °C for 30 min before testing. Compression testing was carried out in two consecutive cycles to 50% deformation, with cross-head speeds of 3 mm/s (pre-test), 1 mm/s (test), and 3 mm/s (post-test). The instrument trigger force was set at 5 g. Hardness (N), adhesiveness (N·s), springiness, cohesiveness, resilience, and chewiness (N·mm) were calculated from the force-time curves.

2.8.5. LF-NMR

The water distribution in emulsion gels was measured by the method of Shen et al. (Shen et al., 2020) with some modifications. LF-NMR measurements were performed using a low-field nuclear magnetic resonance analyzer (NMI20-025V-I, Niumag, Suzhou, China) operating at 21 MHz and 25 °C. Samples were placed in 25 mm NMR tubes, and T₂ relaxation times were determined using the Carr-Purcell-Meiboom-Gill (CPMG) sequence with a 90°–180° pulse spacing of 200 μs, 32 scans, TE = 0.2 ms, 8000 echoes, and a 200 ms repetition interval. The decay curves were analyzed using MultiExp Inv Analysis software to obtain T₂ values.

2.8.6. MRI images

MRI was applied to visualize moisture distribution within the gels. Proton-density images were acquired using a multi-spin-echo sequence. The measurement settings included a repetition time of 1500 ms, an echo time of 0.2 ms, and a proton resonance frequency of 21 MHz. The resulting images were converted into pseudo-color maps to illustrate spatial variations in water mobility.

2.9. Statistical analysis

Measurements were performed in triplicate (n = 3). Data are expressed as mean ± SD. Differences among samples were evaluated using one-way ANOVA followed by Duncan's multiple range test (IBM SPSS 25.0), with p < 0.05 considered significant. Graphs were generated using Origin 2024.

3. Results and discussion

3.1. Physicochemical analysis of SDF

The chemical composition of SDF is shown in Table 1. The yield was 25.36%, with contents of total sugar, uronic acid, and protein at 31.14%, 23.46%, and 15.50%, respectively, and the ζ-potential was −36.61 ± 0.66 mV, indicating a markedly negative surface charge.

Table 1.

Chemical composition of SDF.

Project	SDF
Composition (%)
Yield	25.36 ± 1.00
Total sugar	31.14 ± 1.08
Uronic acid	23.46 ± 0.16
Protein	15.50 ± 0.11
ζ-potential (mV)	−36.61 ± 0.66
Monosaccharide (mol %)
Fuc	0.48
Ara	30.16
Rha	1.66
Gal	15.82
Glc	19.31
Xyl	28.52
Man	0.76
Gal-UA	1.65
Glc-UA	1.64
Molecular weight (MW)
RT (min)	24.22	28.10	31.78
Mw (kDa)	686.56	57.98	5.58

Different letter corner marks indicate significant differences (p˂0.05) between data.

As seen in Table 1, the SDF obtained via ultrasound-assisted hydrogen peroxide treatment showed a multi-peak distribution, with major molecular-weight fractions at approximately 686.56, 57.98, and 5.58 kDa, indicating the coexistence of both high and low molecular weight fractions. Such a distribution may be associated with differences in thickening behavior, interfacial stabilization, and network formation in the emulsion-gel system.

Chromatographic profiles of standard monosaccharides are shown in Fig. 1A. As illustrated in Fig. 1B and Table 1, the SDF was primarily composed of arabinose (30.16%), xylose (28.52%), glucose (19.31%), and galactose (15.82%), with minor amounts of rhamnose, galacturonic acid, glucuronic acid, mannose, and fucose, exhibiting the typical characteristics of a heteropolysaccharide (Shi et al., 2025).

Fig. 1.

Structural characterization of SDF. (A) HPLC chromatogram of standard monosaccharides, (B) HPLC chromatogram of SDF, (C) FTIR spectrum of SDF. SDF: Soluble dietary fiber extracted from bamboo shoot by-products by ultrasound-assisted alkaline hydrogen peroxide treatment.

The FTIR spectrum of SDF (Fig. 1C) displayed characteristic polysaccharide features. The absorption at 3267 cm⁻¹ was assigned to O—H stretching, whereas the band at 2956 cm⁻¹ corresponded to C—H stretching. The 1644 cm⁻¹ band corresponded to C O stretching, suggesting the presence of uronic acid structures (Fu et al., 2020). The 1369 cm⁻¹ peak was related to C—H bending, and the strong peak at 1056 cm⁻¹ was associated with C—O—C and C—O stretching, consistent with cellulose or hemicellulose units.

3.2. Characterization of complexes

3.2.1. Solubility

The aqueous dispersibility of proteins fundamentally underpins their emulsifying, foaming and gelling functionalities (Hu et al,. 2022). As presented in Fig. 2A, the pH-shift treatment markedly enhanced the solubility of native walnut protein isolate from 20.84% to 77.15% (p < 0.05), revealing a critical conformational shift from compact aggregates to a more hydrated and expanded state. The loosening of the tertiary structure exposes hydrophobic residues, fundamentally enhancing the interfacial adsorption capacity and improving emulsion stabilization (Wang et al., 2024). With the addition of SDF, solubility remained at a relatively high and stable level in the range of 0.25%–0.75%, with a value of 79.98% observed at 0.5% SDF. This may be related to enhanced surface charge and the formation of hydration shells that limited protein-protein aggregation (Zhao et al., 2024). At higher SDF levels (≥0.75%), molecular crowding could slightly restrict water accessibility, leading to a minor decline in solubility.

Fig. 2.

(A) Solubility, (B) particle size and ζ-potential, (C) free sulfhydryl content and (D) surface hydrophobicity (H₀) of WPI, PWPI and PWPI-SDF complexes at different SDF concentrations (0–1%, w/v). WPI: Walnut protein isolate; PWPI: pH-shifted walnut protein isolate.

3.2.2. Particle size and ζ-potential

To elucidate how pH-shifting treatment restructures walnut protein particles to enhance colloidal stability and to analyze structural reorganization of the complexes, particle size and ζ-potential were measured. As depicted in Fig. 2B, native WPI had a low surface charge of −18.3 ± 0.66 mV and formed large aggregates with a size of 1122.5 ± 123.03 nm, indicating a conformation prone to irregular clustering. After pH-shifting, the ζ-potential decreased significantly to −27.5 ± 1.2 mV, accompanied by a pronounced decrease in particle size to 652.3 ± 55.4 nm, indicating aggregate disruption and formation of finer, electrostatically stabilized particles favorable for emulsification (Song et al., 2025). This change suggests that pH-shifting exposed more charged groups through protein unfolding, thereby increasing the absolute ζ-potential and strengthening electrostatic repulsion between particles. Such enhanced electrostatic repulsion is important for reducing particle–particle association and improving colloidal dispersion stability. Additional carboxyl groups provided by SDF further enhanced colloidal stability. At 0.5% SDF, particles became smaller and more uniform, mainly because the increased surface charge density strengthened electrostatic repulsion, while the polysaccharide chains of SDF generated steric hindrance that prevented re-aggregation (Wang et al., 2023). At this level, the improved particle dispersion was attributed to the combined effects of electrostatic repulsion and steric stabilization in the PWPI–SDF complexes. Conversely, the enhanced viscosity of the continuous phase provided by excessive SDF restricted molecular mobility, weakening effective protein-polysaccharide interactions, as evidenced by a decrease in ζ-potential and a concomitant increase in particle size, a phenomenon consistent with bridging flocculation at excessive polysaccharide contents.

3.2.3. Free sulfhydryl content

In Fig. 2C, the free sulfhydryl content increased significantly from 18.67 μmol/g in WPI to 24.19 μmol/g after pH-shifting (p<0.05), which indicated the cleavage of intramolecular disulfide bridges and conformational relaxation that exposed previously buried cysteine residues (Jiang et al., 2017). This alkaline-induced unfolding led to more reactive thiol groups being maintained in an accessible state, which is vital for enhancing the protein's surface emulsifying activity at the oil–water interface. Consequently, complexation with SDF further increased thiol content to 27.63 μmol/g at PWPI-S0.5, suggesting that SDF may provide a microenvironment that helped preserve free thiol groups by limiting oxidation and reformation of disulfide bonds. The subsequent decline in free -SH content at higher SDF levels may be mainly related to reduced accessibility of sulfhydryl groups within a denser PWPI-SDF environment, although partial oxidation during enhanced intermolecular association cannot be completely excluded (Ren et al., 2020).

3.2.4. Surface hydrophobicity (H₀)

Surface hydrophobicity serves as an indicator of protein conformational changes and the exposure of nonpolar residues (Wu et al., 2024). As illustrated in Fig. 2D, WPI displayed a low hydrophobicity, consistent with its compact conformation in which most nonpolar residues were buried in the interior. pH-shifting significantly improved the hydrophobicity (p < 0.05), reflecting disruption of intramolecular packing and exposure of aromatic and aliphatic side chains, a trend also reported for pea protein (Jiang et al., 2017). The presence of 0.5% SDF further amplified hydrophobicity, implying that SDF helped stabilize the unfolded protein conformation and facilitated the exposure of these nonpolar domains. Beyond this level, however, hydrated polysaccharide chains likely formed a hydration barrier around the protein surface, which reduced the accessibility of hydrophobic sites to the ANS probe and consequently lowered the apparent hydrophobicity (Lavaei et al., 2022).

3.2.5. Fluorescence spectroscopy

Intrinsic fluorescence spectroscopy provided clear evidence of conformational rearrangements within the protein, reflecting the polarity of the microenvironment surrounding its tryptophan residues. As illustrated in Fig. 3A, a significant quenching of fluorescence intensity was observed, dropping from 312 a.u. to 189 a.u. after pH-shifting, and SDF complexation further indicated that Trp residues migrated from hydrophobic cores to a more polar aqueous environment (Cai et al., 2025). This reduction suggests that pH-shifting loosened the protein structure and exposed previously buried aromatic regions, thereby making PWPI more available for subsequent association with SDF. Furthermore, the progressive decline in Trp fluorescence from 0.25% to 0.5% indicated that PWPI–SDF association was further promoted within this concentration range (Blagodatskikh et al., 2024). Beyond this level, hydrated polysaccharide chains likely formed a hydration barrier around the protein surface, which reduced the accessibility of hydrophobic sites to the ANS probe and consequently lowered the apparent surface hydrophobicity.

Fig. 3.

(A) Intrinsic fluorescence spectra, (B) FTIR spectra, (C) emulsifying activity index (EAI), and (D) emulsifying stability index (ESI) of WPI, PWPI, and PWPI-SDF complexes at different SDF concentrations (0–1%, w/v).

3.2.6. FTIR analysis

FTIR was used to investigate functional group variations and intermolecular interactions among the WPI, PWPI, SDF, and PWPI-SDF complexes. In the O-H/N-H stretching region (3300–3000 cm⁻¹), the absorption band of the PWPI-SDF complex exhibited a distinct redshift of approximately 10 cm⁻¹ compared with WPI, accompanied by a broadening of the peak (Fig. 3B), indicating the formation of stronger hydrogen bonds between PWPI functional groups and the abundant hydroxyl groups of SDF that served as a major driving force for complex formation (Wang et al., 2025). The asymmetric –CH₂ band also shifted slightly, indicating changes in hydrophobic chain packing (Jing et al., 2021). Furthermore, changes in the amide I region (1600–1700 cm⁻¹) were also observed, with the peak shifting from 1630 cm⁻¹ in WPI to 1639 cm⁻¹ in PWPI and 1641 cm⁻¹ in PWPI-S0.75/PWPI-S1, indicating structural rearrangement of the protein after pH-shifting and subsequent complexation with SDF. In addition, the C—O stretching band in the carbohydrate fingerprint region shifted from 1067 cm⁻¹ in WPI to 1062–1063 cm⁻¹ in PWPI-SDF complexes, further supporting the participation of SDF in the formation of the complexes. Overall, these FTIR results indicate that hydrogen bonding and other non-covalent interactions were involved in PWPI–SDF complex formation.

3.2.7. Emulsifying properties

pH-shifting markedly improved the emulsifying properties of WPI by promoting structural unfolding, which exposed additional hydrophobic groups and facilitated its adsorption at the oil–water interface (Wu et al., 2024). Compared with WPI alone, the PWPI-SDF complexes exhibited superior emulsifying capacity. As evidenced in Fig. 3C and D, EAI rose from 0.67 ± 0.04 to 1.40 ± 0.02 m²/g and ESI increased from 21.0 ± 3.5 to 78.07 ± 12.7 min when 0.5% SDF was incorporated (p < 0.05). This improvement may be related to the interfacial association between PWPI and negatively charged SDF, which contributed to improved droplet stabilization. At SDF concentrations above 0.75%, thermodynamic incompatibility occurred as excess SDF bridged adjacent droplets, causing flocculation and lowering droplet dispersibility, thereby diminishing the synergistic stabilization effect (Li et al., 2019).

3.3. Characterization of emulsion

3.3.1. Appearance and storage stability

Fig. 4 showed the visual appearance of emulsions stabilized by PWPI and PWPI-SDF complexes during storage. Macroscopically, higher oil fractions markedly expanded the emulsion volume and showed a thicker and more self-supporting appearance. Emulsions containing 30%–50% oil underwent rapid phase separation within 1 day, which became more pronounced during storage (Fig. 4A). In contrast, formulations with 60%–70% oil maintained a more homogeneous appearance over 14 days and gradually formed a semi-solid, gel-like structure (Fig. 4B). These observations suggest that higher oil loading was associated with improved storage stability.

Fig. 4.

(A) Visual appearance of emulsions prepared with PWPI-SDF complexes containing 0.5% SDF, and (B) emulsions prepared with PWPI-SDF complexes with varying SDF concentrations (0–1%, w/v) at a fixed oil volume fraction of 70%.

3.3.2. Fluorescence microscopy of emulsions

Fluorescence microscopy images, with Nile red coloring the oil phase red and FITC coloring proteins green, revealed a typical oil-in-water emulsion morphology characterized by protein-stabilized oil droplets dispersed uniformly in the aqueous phase. At lower oil contents, the emulsion showed large, irregularly distributed droplets (Fig. 5A), whereas at 70% oil content and 0.5% SDF concentration, the emulsion exhibited more uniform and closely packed droplets, together with a more continuous interfacial film (Fig. 5B). The co-adsorption of PWPI and SDF likely enhanced steric hindrance and electrostatic repulsion, thereby helping to prevent droplet coalescence and improve emulsion stability (Zhao et al., 2023). When the SDF concentration was further increased, excessive free chains in the continuous phase likely induced depletion flocculation, resulting in reduced stability (Wang et al., 2023).

Fig. 5.

Fluorescence microscopy images of emulsions stabilized by PWPI-SDF complexes. (A) Different oil volume fractions with 0.5% SDF: (A–E) 30%, 40%, 50%, 60%, and 70%. (B) Different SDF concentrations at a fixed oil fraction (70%): (a–e) emulsions containing 0–1% SDF. Columns 1–3 represent the oil phase, protein phase, and merged images, respectively. Scale bar = 100 μm.

3.3.3. Rheological behavior

The flow curves (Fig. 6A, B) revealed that all emulsions exhibited typical shear-thinning behavior, as apparent viscosity declined with increasing shear rate. This typical pseudoplasticity was attributed to the progressive breakdown of droplet aggregates under shear (Zhang et al., 2024). Viscosity increased markedly with oil fraction, and the 70% system displayed the strongest resistance to flow, attributable to droplet crowding and a space-filling network. The incorporation of SDF further contributed to structural stabilization by promoting the formation of PWPI-SDF complexes at the interface and by increasing the rheological contribution of the continuous phase through molecular entanglement. Dynamic frequency sweeps further supported these observations. Emulsions with 30–40% oil were viscous-dominated (G′ < G″ at low frequency) and showed only a delayed crossover, reflecting weak elasticity and fragile droplet associations. By contrast, as shown in Fig. 6C, E, the 70% oil emulsion exhibited consistently higher G′ than G″, where droplet compression promoted more pronounced gel-like behavior with significantly elevated modulus (Zhao et al., 2023). At this oil level, moderate SDF supplementation (0.5%) further strengthened the elastic network, whereas excessive SDF reduced both G′ and G″, suggesting that excess fiber chains may induce depletion flocculation or steric hindrance, thereby disturbing the homogeneity of the protein-stabilized network and weakening its mechanical integrity (Zhang et al., 2023; Zuo et al., 2022).

Fig. 6.

Rheological properties of emulsions: (A, B) flow curves of emulsions at different oil volume fractions (30–70%) and SDF levels (0–1%), respectively; (C, E) storage modulus (G') and loss modulus (G") at different oil volume fractions; (D, F) storage modulus (G') and loss modulus (G") at different SDF levels.

3.4. Characterization of emulsion gels

3.4.1. Mechanism of gelation

Fig. 7A demonstrated the FTIR spectra of emulsion gels prepared with varying GDL levels. The amide I band shifted from 1636 cm⁻¹ (emulsion without GDL) to 1626 cm⁻¹ at 1.5% GDL, accompanied by clear band broadening, indicating a transition from disordered to β-sheet structures during gel formation, leading to a rigid gel network (Wang et al., 2023). Simultaneously, the redshift in the carbohydrate region from 1076 to 1037 cm⁻¹ indicated the incorporation of SDF chains into the protein network through hydrogen bonding. By deconvoluting the amide I region of the FTIR spectra, the secondary structure composition of the emulsion gels was determined. As shown in Fig. 7B, the sample without SDF and GDL (EG-S0-G0) exhibited the lowest β-sheet content (31.07%), while α-helix and random coil fractions were relatively high (23.63% and 23.07%). Upon incorporating SDF, the proportion of β-sheet increased markedly from 31.07% to 38.81%, accompanied by reductions in both α-helix and random coil structures (Yuan et al., 2024). This shift suggested that SDF facilitated the exposure of hydrophobic groups and strengthened hydrophobic interactions among WPI molecules, which likely facilitated α-helix unwinding and potentially improving gel network stability (Li et al., 2019). Subsequent acidification with GDL further altered protein conformation. In the initial stage of acidification, β-sheet content continued to increase, α-helix remained low level and β-turn stayed relatively high, indicating that the protein-polysaccharide complexes retained conformational flexibility during network assembly. With increasing GDL concentration, disordered regions progressively transformed into β-sheet, leading to protein-protein interactions and a compact, mechanically reinforced gel network (Wei et al., 2024), which was macroscopically manifested as hardness and chewiness observed in TPA.

Fig. 7.

(A) FTIR spectra of composite emulsion gels induced by different GDL concentrations, and (B) corresponding secondary structure composition from deconvolution of the amide I band.

3.4.2. Rheological behavior of emulsion gels

The rheological behavior of the emulsion gels was evaluated over 0.1–100 rad/s. All emulsion gel samples exhibited typical gel-like behavior, with the storage modulus (G′) remaining higher than the loss modulus (G″) throughout the tested frequency range (Fig. 8A, B), confirming the formation of elastic, solid-like networks (Chen et al., 2025). The gel without SDF and GDL (EG-S0-G0) showed the weakest viscoelasticity. At 100 rad/s, its G′ and G″ were 508.11 and 86.98 Pa, respectively. After SDF incorporation, the viscoelasticity increased noticeably, and in EG-G0, G′ and G″ increased to 1047.72 and 205.88 Pa, respectively. SDF contributed to improved viscoelastic behavior by reinforcing the protein network through hydrogen bonding and molecular entanglement (Wu et al., 2025). With increasing GDL concentration, both moduli further increased, and EG-G1.5 showed the highest values, with G′ and G″ reaching 3310.21 and 846.28 Pa, respectively, at 100 rad/s. The acidification induced by GDL significantly enhanced the viscoelasticity of SDF-containing emulsion gels, indicating decreased electrostatic repulsion and strengthened intermolecular associations (Qin et al., 2023). Meanwhile, the pH reduction from GDL hydrolysis facilitated protein aggregation and increased the value of G′ (Chang et al., 2014). In addition, the gel formed at a high GDL concentration (EG-G2) showed slightly reduced viscoelasticity, suggesting that excessive acidification partially weakened the gel structure.

Fig. 8.

(A) storage modulus (G'), (B) loss modulus (G"), (C) water-holding capacity and (D) cooking loss of GDL-induced composite emulsion gels.

3.4.3. Water retention capacity

Water holding capacity (WHC) reflects the ability of emulsion gels to retain water during processing (Wang et al., 2019). WHC was enhanced by SDF incorporation (Fig. 8C), as flexible polysaccharide chains acted as water-binding sites and reduced pore heterogeneity within the gel matrix. GDL acidification likely promoted the formation of a more continuous network, thereby restricting water migration. At moderate GDL levels, this balance of rigidity and flexibility immobilized water efficiently, whereas excessive acidification compacted the matrix, generating microfractures that facilitated leakage. The cooking loss remained below 10% across all GDL-induced gels (Fig. 8D), whereas EG-S0-G0 showed a significantly higher loss (p < 0.05), confirming that gradual acidification produced uniform networks able to retain water under thermal processing.

3.4.4. Texture analysis

The textural properties of emulsion gels were determined by TPA, and the results indicated that the combined effects of GDL and SDF played an important role in gel formation. As presented in Table 2, in the 0.5% SDF system, hardness increased from 15.82 ± 0.47 N in EG-G0 to 32.48 ± 0.79 N in EG-G1.5, while springiness increased from 0.34 ± 0.03 to 0.73 ± 0.04. This suggests that 1.5% GDL promoted more ordered and compact aggregation and cross-linking of the PWPI-SDF emulsion (Yuan et al., 2024). The enhanced hardness indicated stronger compressive resistance of the network, while the increase in springiness reflected better deformation recovery. At the same GDL level, EG-G1.5 also showed higher hardness and resilience than the gel without SDF, further indicating that SDF reinforced the gel network. When the GDL concentration increased to 2.0%, hardness, adhesiveness, and resilience decreased to 29.83 ± 0.84 N, 17.41 ± 0.09 N·s, and 0.15 ± 0.002, respectively, suggesting that excessive acidification promoted over-aggregation, induced local structural disruption, and weakened the interfacial and matrix network (Zhang et al., 2021). These results indicate that moderate acid-induced cross-linking, together with SDF reinforcement, yielded a compact and elastic droplet-filled network.

Table 2.

Texture profile analysis (TPA) of composite emulsion gels prepared with varying GDL levels.

Sample	Hardness (N)	Chewiness (N·mm)	Adhesiveness (N·s)	Springiness	Cohesiveness	Resilience
EG-S0-G0	10.15 ± 0.663f	1.393 ± 0.173g	4.056 ± 0.314f	0.319 ± 0.011f	0.399 ± 0.005f	0.057 ± 0.005f
EG-G0	15.82 ± 0.465e	2.321 ± 0.057f	7.277 ± 0.163e	0.343 ± 0.025e	0.46 ± 0.004e	0.069 ± 0.007e
EG-G0.5	24.56 ± 1.042d	7.966 ± 0.398e	12.36 ± 0.495d	0.644 ± 0.008c	0.503 ± 0.002d	0.133 ± 0.002d
EG-G1	28.28 ± 0.588c	9.894 ± 0.192d	14.781 ± 0.325c	0.669 ± 0.003bc	0.523 ± 0.003c	0.139 ± 0.002c
EG-S0-G1.5	28.36 ± 0.441c	8.89 ± 0.129c	14.855 ± 0.127c	0.598 ± 0.005d	0.524 ± 0.004c	0.099 ± 0.005c
EG-G1.5	32.48 ± 0.786a	13.105 ± 0.562a	17.926 ± 0.541a	0.732 ± 0.039a	0.552 ± 0.005b	0.164 ± 0.002a
EG-G2	29.83 ± 0.839b	12.021 ± 0.05b	17.408 ± 0.089b	0.691 ± 0.005b	0.584 ± 0.019a	0.153 ± 0.002b

The results are expressed as mean ± standard deviation (SD). Within the same column, values with different superscript letters indicate significant differences (p<0.05).

3.4.5. LF-NMR

LF-NMR was employed to assess water mobility within the emulsion gels based on T₂ relaxation times: T₂₁ (0.1–10 ms, bound water), T₂₂ (10–100 ms, immobilized water), and T₂₃ (100–1000 ms, free water), with relative peak areas denoted as PT₂₁, PT₂₂, and PT₂₃ (Zhao et al., 2024). As shown in Fig. 9A, T₂₂ accounted for >90% of the total signal in all samples, revealing that water was predominantly immobilized within the gel network. The incorporation of SDF caused a pronounced leftward shift of the T₂₁ and T₂₂ peaks. At the same time, PT₂₁ increased and PT₂₃ decreased. These changes reflected that the abundant hydrophilic groups in SDF strengthened protein–water interactions and promoted the conversion of free water into more restricted states. GDL hydrolysis further reduced water mobility by creating an acidic environment that promoted cross-linking at the oil–water interface (Wang et al., 2020). With increasing GDL content, PT₂₁ first decreased and then increased, indicating transient network disruption during pH reduction followed by structural reorganization. At the same time, PT₂₃ decreased steadily and the T₂₃ peak shifted to shorter relaxation times (Fig. 9B). At 2% GDL, however, the overly compact network likely shortened intermolecular spacing and expelled part of the immobilized water into the free state (Li et al., 2019), which was consistent with WHC results. These results demonstrate that SDF mainly strengthened water-binding capacity, while GDL increased network compactness, and both factors jointly regulated water distribution.

Fig. 9.

T₂ relaxation behavior of SDF and GDL-induced emulsion gels: (A) changes in water states of composite emulsion gels, (B) relaxation times and corresponding peak area ratios, and (C) MRI images of emulsion gels. The color scale represents the proton density of water molecules, with red indicating higher proton density and blue indicating lower proton density. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

3.4.6. MRI

Magnetic Resonance Imaging (MRI) provides a non-destructive approach to visualize water distribution in the emulsion gels (Chen et al., 2025). In the pseudo-color images, red regions represent free water with higher proton density, while blue regions correspond to bound water with lower proton density (Zhao et al., 2024). Results showed that gels without SDF exhibited a discontinuous network, exhibiting large areas of red free water. With increasing GDL concentration, red regions progressively diminished, and the EG-G1.5 sample showed high uniformity, reflecting efficient water immobilization and enhanced structural integrity (Fig. 9C), consistent with the significant reduction in PT₂₃ observed in the LF-NMR. In contrast, at 2% GDL, localized red patches reappeared. Such excessive acidification may lead to the formation of a coarse, brittle, and non-uniform network with larger pores, which fails to effectively entrap the oil phase and water, thereby reducing water immobilization (Gao et al., 2020). The MRI results were consistent with the WHC and TPA results, further indicating that moderate acidification combined with SDF improved water retention and gel integrity. The proposed mechanism underlying GDL-induced gel formation, water immobilization, and structural changes is illustrated in schematic Fig. 10.

Fig. 10.

Gelation mechanism for the formation of PWPI-SDF stabilized high-oil emulsion gels induced by GDL.

4. Conclusion

This study developed a high-oil (70% v/v) cold-set emulsion gel based on pH-shifted walnut protein isolate (PWPI) and bamboo shoot soluble dietary fiber (SDF). pH-shifting markedly improved the solubility and interfacial functionality of walnut protein, while appropriate SDF addition enhanced emulsion stability through PWPI-SDF complex formation. At 0.5% SDF and 1.5% GDL, the system exhibited the most favorable gel structure, characterized by strengthened hydrogen bonding, increased β-sheet content, improved hardness and elasticity, and enhanced water immobilization. Overall, these results demonstrate the combined effect of PWPI and SDF in structuring high-oil emulsion gels and support the potential application of bamboo shoot-derived SDF in plant-based emulsion gel systems.

CRediT authorship contribution statement

Liu Liu: Writing – original draft, Visualization, Methodology, Investigation, Data curation. Hai-Bing Ran: Validation, Formal analysis. Li-Kang Qin: Supervision. Yu-Long Jia: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition.

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

Supplementary material.

Data availability

Data will be made available on request.