Modulation of Fabrication and Nutraceutical Delivery Performance of Ovalbumin-Stabilized Oleogel-Based Nanoemulsions via Complexation with Gum Arabic

Protein–polysaccharide complexes, which involve Maillard-type protein–polysaccharide conjugates and electrostatic protein–polysaccharide complexes, have the potential to stabilize oleogel-based nanoemulsions for nutraceutical delivery. Here, ovalbumin (OVA) and gum arabic (GA) were used to prepare OVA–GA conjugate (OGC) and OVA–GA mixture (OGM), followed by the fabrication of astaxanthin-loaded oleogel-based nanoemulsions. Carnauba wax (5% w/w) and rice bran oil were mixed to prepare food-grade oleogel. The successful preparation of OGC was verified by means of SDS-PAGE analysis and free amino groups determination. OGC endowed oleogel-based nanoemulsions with smaller emulsion droplets and higher stability during 30-day storage, implying more outstanding emulsifying capability than OGM. Both OGC-stabilized nanoemulsions and OGM-stabilized nanoemulsions could enhance the extent of lipolysis and the bioaccessibility of astaxanthin compared with oleogel. Meanwhile, OGC exhibited significantly better than OGM, which indicated that OGC-stabilized oleogel-based nanoemulsions possessed more desirable nutraceutical delivery performance than OGM-stabilized oleogel-based nanoemulsions. This study may fill a gap in the influence of different protein–polysaccharide complexes on oleogel-based nanoemulsions and contribute to deeper insights about novel oleogel-based nanoemulsions for their applications in the food industry.


Introduction
Oleogel-based emulsions have been demonstrated to be the outstanding delivery systems for bioactive ingredients [1,2]. When compared with conventional oil-in-water emulsions (without oleogel), oleogel-based emulsions exhibit better long-term storage stability and freeze-thaw stability [3]. The desirable stability for easy storage and transportation, coupled with remarkable delivery performance of bioactive ingredients, has led to the recent increased scholarly attention on oleogel-based emulsions [4][5][6]. Nanoemulsions with high stability to particle aggregation and gravitational separation can significantly enhance the bioavailability of the encapsulated nutrients [7]. It is essential to design novel oleogel-based nanoemulsions with satisfactory nutraceutical delivery performance for the practical application of functional foods. Ovalbumin (OVA), which is over half of the entire protein in egg albumen, is a globular glycoprotein with an ideal amino acid balance [8]. Our preliminary experiments have indicated that OVA cannot stabilize oleogel-based nanoemulsions very well, although it is generally accepted that OVA possesses satisfactory interfacial and emulsifying properties [9]. There is thus a significant SDS-PAGE was performed in accordance with the methodology described previously, with some modifications [24]. The samples including native OVA and glycosylated OVA with different heat treatment times (24 h, 48 h and 96 h) were separated on a precast 12.5% acrylamide separating gel under a voltage of 120 V. Coomassie Brilliant Blue R-250 was applied to stain the gel sheet, which was then decolorized and recorded.

Zeta Potential
A Malvern nanoparticle size potentiometer (Zetasizer Nano zs90) was employed to examine the zeta potential of OGC heated for 96 h over the pH range of 2.0 to 9.0, and the OGM at corresponding pH values were measured as controls.

Free Amino Groups
An ortho-phthaldialdehyde (OPA) method was applied to measure the free amino groups of native OVA and OGC, respectively [25]. OPA (80 mg), 0.1 M sodium tetraborate solution (50 mL), methanol (2 mL), 20% w/w sodium dodecyl sulfate (5 mL) and mercaptoethanol (200 µL) were used to prepare the OPA reagent. Afterwards, 2 mg/mL protein solution (200 µL) and OPA reagent (4 mL) were mixed and then incubated at 35 • C for 2 min. Finally, the absorbance at 340 nm of the samples was examined and the concentration of free amino groups was obtained by means of an L-leucine calibration curve. Figure 1 shows the preparation process of oleogel-based nanoemulsions. Astaxanthin (0.2% w/w) was dissolved in the oleogel structured by 5% w/w carnauba wax under heating to prepare the astaxanthin-loaded oleogel. OGC solutions (20-50 mg/mL) at pH 2.0 were added to the formed oleogel to obtain mixed systems (oil fraction ϕ = 0.9). Subsequently, the mixed systems were sonicated using a SM-900A ultrasonic processor (Nanjing Shunma Instrument Equipment Co., Ltd., Nanjing, China) for 3 min and the frequency was fixed at 20 kHz. The influence of OGC concentrations on the properties of oleogel-based nanoemulsions was also investigated. The same experimental conditions were employed to prepare the oleogel-based nanoemulsions using the OGM solutions instead of the OGC solutions.

Droplet Size of Oleogel-Based Nanoemulsions
The droplet size of oleogel-based nanoemulsions was measured by means of an Intelligent laser particle sizer (Bettersize 2600, Dandong Bettersize Instruments Ltd., Dandong, China). The samples were diluted 100 times using pH-adjusted ultrapure water (pH 2.0) prior to measurement.

Storage Stability Analysis
The obtained oleogel-based nanoemulsions were stored at a refrigerator temperature of 4 °C. The storage stability of the samples was determined by observing and recording the sample during 30-day storage.

Digestion of Oleogel-Based Nanoemulsions
The in vitro gastrointestinal digestion of oleogel-based nanoemulsions was carried out using the method described previously with slight modification [26]. Sodium chloride (2 g) was dissolved in 1 L of pH-preset ultrapure water (pH 1.2) to prepare simulated gastric fluid (SGF). The samples containing 2 g of oleogel were added to SGF (16 mL). The gastric digestion process was started with the addition of SGF (4 mL) containing freshly dissolved pepsin (32 mg). After 120 min at 37.0 ± 0.1 °C, the digestion was terminated by regulating the pH of digesta to 7.5. During the gastric digestion process, CaCl2 (10 mM), bile salt (10 mg/mL) and Tris (50 mM) were dissolved to pH-preset ultrapure water (pH 7.5) for obtaining simulated intestinal fluid (SIF). An equal volume of SIF containing pancreatin (3.2 mg/mL) was added to the gastric digesta for initiating the intestinal digestion process. Over the 120-min incubation period at 37.0 ± 0.1 °C, NaOH solution (0.25 M) was used to keep pH at 7.5 and the volume of NaOH was recorded. Thereafter, the samples were centrifuged at 10,000× g for 40 min to obtain the clear micelle phase.
The release of free fatty acids (FFA) could serve as the parameter to characterize lipolysis of digestion samples. It is generally accepted that 2 mol of FFA could be released from 1 mol of triglycerides and 2 mol of NaOH was thus consumed. The fraction of FFA released (% FFA) was calculated according to Equation (1) [26]: The droplet size of oleogel-based nanoemulsions was measured by means of an Intelligent laser particle sizer (Bettersize 2600, Dandong Bettersize Instruments Ltd., Dandong, China). The samples were diluted 100 times using pH-adjusted ultrapure water (pH 2.0) prior to measurement.

Storage Stability Analysis
The obtained oleogel-based nanoemulsions were stored at a refrigerator temperature of 4 • C. The storage stability of the samples was determined by observing and recording the sample during 30-day storage. The in vitro gastrointestinal digestion of oleogel-based nanoemulsions was carried out using the method described previously with slight modification [26]. Sodium chloride (2 g) was dissolved in 1 L of pH-preset ultrapure water (pH 1.2) to prepare simulated gastric fluid (SGF). The samples containing 2 g of oleogel were added to SGF (16 mL). The gastric digestion process was started with the addition of SGF (4 mL) containing freshly dissolved pepsin (32 mg). After 120 min at 37.0 ± 0.1 • C, the digestion was terminated by regulating the pH of digesta to 7.5. During the gastric digestion process, CaCl 2 (10 mM), bile salt (10 mg/mL) and Tris (50 mM) were dissolved to pH-preset ultrapure water (pH 7.5) for obtaining simulated intestinal fluid (SIF). An equal volume of SIF containing pancreatin (3.2 mg/mL) was added to the gastric digesta for initiating the intestinal digestion process. Over the 120-min incubation period at 37.0 ± 0.1 • C, NaOH solution (0.25 M) was used to keep pH at 7.5 and the volume of NaOH was recorded. Thereafter, the samples were centrifuged at 10,000× g for 40 min to obtain the clear micelle phase.
The release of free fatty acids (FFA) could serve as the parameter to characterize lipolysis of digestion samples. It is generally accepted that 2 mol of FFA could be released from 1 mol of triglycerides and 2 mol of NaOH was thus consumed. The fraction of FFA released (% FFA) was calculated according to Equation (1) [26]: where M Rice bran oil was the molecular mass of the rice bran oil (in g/mol), V NaOH was the volume of NaOH solution (0.25 M) added to the digesta (in L), m NaOH was the molarity of NaOH (in mol/L), w Rice bran oil was the total mass of initially present rice bran oil (in g). The average molecular mass of rice bran oil was taken as 864 g/mol [27].

High-Performance Liquid Chromatography (HPLC) Analysis of Astaxanthin
Astaxanthin content in the micelle phase was examined with an Agilent 1260 HPLC system (Agilent Technologies Inc., Santa Clara, CA, USA) with a YMC-C18 (YMC CO., Ltd., Kyoto, Japan) chromatography column (4.6 mm × 250 mm, 5 µm). The HPLC mobile phase was the combination of (A) methanol and (B) methyl tert-butyl ether. The elution condition was set as follows: 0-10 min, 90% A; 10-30 min, 40% A; 30-40 min, 90% A. The flow velocity was set as 1 mL/min and the sample volume was 30 µL. Quantification of astaxanthin according to a standard curve of astaxanthin was performed at a wavelength of 476 nm.

Determination of Astaxanthin Bioaccessibility
Astaxanthin bioaccessibility was calculated by the following equation after HPLC measurement of astaxanthin in the micelle phase: %bioaccessibility = astaxanthin content in the micelle phase total astaxanthin content in theformulations × 100% (2)

Statistical Analysis
All determinations were carried out in triplicate and all statistical analysis was performed by means of OriginPro 2021.

Formation and Characterization of Oleogel
Various contents of carnauba wax were mixed with rice bran oil to identify the critical gelling concentration. As shown in Figure 2, the obtained rice bran oil samples could flow at the carnauba wax concentration of 2% w/w, 3% w/w and 4% w/w, whereas the food-grade oleogel was successfully obtained at the carnauba wax concentration of 5% w/w because of immobility at room temperature. The formation of oleogel was attributed to crystallization as well as the coalescence of lipids, thereby restricting the movement of rice bran oil as the conventional high-melting-point solid fats [3]. Therefore, 5% w/w was regarded as the critical gelling concentration of carnauba wax-based oleogel and the oleogel was applied to the subsequent experiments in this research.  After the fabrication of OGC by means of a controlled dry heating method, SDS-PAGE analysis was performed in order to confirm the formation of covalent linking and to measure the molecular weight changes of the samples [28]. As depicted in Figure [29]. Moreover, the increased diffusion of the bands for OGC (lane three, lane four and lane five) was possibly due to the complex covalent OGC that one protein molecule was grafted with more than one polysaccharide [30]. Furthermore, a positive correlation was found between dry heating time and glycation extent, since the longer dry heating time of 96 h (lane 5) endowed the OGC with the significantly larger molecular weight of the protein compared with the shorter dry heating times of 24 h (lane 3) and 48 h (lane 4). Therefore, the relatively long dry heating time was conducive to the process of Maillard reaction and the fabrication of OGC.

Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE)
After the fabrication of OGC by means of a controlled dry heating method, SDS-PAGE analysis was performed in order to confirm the formation of covalent linking and to measure the molecular weight changes of the samples [28]. As depicted in Figure [29]. Moreover, the increased diffusion of the bands for OGC (lane three, lane four and lane five) was possibly due to the complex covalent OGC that one protein molecule was grafted with more than one polysaccharide [30]. Furthermore, a positive correlation was found between dry heating time and glycation extent, since the longer dry heating time of 96 h (lane 5) endowed the OGC with the significantly larger molecular weight of the protein compared with the shorter dry heating times of 24 h (lane 3) and 48 h (lane 4). Therefore, the relatively long dry heating time was conducive to the process of Maillard reaction and the fabrication of OGC.

Free Amino Groups
Considering that the above SDS-PAGE analysis indicated that OGC heated for 96 h had a higher glycation extent than OGC heated for 24 h or 48 h, OGC heated for 96 h was

Free Amino Groups
Considering that the above SDS-PAGE analysis indicated that OGC heated for 96 h had a higher glycation extent than OGC heated for 24 h or 48 h, OGC heated for 96 h was chosen for the rest of this study. The amounts of free amino groups of OGC and OVA control were measured in order to identify the involvement of free amino groups in covalent linking between free amino acids of OVA and the carbonyl group of GA, thereby evaluating the glycation extent of OVA [31]. As shown in Figure 3B, the number of free amino groups in OGC dropped by 40.59% compared with that of OVA. The reduction of free amino groups was ascribed to the Maillard reaction between the free amino groups of the OVA and the reducing carbonyl groups of GA [32]. Since the amino acid residues on the protein surface or specific glycation "hot-spots" had priority at glycation modification [33], the surface and specific sites of OVA were glycated by dry heating treatment. The reaction sites of OVA including lysine residues and a small portion of arginine residues were grafted with the carbonyl group during the initial stages of the Maillard reaction [34]. It should be pointed out that the free amino groups of OVA included not only ε-NH 2 of 15 arginine residues and 20 lysine residues but also α-NH 2 in the N-terminus, which was favorable for the process of the Maillard reaction [35].

Zeta Potential
The zeta potential of OGM and OGC was measured to explore the influence of the Maillard reaction on the surface charge of OVA and the change of OVA conformation [36]. Figure 4 shows the zeta potential as a function of pH. The OGC presented a left shift (from On one hand, more negatively charged amino groups (such as -OH and -COOH) were exposed to the environment by the Maillard reaction [37]. On the other hand, the blocking of positively charged amino groups might also be the reason for the decrease of the isoelectric point according to the previous results [26]. According to the results of the free amino groups, it could be speculated that the latter was an essential factor in explaining the experimental phenomenon of the decreased isoelectric point. As indicated by the results of zeta potential measurement, pH 2 was chosen to serve as the satisfactory environment to fabricate oleogel-based nanoemulsions because OGC had the relatively high zeta potential at pH 2.
NH2 of 15 arginine residues and 20 lysine residues but also α-NH2 in the N-terminus, which was favorable for the process of the Maillard reaction [35].

Zeta Potential
The zeta potential of OGM and OGC was measured to explore the influence of the Maillard reaction on the surface charge of OVA and the change of OVA conformation [36]. Figure 4 shows the zeta potential as a function of pH. The OGC presented a left shift (from around pH 5.0 to pH 4.0) in the isoelectric point compared with OGM due to the Maillard reaction. There are two possible explanations for the result. On one hand, more negatively charged amino groups (such as -OH and -COOH) were exposed to the environment by the Maillard reaction [37]. On the other hand, the blocking of positively charged amino groups might also be the reason for the decrease of the isoelectric point according to the previous results [26]. According to the results of the free amino groups, it could be speculated that the latter was an essential factor in explaining the experimental phenomenon of the decreased isoelectric point. As indicated by the results of zeta potential measurement, pH 2 was chosen to serve as the satisfactory environment to fabricate oleogel-based nanoemulsions because OGC had the relatively high zeta potential at pH 2.

Droplet Size of Oleogel-Based Nanoemulsions
Given that high-internal phase emulsions (HIPEs) with an oil fraction (φ) above 0.74 possessed high nutraceutical loading [38], the oleogel-based nanoemulsions (φ = 0.9) were designed and prepared in this research. Nanoemulsions are the non-equilibrium emulsion systems with mean droplet size between 50 and 1000 nm [39]. The droplet sizes of oleogelbased nanoemulsions at different particle concentrations (20-50 mg/mL) were measured to compare OGM-stabilized nanoemulsions and OGC-stabilized nanoemulsions, thereby investigating the effect of the Maillard reaction on the interfacial properties of reaction products. As shown in Figure 5, the droplet sizes of the two oleogel-based nanoemulsions

Droplet Size of Oleogel-Based Nanoemulsions
Given that high-internal phase emulsions (HIPEs) with an oil fraction (ϕ) above 0.74 possessed high nutraceutical loading [38], the oleogel-based nanoemulsions (ϕ = 0.9) were designed and prepared in this research. Nanoemulsions are the non-equilibrium emulsion systems with mean droplet size between 50 and 1000 nm [39]. The droplet sizes of oleogelbased nanoemulsions at different particle concentrations (20-50 mg/mL) were measured to compare OGM-stabilized nanoemulsions and OGC-stabilized nanoemulsions, thereby investigating the effect of the Maillard reaction on the interfacial properties of reaction products. As shown in Figure 5, the droplet sizes of the two oleogel-based nanoemulsions both exhibited a significant decrease with the increase of particle concentrations from 20 mg/mL to 50 mg/mL, which was ascribed to the decline of interfacial tension as the emulsifier concentration increased [40]. Moreover, the OGC-stabilized nanoemulsions had smaller emulsion droplets than OGM-stabilized nanoemulsions at each particle concentration under the same condition, indicating that the Maillard reaction endowed the oleogel-based nanoemulsions with superior emulsifying capability compared to physical mixing. This phenomenon could be explained as follows. First, the exposure of hydrophobic groups in OVA because of the Maillard reaction led to their enhanced affinity for the oil phase. Second, the extension of the hydrophilic carbohydrate moieties into the aqueous phase could hinder droplet aggregation via electrostatic repulsion or steric hindrance [41]. Third, the flexibility of OVA such as the conformational rearrangement of the tertiary protein structure possibly increased due to the Maillard reaction, which might promote the adsorption of OGC at the water-oil interface [42]. As depicted in Figures S1 and S2, the particle size distributions of OGM-stabilized nanoemulsions were broad, while that of OGC-stabilized nanoemulsions were relatively narrow. Consequently, Maillard-type protein-polysaccharide conjugates exhibited relatively desirable interfacial properties compared with electrostatic proteinpolysaccharide complexes.
Third, the flexibility of OVA such as the conformational rearrangement of the tertiary pro-tein structure possibly increased due to the Maillard reaction, which might promote the adsorption of OGC at the water-oil interface [42]. As depicted in Figures S1 and S2, the particle size distributions of OGM-stabilized nanoemulsions were broad, while that of OGC-stabilized nanoemulsions were relatively narrow. Consequently, Maillard-type protein-polysaccharide conjugates exhibited relatively desirable interfacial properties compared with electrostatic protein-polysaccharide complexes. Figure 5. Average droplet size of freshly prepared OGM-stabilized oleogel-based nanoemulsions (NE) and the OGC-stabilized oleogel-based NE (oil fraction φ = 0.9) at different particle concentrations (20-50 mg/mL) at room temperature.

Storage Stability of Oleogel-Based Nanoemulsions
In addition to the droplet size analysis of oleogel-based nanoemulsions, their visual observation was performed to further investigate the emulsifying properties of Maillardtype protein-polysaccharide conjugates and electrostatic protein-polysaccharide complexes. Figure 6 shows that the oleogel-based nanoemulsions at relatively low OGM concentration (20 mg/mL and 30 mg/mL) exhibited an obvious creaming phenomenon. Although the higher OGM concentration of 40 mg/mL could effectively alleviate the problem, the oleogel-based nanoemulsion at the OGM concentration of 50 mg/mL displayed an oiling-out behavior. On the contrary, the oleogel-based nanoemulsions at different OGC concentrations varying from 20 mg/mL to 50 mg/mL had no serum phase, implying better emulsifying properties than OGM. Since storage stability was an important indicator to evaluate the emulsion quality [43], the visual images of oleogel-based nanoemulsions after 30 days of storage at 4 °C were recorded as shown in Figure 6. Regrettably, different heights of bottom serum layer in the oleogel-based nanoemulsions at lower OGM concentrations (c = 20-40 mg/mL) were observed after 30 days of storage, indicating that the phase separation phenomenon became more apparent. It was noteworthy that the turbidity of the serum layer increased with the rise of OGM concentration, which might be ascribed to increased OGM concentration in the serum layer [44]. With regard to OGC-stabilized nanoemulsions, no serum layer emerged during storage. Combined with the results of droplet size analysis, the OGC-stabilized nanoemulsions with smaller droplets

Storage Stability of Oleogel-Based Nanoemulsions
In addition to the droplet size analysis of oleogel-based nanoemulsions, their visual observation was performed to further investigate the emulsifying properties of Maillard-type protein-polysaccharide conjugates and electrostatic protein-polysaccharide complexes. Figure 6 shows that the oleogel-based nanoemulsions at relatively low OGM concentration (20 mg/mL and 30 mg/mL) exhibited an obvious creaming phenomenon. Although the higher OGM concentration of 40 mg/mL could effectively alleviate the problem, the oleogel-based nanoemulsion at the OGM concentration of 50 mg/mL displayed an oilingout behavior. On the contrary, the oleogel-based nanoemulsions at different OGC concentrations varying from 20 mg/mL to 50 mg/mL had no serum phase, implying better emulsifying properties than OGM. Since storage stability was an important indicator to evaluate the emulsion quality [43], the visual images of oleogel-based nanoemulsions after 30 days of storage at 4 • C were recorded as shown in Figure 6. Regrettably, different heights of bottom serum layer in the oleogel-based nanoemulsions at lower OGM concentrations (c = 20-40 mg/mL) were observed after 30 days of storage, indicating that the phase separation phenomenon became more apparent. It was noteworthy that the turbidity of the serum layer increased with the rise of OGM concentration, which might be ascribed to increased OGM concentration in the serum layer [44]. With regard to OGC-stabilized nanoemulsions, no serum layer emerged during storage. Combined with the results of droplet size analysis, the OGC-stabilized nanoemulsions with smaller droplets possessed stronger and more cohesive interfacial membranes, leading to superior stability against film fracture and gravitational separation. Meanwhile, the low frequency and rate of inter-droplet collision at 4 • C resulted in the decline of droplet size growth [45]. But one drawback that arose during storage was the slight oiling-out behavior of the oleogel-based nanoemulsions at lower OGC concentrations (c = 20-40 mg/mL). Whereas the relatively high OGC concentration of 50 mg/mL effectively alleviated the problem, implying that enough OGC concentration was significantly conducive to desirable storage stability of oleogel-based nanoemulsions for application in the food industry. Accordingly, Maillard-type protein-polysaccharide conjugates displayed better performance to stabilize oleogel-based nanoemulsions than electrostatic protein-polysaccharide complexes.
nanoemulsions at lower OGC concentrations (c = 20-40 mg/mL). Whereas the relatively high OGC concentration of 50 mg/mL effectively alleviated the problem, implying that enough OGC concentration was significantly conducive to desirable storage stability of oleogel-based nanoemulsions for application in the food industry. Accordingly, Maillardtype protein-polysaccharide conjugates displayed better performance to stabilize oleogelbased nanoemulsions than electrostatic protein−polysaccharide complexes.

Lipolysis and Bioaccessibility of Astaxanthin in Oleogel-Based Nanoemulsions
In view of the fact that emulsion droplets with a relatively small size and a large surface area could facilitate the rapid release of nutraceuticals in the gastrointestinal tract [7], it could be speculated that an OGC-stabilized nanoemulsion with a smaller droplet size had a better nutraceutical delivery performance than the OGM-stabilized nanoemulsion. Figure 7A shows that the fractions of free fatty acids (FFA) released in both OGCstabilized nanoemulsion (64.5%) and OGM-stabilized nanoemulsion (45.9%) were higher than that in oleogel (8.6%). The extent of lipolysis for OGC-stabilized nanoemulsion with relatively small emulsion droplets was significantly enhanced because of the increased contact area between the pancreatin and the triglycerides [46]. Moreover, the rate of lipolysis rapidly rose during the initial stages of in vitro intestinal digestion but gradually decreased during the digestion for both oleogel-based nanoemulsions. This phenomenon was ascribed to the aggregation of the digestion products and the reduced ratio of enzymes to oil [47].

Lipolysis and Bioaccessibility of Astaxanthin in Oleogel-Based Nanoemulsions
In view of the fact that emulsion droplets with a relatively small size and a large surface area could facilitate the rapid release of nutraceuticals in the gastrointestinal tract [7], it could be speculated that an OGC-stabilized nanoemulsion with a smaller droplet size had a better nutraceutical delivery performance than the OGM-stabilized nanoemulsion. Figure 7A shows that the fractions of free fatty acids (FFA) released in both OGC-stabilized nanoemulsion (64.5%) and OGM-stabilized nanoemulsion (45.9%) were higher than that in oleogel (8.6%). The extent of lipolysis for OGC-stabilized nanoemulsion with relatively small emulsion droplets was significantly enhanced because of the increased contact area between the pancreatin and the triglycerides [46]. Moreover, the rate of lipolysis rapidly rose during the initial stages of in vitro intestinal digestion but gradually decreased during the digestion for both oleogel-based nanoemulsions. This phenomenon was ascribed to the aggregation of the digestion products and the reduced ratio of enzymes to oil [47]. In vitro bioaccessibility was measured to investigate the nutraceutical delivery performance of oleogel-based nanoemulsions. As depicted in Figure 7B, the astaxanthin bioaccessibility of either OGC-stabilized nanoemulsion (42.7% ± 2.7%) or OGM-stabilized nanoemulsion (33.1% ± 1.9%) was higher than that of oleogel (5.5% ± 0.7%), indicating that In vitro bioaccessibility was measured to investigate the nutraceutical delivery performance of oleogel-based nanoemulsions. As depicted in Figure 7B, the astaxanthin bioaccessibility of either OGC-stabilized nanoemulsion (42.7% ± 2.7%) or OGM-stabilized nanoemulsion (33.1% ± 1.9%) was higher than that of oleogel (5.5% ± 0.7%), indicating that the oleogel-based nanoemulsions displayed outstanding nutraceutical delivery performance. Given that FFA promoted the formation of mixed micelles and the micelles could be used to solubilize astaxanthin, the release of astaxanthin could be regulated through the controlled lipolysis of oleogel-based nanoemulsions [48]. The structural destruction of oleogel-based nanoemulsions led to the exposure of the astaxanthin hydrophobic core, thereby contributing to the transfer of astaxanthin into the mixed micelles and improving astaxanthin bioaccessibility [49]. In addition, contrary to an inhomogeneous appearance of OGM-stabilized nanoemulsions, the OGC-stabilized nanoemulsion was homogeneous on the whole. This phenomenon indicated that OGC was a desirable emulsifier to stabilize oleogel-based nanoemulsions for nutraceutical delivery when compared with OGM. Furthermore, Figure 7 shows that the astaxanthin bioaccessibility positively correlated with both the extent of lipolysis and the visual observation of oleogel-based nanoemulsions. This result systematically demonstrated that oleogel-based nanoemulsions with outstanding nutraceutical delivery performance were successfully fabricated by Maillard-type OVA-GA conjugate in this study.

Conclusions
In summary, the OVA-GA conjugate (OGC) and the OVA-GA mixture (OGM) were prepared to stabilize food-grade oleogel-based nanoemulsions. The comparisons between electrostatic protein-polysaccharide complexes and Maillard-type protein-polysaccharide conjugates were subsequently performed. As indicated by the results, the OGC-stabilized nanoemulsions had smaller emulsion droplets and superior storage stability than OGMstabilized nanoemulsions, indicating that OGC had more desirable emulsifying capability to stabilize oleogel-based nanoemulsions than OGM. The astaxanthin-loaded OGC-stabilized nanoemulsions, which could substantially improve the extent of lipolysis and astaxanthin bioaccessibility, exhibited a relatively homogeneous appearance, unlike the astaxanthinloaded OGM-stabilized nanoemulsions. The outstanding nutraceutical delivery performance may endow OGC-stabilized oleogel-based nanoemulsions with great potential in the application of functional foods. This research focusing on the comparisons between electrostatic protein-polysaccharide complexes and Maillard-type protein-polysaccharide conjugates may also offer insights to the design, characterization and application of proteinpolysaccharide complexes.