Enhancing bioaccessibility of camellia seed oil through whey protein isolate–quercetin complex stabilized emulsions Authors: Wenxin Zha, Qingqing Li, Yu Yuan, Likang Qin, Haiyan Zhong, Yong Zhu Published: 2026 ABSTRACT AbstractCamellia seed oil (CSO) exhibits poor dispersibility and low bioaccessibility during digestion. In this study, a whey protein isolate–quercetin (WPI–Que) complex was developed to fabricate CSO-in-water emulsions. The optimal complex (10 mg Que. was dissolved in 100 mL WPI solution) showed favorable physicochemical and emulsifying properties, and the formation of the complex originated from quercetin-induced conformational changes of WPI structure. Furthermore, the emulsion consisting of 5 g CSO and 10 mL emulsifier solution exhibited a uniform droplet distribution and an absolute zeta potential above 30 mV. More importantly, the emulsion exhibited higher bioaccessibility compared to bulk oil, with the maximum FFA release of 69.33 ± 0.99% and lipolysis rate of 0.0531 min−1. These findings suggest that the WPI–Que complex can be used as emulsifier for constructing CSO delivery in functional food systems.Keywords: Camellia seed oil, Whey protein isolate, Quercetin, Emulsion, Bioaccessibility Introduction Camellia seed oil (CSO) is a woody edible oil in China, which contains more than 90% unsaturated fatty acids and diverse bioactive compounds (He et al., 2023; Li et al., 2023). Numerous studies have demonstrated that CSO exhibits multiple health-promoting effects, including antioxidant, hypolipidemic and cardioprotective activities (Qin et al., 2024). These findings suggest that CSO can serve as an ingredient in functional foods, whereas it is hardly incorporated into water-based food systems due to its hydrophobic nature and poor aqueous dispersibility (Wang et al., 2024). In addition, the dispersion state of lipids results in their poor digestive accessibility because lipid digestion depends on the interfacial area available for lipase action (McClements, 2018). Emulsification is regarded as an effective strategy for improving the dispersibility of hydrophobic oils, as well as enhancing their bioaccessibility both in vitro and in vivo (Yang et al., 2018; Zeng et al., 2023). Whey protein isolate (WPI) is extensively employed in the food industry owing to its excellent interfacial activity and recognized safety (Rafiei et al., 2025). However, the emulsification performance of WPI can be impaired under unfavorable environmental conditions, particularly near the isoelectric point or at high ionic strength, which may result in droplet aggregation (Keramat & Ehsandoost, 2023; Tao et al., 2023). Incorporating polyphenols into proteins has been confirmed as an effective strategy for improving emulsifying properties (Jauregi et al., 2021; Ozdal et al., 2013), whereas this improvement is strongly associated with polyphenol types, concentration and interaction mode(Chen et al., 2022; Cheng et al., 2023).It has been suggested that quercetin (Que) and WPI influence each other when combined; specifically, Que. modulates the proteins' structural and functional properties, in turn, WPI improve the aqueous dispersibility and bioaccessibility of quercetin (Chen et al., 2022; Cheng et al., 2023). Based on these considerations, we designed a composite emulsifier consisted of WPI and Que., and its formulation was optimized with respect to physicochemical characteristics and emulsifying performance. Moreover, it was confirmed that the intermolecular forces existed between WPI and Que., and the conformational changes of WPI were induced by Que. The composition of the emulsion system was also optimized in terms of colloidal properties and microstructure; the superiority in bioaccessibility of the emulsion was verified by using an in vitro digestion model coupled with mathematical fitting methods. Materials and methods Materials and chemicals The refined CSO was supplied by Guizhou Hengshengyuan Agricultural Development Co., Ltd. (Tianzhu, Guizhou, China). WPI was purchased from MSG company (Eden Prairie, MN, USA). Que. (purity ≥95%) and pancreatic lipase (porcine source, 125 U/mg) were obtained from Sigma-Aldrich (St. Louis, MO, USA). Sodium dodecyl sulfate (SDS), Nile Red and pepsin (porcine source, 3000 U/mg) were purchased from Rhawn Scientific Co., Ltd. (Shanghai, China). Nile Blue A, pancreatin (130 U/mg) and bile salts (cholic acid content ≥75%) were purchased from Macklin Biochemical Co., Ltd. (Shanghai, China). α-Amylase (2000 U/g) was purchased from Yuanye Bio-Technology Co., Ltd. (Shanghai, China). Inorganic salts (KCl, KH2PO4, NaHCO3, NaCl, MgCl2·6H2O, (NH4)2CO3, CaCl2(H2O)2), HCl, NaOH, KBr and other chemicals were of analytical grade. Preparation of WPI–Que complex The WPI–Que complex emulsifier was prepared according to previous studies with minor modifications (Chen et al., 2022; Wang et al., 2023). WPI was dissolved in deionized water (2%, w/v) and hydrated overnight at 4 °C, and then 4 aliquots of quercetin (10, 20, 40 and 80 mg) were separately added into 4 different WPI solutions denoted as C₁₀, C₂₀, C₄₀, and C₈₀, respectively. The mixtures were stirred continuously for at least 4 h at ambient temperature and further stored overnight at 4 °C. Part of each aqueous dispersion was freeze-dried for 48 h for subsequent experiments. Preparation of CSO emulsions CSO emulsions were prepared by mixing CSO with the WPI–Que complex solution at ratios of 1:10, 2:10, 3:10, 4:10, 5:10, 6:10, 7:10, and 8:10 (w/v), denoted as E1, E2, E3, E4, E5, E6, E7, and E8, respectively. The oil-to-complex solution ratios were selected with reference to previous studies in which the oil-phase fraction was varied to optimize the formulation and stability of oil-in-water emulsions (Yang et al., 2018; Zeng et al., 2023).The resulting mixtures were homogenized at 20,000 rpm for 2 min using a high-speed disperser and stored at 4 °C in the dark until further analysis. Determination of particle size, zeta potential and polydispersity index (PDI) of WPI–Que complexes and CSO emulsions The WPI–Que complexes and CSO emulsions were appropriately diluted with deionized water and transferred into disposable cuvettes for measurement using a laser particle size analyzer at 25 °C. The particle size and PDI were determined based on the principle of dynamic light scattering (DLS), while the zeta potential was calculated from the electrophoretic mobility. Determination of WPI–Que interactions using FT-IR spectroscopy FT-IR spectroscopy was used to analyze the interactions of WPI and Que., with slight modifications to the previously reported method (Wang, Xie, et al., 2022). The freeze-dried WPI-Que complex powders were thoroughly mixed with potassium bromide (KBr) at a mass ratio of 1:100 (w/w), finely ground, and pressed into pellets for spectral analysis. FT-IR spectra were recorded over the wavenumber range of 4000–400 cm−1 with a resolution of 4 cm−1, and each sample was scanned 32 times. Determination of protein conformation using CD spectroscopy Far-UV CD spectroscopy was used to analyze the secondary structure of WPI in the complexes, with slight modifications to the previously reported method(Feng et al., 2019). The samples were diluted 100-fold with 10 mM PBS buffer (pH 7.2–7.4), and the diluted complexes (200 μL) were transferred into the 1 mm quartz cuvettes and analyzed at room temperature using a far-UV CD spectrometer in the wavelength range of 190–240 nm. The raw data were smoothed using Spectra Manager software and were further analyzed using an online protein secondary structure estimation platform with the SELCON algorithm to estimate the relative contents of α-helix, β-sheet, β-turn, and random coil structures. Determination of emulsifying activity and stability of WPI–Que complexes The emulsifying activity and stability of WPI–Que complexes were determined following the method described by Li (Li et al., 2014) with slight modifications. The WPI–Que complexes (8 mL) were mixed with CSO (2 mL). The resulting mixtures was homogenized (20,000 rpm, 1 min) using a high-speed disperser to form emulsions. Aliquots (50 μL) of the emulsion were collected from a fixed position (5 mm above the bottom of the tube) at 0 and 10 min, and then each aliquot was diluted with 5 mL of 0.1% (w/v) sodium dodecyl sulfate (SDS) solution (1:100 dilution). Subsequently, the absorbance of the diluted emulsions was measured at 500 nm with a UV–Vis spectrophotometer. The emulsifying activity index (EAI) and emulsion stability index (ESI) were calculated referring to the equations: where f indicates the dilution factor (100), φ indicates the proportion of CSO volume to total (φ = 0.2), C indicates the WPI concentration in the aqueous phase (2%, w/v), A0 and A10 indicate the absorbance of the emulsions collected at 0 min and 10 min, respectively. Observation of microstructure and morphology of CSO emulsions The CSO emulsion (10 μL) was deposited onto a glass slide and observed under an optical microscope (40×) to capture the microstructure of the emulsion droplets. Digital images were captured using a CCD camera system connected to the microscope. The interfacial structure of CSO emulsions was observed using confocal laser scanning microscopy (CLSM). The CSO emulsion (500 μL) was mixed with 20 μL of the mixed dye solution prepared by combining Nile Red (2 mg/mL) and Nile Blue A (2 mg/mL) solutions in equal volumes. The stained emulsion was placed on a glass slide and observed using CLSM. Nile Red, which selectively stains the oil phase, was excited at 528 nm with emission recorded at 576 nm. The WPI–Que complex, labeled with Nile Blue A, was excited at 631 nm with emission recorded at 660 nm. All fluorescence images were digitally acquired and processed. The stability of the optimized CSO emulsion under different solution environment The stability of the optimized CSO emulsion (E5) was evaluated under various pH (2, 4, 6, 8 and 10, adjusted with 0.1 M HCl or NaOH) and ionic strength (50, 100, 150 and 200 mmol/L NaCl) conditions. Stability was assessed by visual observation (using a camera), as well as by measurements of particle size and zeta potential (as described in Section 2.4). Procedures of in vitro digestion In vitro digestion was carried out following the standardized INFOGEST 2.0 static digestion protocol (Brodkorb et al., 2019) and this study (Premjit & Mitra, 2023). The CSO emulsion (E5, 5 mL) was transferred into a wide-mouthed conical flask and mixed with 4 mL of simulated salivary fluid (SSF) (Table 1). The mixture was neutralized prior to the addition of 1 mL of α-amylase (75 U/mL in the final mixture) and then incubated in a shaking water bath (150 rpm) at 37 °C for 5 min to simulate oral digestion. Table 1The composition of electrolyte stock solutions.Buffer Solution (1.25×)SSF(pH = 7) (mM)SGF(pH = 3) (mM)SIF(pH = 7) (mM)KCl15.16.96.8KH2PO43.70.90.8NaHCO313.62585NaCl–47.238.4MgCl2(H2O)60.150.120.33(NH4)2CO30.060.5–HCl1.115.68.4CaCl2(H2O)21.50.150.6 The composition of electrolyte stock solutions. The digesta collected from oral digestion was mixed with 8 mL of simulated gastric fluid (SGF) (Table 1). The pH of the mixture was titrated to 3.0 by adding 0.3 M HCl before adding 2 mL of pepsin (2000 U/mL in the final mixture). The mixture was then incubated in the shaking water bath (150 rpm) at 37 °C for 2 h to simulate gastric digestion. Eventually, the digesta obtained from gastric digestion was mixed with 16 mL of simulated intestinal fluid (SIF) (Table 1). The pH of the mixture was adjusted to 8.0 using 1 M NaOH. Then, 4 mL of intestinal digestive juice (containing bile salts, pancreatin, and additional pancreatic lipase) was added to achieve final concentrations of 10 mM, 100 U/mL, and 2000 U/mL, respectively. The mixture was incubated in the shaking water bath at 37 °C for 2 h to simulate intestinal digestion. The control was prepared by mixing bulk CSO (2.500 g) and deionized water (2.500 g), the CSO emulsion and control were subjected to the above procedures in triplicate. Determination of released free fatty acids The extent of lipid digestion was monitored by measuring the released free fatty acids (FFA) according to the study (Li & McClements, 2010). Bulk CSO and CSO emulsions were sequentially subjected to the in vitro digestion protocol referring to section 2.10. During the intestinal digestion, the digesta was completely neutralized by titrating with NaOH solution (1 mol/L) under continuous stirring at 37 °C, and the cumulative volumes of consumed NaOH were recorded at 1, 3, 5, 10, 20, 30, 60, 90, and 120 min. It is generally assumed that pancreatic lipase specifically hydrolyzes the ester bonds at the sn-1 and sn-3 positions of the triacylglycerol (TAG) backbone, thereby releasing two fatty acids per TAG molecule. Consequently, the percentage of FFA release (FFA%) was calculated using the following equation: Where VNaOH indicates the volume (L) of NaOH, C(NaOH) indicates the molar concentration of NaOH (1 mol/L), M(oil) indicates the average molecular weight of CSO (882 g/mol), W(oil) indicates the initial mass of oil in the digestion system (g). The FFA data were fitted using a three-parameter first-order kinetic model: where FFAt indicates the FFA release (%) at digestion time t, FFA0 indicates the initial FFA release (%), FFA∞ indicates the maximum FFA release (%), and k indicates the first-order release rate constant (min−1). Nonlinear regression analysis was performed using Origin software, and the goodness of fit was evaluated based on the coefficient of determination (R2) and reduced chi-square values. Data analysis All results are presented as mean ± SD of three independent replicates. Statistical analysis was performed using one-way analysis of variance (ANOVA) followed by Waller–Duncan multiple range tests. Differences were considered significant at p < 0.05 using IBM SPSS Statistics 26.0. Results and discussion Physicochemical characterization of WPI–que complexes The incorporation of Que. into WPI induced changes in the zeta potential, particle size and PDI of the colloidal system. Zeta potential is a key parameter for characterizing the surface charge of particles in liquid dispersions, reflecting their electrostatic stability and the strength of interparticle interactions (Retamal Marín et al., 2017). Hence, the absolute value of zeta potential is positively associated with the colloidal stability (Chen et al., 2025), and the PDI indicates the uniformity of the particle size distribution (Liao & Huang, 2022). In general, a larger absolute value of zeta potential enhances electrostatic repulsion and reduces aggregation, leading to a decrease in the measured particle size and PDI (Liu & Chen, 2023). The complexes of C0, C10, C20, C40, C80 are shown in Fig. 1. Notably, the complex of C10 exhibited the highest absolute value of zeta potential (16.26 ± 0.03 mV) (Fig. 2) and showed a reasonable particle size and PDI (Fig. 2). This may be originated from the fact that more negatively charged amino acid residues were exposed to the protein surface (Mantovani et al., 2018). Nevertheless, the remaining samples presented a bridging effect because excess phenolic molecules linked with protein chains, resulting in the charge shielding and protein aggregation (Jia et al., 2022; Khan et al., 2024; Zdovc et al., 2022). These results suggested that the C10 complex was suitable for preparing a stable CSO emulsion system. Appearance diagram of WPI-Que complexes. The zeta potential, particle size, and PDI of WPI-Que complexes. Within the same group, values bearing different letters are significantly different from each other (p < 0.05). Intermolecular forces between WPI and Que FT-IR analysis was used to provide spectroscopic evidence for the interaction between WPI and Que. The FT-IR results showed that the interaction between WPI and Que. was mainly associated with non-covalent interactions. Specifically, the band around 3290 cm−1 in the amide A region gradually narrowed with increasing Que. content, which suggested changes in the hydrogen-bonding environment and the possible involvement of hydrogen bonding between WPI and Que. (Ozdal et al., 2013). In addition, the peak near 1649 cm−1 in the amide I region showed no obvious shift, whereas a slight red shift was observed near 1542 cm−1 in the amide II region, suggesting changes in C—N stretching and N—H bending vibrations (Kong & Yu, 2007). These spectral variations indicate that quercetin binding may induce conformational rearrangement of WPI and alter the microenvironment of polar and hydrophobic groups. The above results are shown in Fig. 3. FT-IR spectrogram of WPI-Que complexes. Structural changes of WPI after quercetin binding CD spectroscopy was further employed to evaluate changes of WPI in the secondary structure after quercetin binding. The CD spectra exhibited a positive peak near 197 nm and negative peaks around 209 and 220 nm, indicating the presence of α-helix structures. Notably, the positive peak near 197 nm was more intense in C10 complex than in the other samples, suggesting that an appropriate amount of Que. affected the spatial conformation of WPI. The above results are shown in Fig. 4. The fitted results indicated changes in the relative contents of α-helix, β-sheet, β-turn, and random coil structures after quercetin binding. In particular, the β-sheet content showed a decreasing trend with increasing Que. content, as shown in Table 2. Similar structural changes have been reported in whey protein–quercetin systems (Cheng et al., 2023). CD spectrogram of WPI-Que complexes. Table 2Secondary structures of WPI-Que complex.Complexα-helix/%β-sheet/%β-turn/%Random/%C049.715.310.124.9C1034.611.023.131.3C2052.22.113.532.1C4053.70.06.939.4C8059.00.89.131.1 Secondary structures of WPI-Que complex. Emulsifying properties of WPI–Que complexes Emulsifying activity is defined as the maximum amount of oil that can be emulsified by a fixed amount of the protein, while emulsion stability refers to the ability of an emulsion to resist phase separation (Pearce & Kinsella, 1978), both of which are positively associated with the affinity and adsorption capacity of emulsifiers at the oil–water interface (Yan et al., 2022). The EAI and ESI of the WPI–Que complexes were higher than those of WPI alone (Fig. 5), notably, the C10 complex exhibited superior ESI (70.75 ± 0.58%) and satisfactory EAI (0.76 ± 0.01 m2/g). Coincidentally, the FT-IR determination results showed that WPI and Que. were interacted via hydrogen bound, concurrently, the conformation of WPI was changed by the interaction of Que. observed from CD analyses (as shown in 3.2). The non-covalent WPI–Que interactions and conformational rearrangement of WPI may facilitate the adsorption and organization of the oil–water interface (Wang et al., 2025; Xiao et al., 2025). The EAI and ESI of WPI-Que complexes. Within the same group, values bearing different letters are significantly different from each other (p < 0.05). Physicochemical characterization of CSO emulsions Based on the above results, the C10 complex was employed for preparing the CSO emulsion system (Fig. 6). This study further analyzed the factors affecting the system's stability, including zeta potential and particle size, which are commonly used to evaluate the stability of colloidal systems. The absolute zeta potential values of all samples exceeded 30 mV (Fig. 7). This indicated that sufficient electrostatic repulsion forces were present in these colloidal droplets (Zhang et al., 2025). In contrast, the particle size of these colloidal droplets showed an upward trend with the increasing oil content, which ranged from 472.30 ± 3.82 nm to 869.05 ± 40.66 nm (Fig. 7). A significant increase was observed between E₅ and E₆ (p < 0.05). This indicated that the ratio of CSO to emulsifier complex solution at 5:10 was close to the saturation point of emulsifier coverage. The fine particle size of colloidal droplets may be attributed to the high emulsifier-to-oil ratio, which provided sufficient interfacial coverage for the oil droplets. As the oil content increased, the colloidal droplets were likely to aggregate, resulting in larger particle sizes (Lee et al., 2025). Optical microscopy showed a trend consistent with the particle size distribution data, and E5 displayed a relatively uniform size distribution (Fig. 8). To further elucidate this, the interfacial structure of the E5 emulsion was characterized using CLSM (Fig. 9). The continuous structures were observed, where the oil phase (colored yellow) was encapsulated by the WPI (colored red). This confirmed that the WPI–Que complex successfully formed an intact film at the oil–water interface. CLSM images suggested that the WPI–Que complex evenly surrounded the oil droplets. It synergistically worked with the electrostatic repulsion (zeta potential >30 mV) to prevent droplet flocculation (Song et al., 2025). Consequently. The E5 emulsion is the optimal formulation among all the samples tested. Appearance diagram of CSO emulsions. The zeta potential and particle size of CSO emulsions. Values bearing different letters are significantly different from each other (p < 0.05). Images of CSO emulsions monitored by optical microscopy. Scale bar: 20 μm. CLSM images of the CSO emulsion (E5) stained with Nile blue A (A), stained with Nile red (B), stained with Nile blue A and Nile red (C). The WPI (red) and oil phase (yellow) were stained by Nile blue A and Nile red, respectively. Scale bar: 20 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) The stability of the optimized CSO emulsion under severe pH and ionic strength To evaluate the environmental stability of the optimized CSO emulsion (E5), we further determined the zeta potential, particle size and appearance of the E5 emulsion under severe pH and ionic strength (various pH and ionic strengths). The zeta potential showed a decreasing trend as the pH increased from 2.0 to 10.0, and the emulsion presented the lowest absolute zeta potential (3.81 ± 0.42 mV) at pH 4. Concurrently, the particle sizes ranged from 659 ± 40.73 nm to 1059.35 ± 84.78 nm, and the emulsion was unstable at pH 4 (Fig. 11 a). The isoelectric point of WPI is close to pH 4, under which electrostatic repulsion between droplets is reduced (Shang et al., 2023), this promoted the aggregation of droplets observed from apparent phenomenon observation experiment (Fig. 10 a). In contrast, the absolute zeta potential was significantly affected by increasing ionic strength, all of which were below 20 mV, suggesting that salt ions screened the electrostatic repulsion between emulsion droplets (Harnsilawat et al., 2006). Nevertheless, the apparent phenomenon observation experiment showed that these emulsions still exhibited stable behavior over the entire tested NaCl concentration range (50, 150 and 200 mmol/L), as shown in Fig. 10 b. These results suggested that the stability of the emulsion system was not solely determined by electrostatic repulsion but also involved steric effects associated with the protein–polyphenol interfacial layer, as proposed by Tao et al. (Tao et al., 2023). Similarly, the particle size was positively affected by incerasing ionic strength, which reached a maximum (1109.63 ± 52.98 nm) at 100 mmol/L NaCl (Fig. 11 b), which can be attributed to the reduced electrostatic repulsion that normally prevents particle aggregation (Zhu et al., 2020). While, the particle size of emulsions showed a decreasing trend as the NaCl content ranged from 100 mmol/L to 200 mmol/L, this may ascribe to interfacial packing and the balance between electrostatic and steric interactions (Srinivasan et al., 2000; Zhang et al., 2022). The evolution of particle size and zeta potential of E5 emulsion under severe pH and ionic strength. Within the same group, values bearing different letters are significantly different from each other (p < 0.05). Apparent diagram of E5 emulsion under severe pH (a) and ionic strength (b), the numbers bearing the upper part of these samples indicate the pH of the corresponding sample (a) and the NaCl content (mmol/L) of the corresponding sample(b). Apparent bioaccessibility of CSO emulsions evaluated by FFA release In vitro gastrointestinal digestion models are widely used to assess the bioaccessibility of lipids. During gastrointestinal digestion, triacylglycerols are hydrolyzed to release FFAs and monoacylglycerols (Martins et al., 2024), both of which interact with bile salts to form mixed micelles, thereby contributing to lipid bioaccessibility (Wang, Li, et al., 2022). Since lipid digestion mainly occurs in the intestinal phase under the action of pancreatic enzymes and bile salts, the release of FFAs during intestinal digestion was monitored to evaluate the effect of emulsification on lipid digestion. The FFA release curves of bulk CSO and the CSO emulsion are shown in Fig. 12. The CSO emulsion showed a higher apparent FFA release (71.79 ± 0.98%) than the bulk CSO (53.74 ± 1.15%) after intestinal digestion at 120 min. The content of released FFAs from CSO emulsion was consistently higher than that of the CSO over the whole digestion. Herein, we further studied the apparent bioaccessibility of bulk CSO and CSO emulsions using mathematical fitting methods. The release of fatty acids followed first-order kinetics, and the CSO emulsions showed a higher apparent bioaccessibility than bulk CSO. Specifically, the maximum FFA release (69.33 ± 0.99%) and lipolysis rate (0.0531 min−1) of CSO emulsions were higher than the corresponding values for bulk CSO (Table 3). Previous studies have suggested that emulsions can provide a larger oil–water interfacial area for lipase action (Liu et al., 2021; Wijaya et al., 2020). In addition, bile salts in the simulated intestinal tract promote FFA release. Specifically, they are rapidly adsorbed to the oil–water interface, leading to the displacement of proteins (WPI) from the interface, which facilitates lipase anchoring onto lipid droplets and consequently enhances lipid bioaccessibility (Luo et al., 2023). The curves of FFA release from bulk CSO and CSO emulsion under simulated digestion condition. Within the curve, values bearing different letters are significantly different from each other (p < 0.05). Table 3Kinetic parameters for FFA release from bulk CSO and the CSO emulsion.SampleFFA0 (%)FFA∞ (%)k (min−1)R2Bulk CSO1.66 ± 0.50b52.28 ± 1.50b0.04230.9969CSO emulsion5.76 ± 0.90a69.33 ± 0.99a0.05310.9982Values are expressed as mean ± fitting standard error. FFA release was fitted by the first-order kinetic model: FFAt=FFA∞−FFA∞−FFA0·e−kt. Within the same column, values bearing different letters are significantly different (p < 0.05). Kinetic parameters for FFA release from bulk CSO and the CSO emulsion. Values are expressed as mean ± fitting standard error. FFA release was fitted by the first-order kinetic model: FFAt=FFA∞−FFA∞−FFA0·e−kt. Within the same column, values bearing different letters are significantly different (p < 0.05). Conclusion The bulk CSO shows a low bioaccessibility, which limits its application in functional foods. This study designed a CSO emulsion, which was sustainedly stabilized by a WPI–Que complex consisted by 10 mg quercetin and 100 mL WPI solutions. The CSO emulsion showed favorable colloidal characteristics as the ratio of CSO to WPI–Que complex solution was at 5:10 (w/v), and the emulsion exhibited a higher bioaccessibility compared to bulk oil in terms of the maximum FFA release and lipolysis rate. These findings suggest that the WPI–Que complexes can be served as promising food-grade emulsifier for constructing CSO delivery system and improving the potential bioaccessibility of CSO. CRediT authorship contribution statement Wenxin Zha: Writing – original draft, Software, Methodology, Data curation. Qingqing Li: Writing – original draft, Software, Methodology, Data curation. Yu Yuan: Software, Data curation. Likang Qin: Supervision, Funding acquisition. Haiyan Zhong: Supervision, Investigation. Yong Zhu: Writing – review & editing, Validation, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization. Funding This study was financially supported by the projects of Guizhou Characteristic Forestry Program (Te Lin Yan 2020-19), the National Natural Science Foundation of China (32460410). 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 statement The raw experimental data can be provided by the corresponding author if requested by the publisher. References Associated Data Data Availability Statement The raw experimental data can be provided by the corresponding author if requested by the publisher.