Preparation and functional characterization of buckwheat protein glycosylation products

Abstract

Buckwheat protein (BP) exhibits a nutritionally balanced amino acid profile and demonstrates multiple bioactive properties, although certain functional limitations restrict its broader application. This study systematically optimized the reaction conditions for both wet/dry-heat glycosylation between BP and Fagopyrum esculentum polysaccharide (FEP), characterized the effects of these conjugation methods on BP's functional properties, and successfully employed the resulting bioconjugates for vitamin D₃ (VD₃) encapsulation. The data demonstrate that the wet glycosylation method achieved a degree of grafting (DG) of 17.72 % for BP, which significantly enhanced both in vitro protein digestibility and viscoelastic properties. The dry glycosylation reaction yielded a DG of 16.53 %, resulting in a marked improvement in bile acid–binding capacity. Emulsions prepared with the glycosylated bioconjugates showed markedly improved VD₃ stability. These findings collectively suggest that glycosylation modification represents an effective approach for simultaneously enhancing both the functional characteristics of BP and the stabilization of VD₃.

Highlights

• •Optimized conditions for wet/dry glycosylation of buckwheat protein.
• •Wet glycosylation products with improved digestion and rheological properties.
• •Dry glycosylation products have higher bile acid binding capacity.
• •Glycosylation product emulsions significantly improve vitamin D₃ stability.

1. Introduction

Buckwheat is a gluten-free pseudocereal exhibiting grain-like chemical composition and applications (L. Wu et al., 2021; J. Zhang et al., 2023). Buckwheat protein (BP) contains all eight essential amino acids at levels meeting FAO/WHO recommendations for children and adults, confirming its nutritional adequacy (D. Li et al., 2023; F. Zhu, 2021). Additionally, BP contains low levels of alcohol-soluble gluten proteins, making it a suitable cereal-alternative food source for celiac disease patients. To date, the reported biological activities of BP include cholesterol-lowering, antihypertensive, antibacterial, and antioxidant properties. As a high-quality plant-derived protein, BP exhibits significant potential for functional food development (F. Zhu, 2021). The amino acid profile of BP comprises approximately 40 % hydrophobic residues (including non-polar and aromatic amino acids), which impart distinct amphiphilic characteristics that enhance complexation with lipophilic bioactive compounds (Yu et al., 2022). Compared with soy protein, BP has not been fully used in the food industry because of its insufficient functional characteristics (emulsification, gelation, and foaming) (Jin, Ohanenye, & Udenigwe, 2022). While BP demonstrates promising potential, further research is required to optimize its physicochemical and functional properties for widespread application as a plant-based protein alternative in food product development.

The Maillard reaction initiates with glycosylation, a non-enzymatic condensation between reducing sugars and free amino groups. First characterized by Louis-Camille Maillard in 1912, this reaction represents the foundational step in the complex Maillard reaction cascade that ultimately leads to food browning (Chiang, Eyres, Silcock, Hardacre, & Parker, 2019). The early and intermediate stages of the Maillard reaction induce structural modifications that generate bioactive glycosylation products capable of enhancing protein functionality. However, advanced Maillard reaction stages promote the formation of undesirable off-flavors and potentially harmful polymeric compounds (De Oliveira, Coimbra, & d. R., de Oliveira, E. B., Zuñiga, A. D. G & Rojas, E. E. G., 2016). Polysaccharides can effectively modulate the Maillard reaction by serving as alternative glycosylation agents to reducing sugars. Their structural characteristics (including high molecular weight and limited reducing ends) restrict reaction advancement while concurrently improving protein functional properties (Sheng et al., 2020). Under appropriate temperature and water activity conditions, glycosylation can occur spontaneously, promoting covalent coupling between adjacent protein and polysaccharide molecules. The extent of reaction between proteins and carbohydrates can also be characterized by the degree of grafting (DG). Proteins can form covalent bonds with polysaccharides, which consume free amino acids within the proteins. Consequently, the DG can be calculated by monitoring the alterations in the free amino acid content within the reaction system (Chen et al., 2025).

Natural polysaccharides can be extracted from animal, plant, and microbial sources, with plants representing the most abundant source of food-grade polysaccharides. A defining characteristic of plant polysaccharides is their structural stability, conferred by strong intermolecular interactions. These properties (including structural rigidity and pronounced hydrophilicity) enable plant polysaccharides to significantly influence the textural and stability characteristics of food formulations (Shao et al., 2020). Furthermore, Fagopyrum esculentum polysaccharide (FEP) represents a key bioactive component of buckwheat, demonstrating multiple physiological functions such as hypoglycemic, antitumor, immunomodulatory, and antioxidant activities. In recent years, this multifunctional compound has garnered significant attention in the fields of food science, nutrition, and medicine (X.-T. Wang et al., 2016). Zemnukhova, Sukhoverkhov, Shkorina, Kovekhova, and Tomshich (2007) found that the differences in the structural characteristics of polysaccharides extracted from buckwheat might be due to the variations in extraction, separation, and purification methods, as well as the use of different types of buckwheat. Previous studies have demonstrated that the quantity of reducing groups may significantly influence carbohydrate reactivity (Chailangka et al., 2022). In the present study, the FEP employed exhibited a notably high reducing sugar content of 23.15 % (w/w), the high steric hindrance of FEP controls the reaction rate, and the high content of reducing sugar is conducive to the progress of the reaction. In food processing systems containing reducing sugars and protein components, this reaction is inevitable. When the drying parameters of buckwheat flour are not properly controlled, glycosylation is easily induced, which leads to damage of the natural functional characteristics of BP and simultaneously promotes the Maillard browning reaction, generating undesirable pigmentation that affects the sensory quality of the product. To address these issues, this study utilized endogenous FEP (rather than exogenous sugars) for specific glycoconjugation with BP, aiming to elucidate the mechanisms underlying processing-induced quality deterioration of BP while fully exploiting the intrinsic affinity between homologous biomolecules for protein modification.

Vitamin D₃ (VD₃) is an essential fat-soluble vitamin. Its deficiency can lead to rickets in children and osteomalacia in adults (Ajdžanović et al., 2024). VD₃ deficiency remains a major public health issue worldwide, chiefly caused by limited sunlight exposure that impairs cutaneous VD₃ production (Gupta et al., 2019). Moreover, VD₃ is environmentally labile and undergoes oxidative degradation, compromising its bioavailability (T. Wu, Liu, & Hu, 2022). Unprotected VD₃ undergoes rapid degradation under gastric conditions. To improve its physicochemical stability, aqueous solubility, and oral bioavailability, numerous encapsulation systems have been explored for the targeted delivery of VD₃ (Maurya, Shakya, Bashir, Jan, & McClements, 2023). The formation of electrostatic complexes between proteins and polysaccharides has attracted considerable research attention for emulsion stabilization. These biopolymer complexes demonstrate superior resistance to environmental stresses compared to emulsions stabilized solely by proteins or polysaccharides. Furthermore, the protein components can effectively bind VD₃, thereby enhancing both its stability and bioavailability (Delavari et al., 2015).

This thesis focuses on BP as the research subject, employing FEP through wet/dry glycosylation for protein modification. The study systematically investigated the effects of different glycosylation methods on the DG, optimized the respective reaction conditions, characterized the functional properties of the resulting conjugates, and successfully utilized these bioconjugates to prepare emulsion-based delivery systems for VD₃ encapsulation and stabilization. This study aims to expand the applications of polysaccharides in modifying the processing characteristics of BP, thereby establishing a theoretical foundation for enhancing BP's functional properties. The research provides both theoretical frameworks and technical references to facilitate the broader utilization of BP in the food industry.

2. Materials and methods

2.1. Materials

Buckwheat was obtained from local markets in Changchun, China. Fagopyrum esculentum polysaccharide (Purity >70 %) was sourced from Lanzhou Weite Biotechnology Co., Ltd. (Gansu, China). Soybean oil was purchased from a Changchun supermarket (Jilin Province, China). Glucose assay kit (Beijing Boxbio Science & Technology Co., Ltd., Beijing), BCA protein quantification kit (Shanghai Yuanye Bio-Technology Co., Ltd., China), DPPH radical scavenging capacity assay kit (Shanghai Yuanye Bio-Technology Co., Ltd., China), Hydroxyl radical scavenging assay kit (Shanghai Yuanye Bio-Technology Co., Ltd., China), Total antioxidant capacity (T-AOC) assay kit (Shanghai Yuanye Bio-Technology Co., Ltd., China). Pancreatic enzymes (8 × USP specifications) and pepsin (≥ 2500 units/mg protein) were obtained from Sigma-Aldrich (St. Louis, MO, USA). All of the chemicals and reagents used in our experiments were analytical grade.

2.2. Preparation of BP

Buckwheat flour was defatted by mixing with petroleum ether (1:5 w/v) under continuous stirring for 4 h at room temperature. The degreased flour was then suspended in deionized water (1:10 w/v), and the pH was adjusted to 10.0 using 1 M NaOH. The alkaline solution was centrifuged at 8000 rpm for 20 min (TLXJ-JIB centrifuge, Nanjing Anting Scientific Instrument Co., Shanghai). The collected supernatant was acidified to pH 4.5 with 1 M HCl and stored at 4 °C overnight to facilitate protein precipitation. The resulting precipitate was neutralized to pH 7.0 with distilled water and lyophilized (LGE-50E freeze dryer, Sihuan Furey Scientific Instrument Co., Beijing) to obtain purified BP. (The protein purity was determined to be 88.520 % using the Kjeldahl nitrogen determination method.)

2.3. Preparation of glycosylation product

10 mg/mL BP solution was prepared in PBS and then mixed with FEP at predetermined mass ratios. The mixture was homogenized by vortex mixing (4500 rpm) at 25 °C for 1 h to form the mixed BP-FEP solution. This solution was lyophilized to obtain the BP-FEP mixture (Mix). The mixed solution was adjusted to the predetermined pH using 0.1 M NaOH/HCl. Glycosylation was conducted by incubating the solution in a precision-controlled water bath at a set temperature and time under constant agitation. The reaction was immediately quenched in an ice-water bath upon completion, followed by lyophilization. The freeze-dried product was ground through a stainless steel sieve to obtain the wet-glycosylated BP-FEP conjugate (designated as W-BP-FEP). The reaction mixture was transferred to a glass Petri dish, and the relative humidity (79 %) was maintained using a saturated KBr solution. The system was incubated in a constant-temperature dryer at a set temperature and time. Upon completion, the reaction was immediately terminated by ice-bath quenching. The product was subsequently lyophilized to obtain the dry-glycosylated BP-FEP conjugate (designated D-BP-FEP).

2.4. Determination of the DG

The DG was determined using the ortho-phthalaldehyde (OPA) method, with modifications adapted from Jiang et al. (2025). The OPA solution was prepared by dissolving 80 mg of OPA powder in 2 mL of anhydrous ethanol, combining it with 50 mL of 0.1 mol/L sodium tetraborate solution, 5 mL of 20 % (w/v) sodium dodecyl sulfate, and 200 μL of β-mercaptoethanol. The mixture was then topped up to a final volume of 100 mL with distilled water. The samples were then diluted, and 200 μL of the samples were collected and mixed with 4 mL of OPA reagent. The prepared mixtures were subsequently incubated in a water bath set at 35 °C for a duration of 2 min. Afterward, the absorbance of these mixtures was determined at 340 nm using a UV spectrophotometer (TU-1901, Puritan General Instrument Co., Ltd., Beijing, China).

$$\mathrm{DG}\left(\%\right)=\left(\frac{\text{The absorbance of sample before reaction}-\text{the absorbance of the grafts}}{\text{The absorbance of sample before reaction}}\right)\times 100$$ (1)

2.5. Optimization of glycosylation reaction conditions

The glycosylation reaction was performed following the protocols outlined in Sections 2.3. Initial single-factor experiments were conducted to evaluate the effects of glycoprotein ratio, temperature, time, and pH on the DG (Chen et al., 2025; Z. Li et al., 2024). Based on these results, the Box-Behnken design was implemented in response surface methodology (RSM) to avoid extreme-condition protein denaturation, with the DG serving as the primary optimization parameter. The experimental design matrix is presented in Table 1.

Table 1.

Single factor test factors and levels.

	Factor
	Sugar-protein ratio	Temperature(°C)	Time(min)	pH
	3:1	30	60	5
	2:1	45	90	7
Level	1:1	60	120	9
	1:2	75	150	11
	1:3	90	180	13

2.6. Evaluation of the bioactive properties of glycosylated products

2.6.1. Swelling capacity

The swelling capacity of the samples was determined using the method described by Cui et al. (2024). Samples (10 mg) were hydrated in 1 mL water, and vortexed. After boiling and rapid cooling, centrifugation (2000 rpm, 5 min) was performed. After carefully removing the supernatant, the precipitate mass was determined by gravimetric analysis.

$$\text{Swelling capacity}\left(\mathrm{g}/\mathrm{g}\right)=\frac{\mathit{Sediment\; quality}}{\mathit{Sample\; quality}}\times 100$$ (2)

2.6.2. Fat-binding capacity

Fat-binding capacity was quantified according to the gravimetric method of Davoudi, Azizi, and Barzegar (2022), with the following modifications: 5 mg sample was homogenized in 1 mL of soybean oil (vortex mixing, 1 min), followed by continuous shaking (1,500 rpm, 30 min). After 30 min of static incubation, the mixture was centrifuged (2000 rpm, 5 min). The supernatant was carefully removed, and the oil-bound pellet mass was measured gravimetrically.

$$\text{Fat}-\text{binding}\text{capacity}\left(\mathrm{g}/\mathrm{g}\right)=\frac{\mathit{Sediment}\mathit{quality}}{\mathit{Sample}\mathit{quality}}\times 100$$ (3)

2.6.3. Bile acid - binding capacity

The bile acid-binding capacity was determined according to the in vitro method of Doubilet (1936). 200 mg aliquot of cholic acid was dissolved in 4.7 mL of 0.1 M NaOH and diluted to 200 mL. 25 mg of the sample was incubated with 10 mL of this cholic acid solution at 37 °C for 2 h with continuous stirring. Subsequently, 1 mL of the reaction mixture was combined with an equal volume of furfural reagent (1:1 v/v), followed by the addition of 5 mL concentrated sulfuric acid (98 %) with vigorous vortexing. After 5 min incubation in an ice bath, the solution was heated at 70 °C for 5 min in a water bath, then immediately cooled on ice for another 5 min. The absorbance was measured at 490 nm using an enzyme meter (Spectramax 190 BMG Labtech Company, Germany).

2.6.4. DPPH radical scavenging activity

25 mg sample was extracted with 400 μL of nitrogen radical extract at 40 °C for 30 min in a temperature-controlled water bath. The mixture was centrifuged at 10,000 rpm for 10 min using a refrigerated centrifuge. Three experimental sets were prepared in parallel: Blank control tube was prepared by combining 100 μL nitrogen radical extract with equal volumes (450 μL each) of absolute ethanol and DPPH solution. Sample test tube was prepared by combining 100 μL of sample solution with 450 μL of absolute ethanol and 450 μL of DPPH solution. Sample control tube was prepared by mixing 100 μL of sample solution with 900 μL of absolute ethanol. All procedures followed the DPPH Radical Scavenging Activity Kit. After vortex mixing, the reaction systems were incubated at 25 °C under light-protected conditions for 30 min. Absorbance was measured at 517 nm using an enzyme marker.

$$\text{DPPH radical scavenging activity (\%)=}\frac{{\mathrm{A}}_0-\left({\mathrm{A}}_1-{\mathrm{A}}_2\right)}{{\mathrm{A}}_0}\times 100$$ (4)

Here: A₀ is the absorbance value of the Blank control; A₁ is the absorbance value of the Sample test; A₂ is the absorbance value of the Sample control.

2.6.5. OH radical scavenging ability

The hydroxyl radical scavenging activity was determined according to the Hydroxyl Radical Scavenging Assay Kit. Five reaction systems were prepared in parallel: Blank control tube was prepared by combining 0.2 mL ·OH buffer with 0.8 mL distilled water. Undamaged control tube was prepared by sequentially adding 0.15 mL 1,10-phenanthroline, 0.2 mL ·OH buffer, 0.1 mL ferrous chromogenic agent, and 0.55 mL distilled water. Oxidatively damaged tube was prepared by mixing 0.15 mL 1,10-phenanthroline, 0.2 mL ·OH buffer, 0.1 mL ferrous chromogenic agent, and 0.45 mL distilled water. Sample control tube was prepared with 0.2 mL ·OH buffer, 0.7 mL distilled water, and 0.1 mL sample solution. Sample assay tube was prepared by combining 0.15 mL 1,10-phenanthroline, 0.2 mL ·OH buffer, 0.1 mL ferrous chromogenic agent, and 0.35 mL distilled water, followed by the addition of 0.1 mL sample solution and 0.1 mL oxidant. Subsequently, 300 μL aliquots from each reaction tube were transferred to a 96-well microplate in triplicate. Absorbance was measured at 510 nm using an enzyme marker.

$$\text{The}\mathrm{OH}\text{radical scavenging ability }\left(\%\right)=\frac{\left({\mathrm{A}}_3-{\mathrm{A}}_4\right)-\left({\mathrm{A}}_2-{\mathrm{A}}_0\right)}{{\mathrm{A}}_1-{\mathrm{A}}_2}\times 100$$ (5)

Here: A₀ is the absorbance value of the Blank control tube; A₁ is the absorbance value of the Undamaged control tube; A₂ is the absorbance value of the Oxidatively damaged tube; A₃ is the absorbance value of the Sample assay tube; and A₄ is the absorbance value of the Sample control tube.

2.6.6. ABTS radical scavenging ability

The ABTS radical scavenging capacity was quantified using a Total Antioxidant Capacity Assay Kit. The ABTS working solution was prepared by mixing equal volumes (1:1 v/v) of ABTS stock solution and oxidant solution, followed by dark incubation at 25 °C for 12 h. For analysis, 280 μL of diluted ABTS working solution was combined with 7 μL of sample solution (40:1 v/v) in a 96-well microplate. After 5 min of reaction at 25 °C under lucifuge conditions. The absorbance of the sample at 405 nm was measured by an enzyme marker.

$$\text{ABTS radical scavenging ability}\left(\%\right)=\frac{{\mathrm{A}}_1-{\mathrm{A}}_0}{A_1}\times 100$$ (6)

Here: A₁ is the absorbance of ABTS working solution and distilled water; A₀ is the absorbance of ABTS working solution and sample.

2.7. In vitro digestion

The digestive properties of the glycosylation products were characterized using an in vitro simulated gastrointestinal digestion model following the methodology of Zhang et al. (2023). The gastric digestion phase was initiated by pre-warming simulated gastric fluid (SGF, pH 3.0) to 37 °C in a water bath. An equal volume (1:1 v/v) of sample solution and preheated SGF were homogenized by vortex mixing and incubated at 37 °C for 2 h with continuous shaking. Enzymatic activity was terminated by boiling (100 °C, 5 min), followed by immediate cooling on ice. For intestinal digestion, the gastric digestate was adjusted to pH 7.0 using 0.1 M NaOH, then mixed with an equal volume of simulated intestinal fluid (SIF, pH 7.0). The mixture was incubated under identical conditions before final enzyme inactivation by boiling and ice-cooling.

2.7.1. Digestibility measurement

The digestive fluid was centrifuged at 5000 rpm for 10 min at 4 °C. The protein content of the resultant supernatant was quantified using the BCA protein quantification kit.

$$\text{Digestibility}\left(\%\right)=\frac{\mathit{Sample\; protein\; content}-\mathit{Prote}\mathrm{i}n\mathit{content\; in\; supernatant}}{\mathit{Sample\; protein\; content}}\times 100$$ (7)

2.7.2. Determination of free amino acid content

The free amino acid content of the digestive samples was quantified using the ninhydrin assay. Briefly, an equal volume (1:1 v/v) of sample digest and ninhydrin solution (2 % w/v) was mixed and heated at 100 °C for 15 min. After cooling to room temperature, the absorbance was measured at 570 nm using an enzyme marker. A standard curve was prepared using glycine (0–100 μg/mL) for quantification.

2.7.3. Determination of peptide content

An equal volume of 10 % (w/v) trichloroacetic acid (TCA) was added to the sample digest and vortex-mixed for 30 s. The mixture was allowed to stand at 25 °C for 10 min to facilitate protein precipitation, followed by centrifugation at 5000 rpm for 10 min at 4 °C. The peptide content in the resultant supernatant was subsequently quantified using the BCA protein quantification kit.

2.8. Glucose availability

The glucose availability was performed according to the method described by Rodríguez et al. (2008). 500 mg sample and glucose standard were co-dissolved in 10 mL of 0.1 M PBS (pH 7.4). The digestion process was performed identically in Section 2.7 of this work. Following digestion, the supernatant was collected by centrifugation (5000 rpm, 10 min, 4 °C) for subsequent glucose quantification via a Glucose assay kit.

2.9. Gel characterization

2.9.1. Preparation of gels

The protein gels were prepared using a thermal induction method as described by Sun et al. (2024). 10 % (w/v) sample solution was prepared by continuous stirring for 2 h at 25 °C to ensure complete dissolution, followed by overnight hydration at 4 °C. The solution was then transferred to sealed vials and heated at 90 °C for 30 min in a temperature-controlled water bath. Subsequently, the samples were rapidly cooled to 25 °C using an ice-water bath and stored at 4 °C for 12 h to facilitate gel formation.

2.9.2. Rheological property

The rheological properties of the gel were characterized using a rheometer (MCR302, Anton Paar GmbH, Australia) following the methodology established by Li et al. (2025). The shear-dependent viscosity of the gels was measured over a shear rate range of 0.1–10 s⁻¹ using a rheometer. Dynamic oscillatory measurements were conducted to determine the storage modulus (G') and loss modulus (G") within a frequency range of 0.1–10 Hz at 25 °C, with a fixed strain amplitude of 1 %.

2.9.3. Moisture distribution

Gel samples were subjected to LF–NMR analysis (MESOMR23–040–VI, Newman Technology Co., Ltd., Shanghai) at 32 °C. Samples were evaluated for moisture distribution via T₂ relaxation measurements using the CPMG sequence.

2.10. Preparation of VD₃-loaded emulsions using glycosylation products

A 2 % (w/w) glycosylated product solution was prepared in PBS and adjusted to pH 7.0 using 0.1 M NaOH/HCl. VD₃ was first dissolved in soybean oil. The oil phase (VD₃-enriched soybean oil) was then mixed with the aqueous phase (glycosylated product solution) at a 20 % oil fraction (ϕ = 0.2). Emulsification was performed using a high-speed homogenizer (T25, IKA, Germany) under varying conditions for single-factor optimization: rotational speed (8000–12,000 rpm in 1000 rpm increments) and duration (2–10 min in 2–min intervals).

2.10.1. VD₃ loading

The encapsulation efficiency of VD₃ was determined according to the method described by Zhu et al. (2024). Briefly, 1 mL of the emulsion was aliquoted into a 10 mL centrifuge tube, followed by sequential extraction steps: Ethanol (2 mL) was added and vortex-mixed, n-Hexane (2 mL) was introduced and gently shaken, Centrifugation at 4000 rpm for 2 min, The upper n-hexane phase was carefully pipetted, This extraction process was repeated three times. The collected upper organic phase was analyzed by measuring absorbance at 265 nm using a UV spectrophotometer (TU-1901, Puritan General Instrument Co., Ltd., Beijing, China). The encapsulation efficiency was calculated as:

$${\mathit{VD}}_3\text{loading}\left(\%\right)=\frac{V{D}_3\text{content in}\text{the}\text{oil}\mathrm{phase}}{{\mathit{VD}}_3\text{content in lotion}}\times 100$$ (8)

2.10.2. Turbidity

The absorbance of the emulsion was measured at 600 nm using an enzyme marker.

2.10.3. Centrifugal stability

10 mL aliquot of the emulsion was transferred to a graduated 15 mL conical centrifuge tube, and the initial total height was measured using a digital caliper. Centrifugation was performed at 5000 rpm for 5 min. The height of the separated supernatant layer was subsequently measured under identical conditions. The creaming index (CI) was calculated as:

$$\mathrm{CI}\left(\%\right)=\frac{\text{Height}\text{of supernatant}\text{after}\text{centrifugation}}{\text{Total height of the emulsion sample}}\times 100$$ (9)

2.10.4. Potential and particle size

The Zeta potential and particle size distribution of the emulsions were analyzed using a dynamic light scattering system (Zetasizer Pro, Malvern Panalytical, UK). Measurements were conducted at 25.0 ± 0.5 °C with three independent replicates per formulation.

2.10.5. Environmental stability of VD₃

2.10.5.1. pH stability test

Different pH buffer solutions were prepared and mixed with the emulsion at a 1:1 (v/v) ratio. The mixtures were shaken at 25 °C for 1 h, followed by VD₃ content analysis.

2.10.5.2. Ionic strength test

The emulsion was combined 1:1 (v/v) with NaCl solutions of varying concentrations, and the VD₃ content was quantified.

2.10.5.3. Thermal stability test

To assess thermal stability, the emulsion was incubated at 4 °C, 30 °C, and 90 °C for 5 h, with aliquots withdrawn hourly for VD₃ analysis.

2.10.5.4. Photostability test

For photodegradation assessment, the emulsion was exposed to UV light for 5 h, with samples collected at 1 h intervals for VD₃ quantification.

2.11. In vitro simulated digestion of emulsions

The digestive behavior of the emulsion was evaluated using an established in vitro digestion model. Post-digestion VD₃ retention was analyzed according to the method described in Section 2.10.1. Emulsion droplet characteristics, including mean particle size and Zeta potential, were determined following the analytical protocol outlined in Section 2.10.4.

2.12. Statistical analysis

All data are presented as mean ± standard deviation (SD), with at least three independent replicates per group. Significance was determined by t-test or one-way analysis of variance (ANOVA) (P < 0.05), with analyses performed using GraphPad Prism 9.5.0 and Origin 2022.

3. Results and discussion

3.1. Results of a one-factor test for glycosylation reactions

The results of the single-factor experiment were presented in Fig. 1(A-D). Glycosylation is intrinsically a base-catalyzed reaction. However, extreme pH conditions (either excessive acidity or alkalinity) induce protein conformational changes, while improper reactant concentrations (either too high or too low) reduce molecular collision frequency, thereby decreasing the DG. Reaction temperature and duration similarly influence DG because low temperatures limit amino acid surface accessibility, whereas high temperatures cause protein denaturation. Prolonged reaction times may promote protein aggregation, which adversely affects polysaccharide conjugation. Consequently, systematic optimization of glycosylation conditions is essential for achieving optimal DG (Dai et al., 2023).

Fig. 1.

Effects of glycosylation reaction conditions on protein grafting efficiency. (A) Sugar-protein ratio, (B) pH, (C) Temperature, (D) Time. Data are mean ± standard deviation. Different superscript letters indicate significant differences (P < 0.05).

3.2. Glycosylation reaction response surface test results

As shown in Table 2, the constructed response surface model for glycosylation showed high statistical significance (P < 0.01). The non-significant P-value of the lack-of-fit term (P > 0.05) indicated good model adequacy, confirming that the model accurately represents the experimental data. Furthermore, these results demonstrated that all investigated variables significantly influence the DG. Analysis of variance for the glycosylation response surface model indicated the model's predictive capability for wet glycosylation conditions. The three-dimensional response surface plots and corresponding contour plots reveal significant interactive effects among process variables. Furthermore, the DG exhibited marked sensitivity to variations in reaction conditions, as evidenced by the pronounced curvature in the response surfaces (Fig. 2). Through response surface methodology optimization, this study established distinct optimal conditions for both wet and dry glycosylation processes. For the wet glycosylation process, the optimized parameters were: glycoprotein mass ratio of 1:1, reaction temperature of 75 °C, duration of 90 min, and pH 7.4. The model predicted a DG of 17.88 %, while experimental validation yielded 17.72 %. The dry glycosylation process showed optimal performance under different conditions: glycoprotein mass ratio of 1:1, temperature of 60 °C, reaction time of 150 min, and pH 9.3. The predicted DG (17.97 %) was slightly higher than the experimental value (16.53 %).

Table 2.

Protein glycosylation degree and response surface model analysis.

Source	square sum	degrees of freedom	mean square	F-value	P-value	Significance	square sum	mean square	F-value	P-value	Significance
Model	187.83	14	13.416	960.595	<0.001	significant	500.07	35.72	216.83	< 0.0001	significant
A	0.139	1	0.139	9.990	0.006	**	3.76	3.76	22.82	0.0003	**
B	0.304	1	0.301	21.765	0.0004	**	2.72	2.72	16.52	0.0012	**
C	19.096	1	19.096	1367.234	<0.0001	**	9.88	9.88	59.97	< 0.0001	**
D	5.135	1	5.135	367.659	<0.0001	**	2.78	2.78	16.85	0.0011
AB	0.006	1	0.006	0.441	0.5173		1.44	1.44	8.71	0.0105
AC	0.016	1	0.013	1.173	0.2970		0.9801	0.9801	5.95	0.0286
AD	2.507	1	2.507	179.524	<0.0001	**	0.0844	0.0844	0.5123	0.4859
BC	3.381	1	3.381	242.131	<0.0001	**	7.72	7.72	46.85	< 0.0001
BD	2.463	1	2.463	176.364	<0.0001	**	1.43	1.43	8.70	0.0106
CD	4.137	1	4.137	296.202	<0.0001	**	0.0858	0.0858	0.5211	0.4822
A2	98.926	1	98.926	7082.721	<0.0001	**	384.95	384.95	2336.83	< 0.0001
B2	18.838	1	18.808	1346.121	<0.0001	**	52.15	52.15	316.60	< 0.0001
C2	69.095	1	69.095	4946.949	<0.0001	**	111.42	111.42	676.37	< 0.0001
D2	31.117	1	31.117	2227.899	<0.0001	**	106.98	106.98	649.43	< 0.0001
Residual	0.195	14	0.013				2.31	0.1647
Lost proposal	0.107	10	0.010	0.487	0.8372		2.15	0.2154	5.66	0.6547
Pure terror	0.088	4	0.02203				0.1522	0.0381
Total error	188.238	28					502.38	502.38
R2	0.995						0.995
Radj2	0.991						0.990
Rpred2	0.996						0.974
C.V.%	1.16						5.72
Adeq Precision	99.7						43.92

** indicates highly significant difference (P < 0.01), * indicates significant difference (P < 0.05).

Fig. 2.

Box-Behnken Response Surface Analysis for Glycosylation: 3D and Contour Plots. (A) 3D and contour plots of wet glycosylation reaction characteristics, (B) 3D and contour plots of dry glycosylation reaction characteristics.

3.3. Characterization of biological activities in protein glycosylation products

The lipid-lowering effect is closely associated with fat-binding capacity, swelling properties, and cholesterol-lowering activity (Qin et al., 2021). However, glycosylation-induced reduction in swelling capacity may impair these functional attributes (Fig. 3A). This suggests that polysaccharide conjugation introduces additional hydroxyl groups, enhancing protein-water affinity and thereby increasing protein solubility while reducing swelling capacity. As shown in Fig. 3B, Glycosylation significantly enhanced fat-binding capacity, with increases of 131.69 % (W-BP-FEP) and 133.99 % (D-BP-FEP), attributable to the hydroxyl groups introduced by polysaccharide conjugation. Hydroxyl groups play a more important role in fat binding capacity (Qin et al., 2021). Moreover, the reduction in the hydrophobicity of the glycosylated product surface also confirmed the introduction of hydroxyl groups (Supplementary material 1). As shown in Fig. 3C, the glycosylation reaction significantly enhanced the bile acid binding capacity of BP (P < 0.05). This improvement may be attributed to the introduction of polar groups from polysaccharides during the reaction and the formation of a viscous macromolecular network through polysaccharide-water interactions, which could physically hinder bile acid reabsorption (Ragot, Russell, & Schneeman, 1992). Fig. 3D-E indicated that glycosylation significantly enhanced the radical scavenging capacity of BP. The unmodified BP demonstrated radical scavenging activities with the following efficiencies: 81.17 % for DPPH radical scavenging activity, 68.77 % for ABTS radical scavenging capacity, and 54.15 % for OH radical scavenging. The DPPH radical scavenging activity showed a significant increase (P < 0.05), which can be attributed to three key effects of FEP conjugation with BP amino acids: increased tertiary structural flexibility, exposure of previously buried amino acid residues, and enhanced hydrogen-donating capacity (Feng, Cai, Wang, Li, & Liu, 2018). Similarly, glycosylation improved ABTS radical scavenging capacity (P < 0.05) through protein structural unfolding that generated additional hydroxyl/reducing groups while enhancing both electron-and hydrogen-donating capabilities. Most notably, OH radical scavenging increased by 37.72 % for W-BP-FEP and 43.46 % for D-BP-FEP, achieved through dual mechanisms: polysaccharide-mediated hydrogen donation and Fe²⁺ chelation that suppresses radical generation (J. Wang et al., 2021). In addition, the formation of furans, ketones, heterocyclic compounds, and brown melanin-like substances during the thermal decomposition of Amadori compounds in the Maillard reaction enhances the antioxidant performance of the protein (Dong et al., 2023).

Fig. 3.

Characterization of biological activities in protein glycosylation products. (A) Swelling capacity, (B) Fat binding capacity, (C) Bile acid binding capacity, (D) ABTS radical scavenging ability, (E) DPPH radical scavenging activity, (F) OH radical scavenging ability. Different superscript letters indicate significant differences (P < 0.05).

3.4. Influence of glycosylation on digestive stability and glucose availability in vitro

As shown in Fig. 4A-C, glycosylation improved protein digestibility, peptide content, and free amino acid release, likely due to structural unfolding that facilitates greater accessibility to enzymatic cleavage sites. While some studies report increased digestibility post-glycosylation and others show decreases, these differences likely stem from variations in protein characteristics, polysaccharide size, and glycosylation-induced structural changes (Joubran, Moscovici, Portmann, & Lesmes, 2017). In this study, polysaccharide grafting via Maillard reaction disrupts the native compact/disulfide-linked structure of BP, facilitating protease access. Concurrently, FEP prevents aggregation and sustains solubility through hydrophilic interactions, promoting digestive enzyme activity. Glucose availability measurements are used to assess the sensitivity of a substance to glucose (De Moura et al., 2011). The significant reduction in glucose availability (Fig. 4D) correlates with rheological measurements (increased viscosity) and glucuronic acid presence (Zhao et al., 2017).

Fig. 4.

Effects of glycosylation on invitro digestibility, glucose bioaccessibility, and gelation properties of BP. (A) Digestibility, (B) Peptide content, (C) Free amino acid content, (D) Glucose availability, (E) Apparent viscosity, (F) Energy storage modulus, (G) Loss modulus, (H) Water distribution. Different superscript letters indicate significant differences (P < 0.05).

3.5. Gel characterization

As shown in Fig. 4D, the gel exhibited typical pseudoplastic fluid behavior within the shear rate range of 0.1–10 s⁻¹. This rheological property results from the shear-induced disruption of irregular polymer coils and their subsequent alignment parallel to the flow direction (Inglett, Chen, & Liu, 2015). Compared to native BP gels, the glycosylated products formed significantly more viscous gels. This enhancement results from the glycosylation-induced branching of BP molecules, which promotes intermolecular entanglement. Across the entire frequency range tested, all gel samples exhibited G' greater than G″, demonstrating frequency-dependent behavior characteristic of weak gel systems with dominant elastic properties (Mao, Ren, Ye, Kong, & Tian, 2023). Fig. 4E-F shows elevated G' and G″ in glycosylated gels versus native BP gels, indicating enhanced viscoelasticity from FEP-mediated hydrogen bonding, hydrophobic interactions, and disulfide bonds that stabilize the gel network (Sun et al., 2024). Table 3 demonstrated that the gel samples primarily contained free water (T₂₃). The glycosylated products showed significantly attenuated T₂₃ signal amplitudes (P < 0.05), indicating polysaccharide-mediated conversion of free water to bound states. This shift suggests enhanced molecular interactions between the gel network components and water molecules, resulting in a more tightly cross-linked gel structure.

Table 3.

Effect of glycosylation reactions on the water composition of buckwheat proteins.

Groups	PT21/%	PT22/%	PT23/%
BP	3.127 ± 0.023a	3.927 ± 0.102c	92.946 ± 0.112b
Mix	2.713 ± 0.072b	3.680 ± 0.082d	93.607 ± 0.108a
W-BP-FEP	2.809 ± 0.087b	6.911 ± 0.115a	90.280 ± 0.131d
D-BP-FEP	3.247 ± 0.095a	4.788 ± 0.103b	91.965 ± 0.141c

Data are mean ± standard deviation. Different superscript letters in the column indicate significant differences (P < 0.05).

3.6. Preparation of VD₃-loaded emulsions using glycosylated products

As shown in Fig. 5A-B, the encapsulation efficiency initially increased with processing time before reaching a plateau. Studies have shown that excessive homogenization speed reduces emulsion droplet size below optimal levels, accelerating Brownian motion-induced protein aggregation and consequently decreasing encapsulation rates (da Silva-Padilha, Oliveira Júnior, Francisco, & da Cunha, 2024). Maximum encapsulation efficiency was obtained at 11,000 rpm for 6 min, establishing these as optimal conditions for VD₃ incorporation.

Fig. 5.

Effects of glycosylated protein emulsions on VD₃ encapsulation efficiency and environmental stability. (A) Effect of homogenization time on VD₃ encapsulation rate, (B) Effect of homogenization speed on VD₃ encapsulation rate, (C) Emulsion ionic stability, (D) Emulsion 4 °C stability, (E) Emulsion 30 °C stability, (F) Emulsion 90 °C stability, (G) Emulsion pH stability (H) Emulsion UV stability. Data are mean ± standard deviation. Different superscript letters indicate significant differences (P < 0.05).

3.7. VD₃ emulsion properties

Table 4 demonstrated significantly higher VD₃ encapsulation efficiency in wet glycosylation product emulsions compared to their dry counterparts (P < 0.05). This enhanced performance likely results from superior emulsification properties inherent to the W-BP-FEP. Emulsion turbidity serves as an indicator of both particle size and colloidal stability. Larger emulsion particles generally exhibit higher turbidity but lower stability, while smaller particles demonstrate reduced turbidity with enhanced stability. CI values directly correlate with emulsion stability, where lower CI indicates greater stability (W. Wang et al., 2022). The VD₃-loaded emulsion prepared with wet glycosylation products exhibited lower turbidity and CI, indicating enhanced stability, despite having larger average particle sizes. This may be attributed to the higher encapsulation efficiency achieved through the wet glycosylation process. The research found that the embedding rate of the glycosylated product emulsion VD₃ was positively correlated with the particle size of the emulsion (Supplementary material 2). Moreover, the thick interfacial film formed by the covalent protein-polysaccharide can resist droplet coalescence.

Table 4.

Properties of VD₃ embedded in the emulsion of glycosylation products.

groups	embedding rate(%)	turbidity	CI(%)	particle size(μm)	Zeta potential(mV)	PDI index
Wet Glycosylation Product Emulsion	43.446 ± 0.417⁎⁎	3.627 ± 0.046⁎⁎	6.021 ± 0.232⁎	5.002 ± 0.032⁎⁎	−10.900 ± 0.213	0.217 ± 0.021⁎
Dry Glycosylation Product Emulsion	27.480 ± 0.328⁎⁎	4.772 ± 0.088⁎⁎	9.231 ± 0.261⁎	2.429 ± 0.072⁎⁎	−10.500 ± 0.382	0.873 ± 0.042⁎

Data are mean ± standard deviation. ** indicates highly significant difference (P < 0.01), * indicates significant difference (P < 0.05).

3.8. Emulsion environmental stability

VD₃ encapsulated in glycosylated emulsion systems exhibited significantly enhanced environmental stability compared to free VD₃ (P < 0.05). This improvement originates from three synergistic mechanisms: glycosylation-mediated protein restructuring that anchors hydrophobic moieties at the oil-water interface, forming a viscoelastic protective barrier; extended polysaccharide chains in the continuous phase that provide steric stabilization, and the resultant composite interface that effectively prevents droplet coalescence. These structural advantages collectively enhance VD₃ retention under various environmental stress conditions.

3.9. Stability of emulsion digestion

As shown in Table 5, the particle size significantly increased during simulated gastric digestion compared to the original emulsion (P < 0.05). This size enlargement resulted from three concurrent processes: pepsin-induced protein degradation disrupting the interfacial structure, droplet coalescence under acidic conditions, and protein flocculation that further promoted emulsion droplet aggregation. Following intestinal digestion, the emulsion particle size significantly decreased (P < 0.05), likely due to trypsin-mediated hydrolysis of interfacial proteins. This enzymatic degradation further disrupted the emulsion structure, leading to smaller droplet sizes. The emulsion exhibited a significantly higher absolute Zeta potential value after simulated digestion compared to the original emulsion (P < 0.05). This increase can be attributed to the presence of undigested lipid droplets and interfacial interactions between newly formed digestion products and residual components. The significantly higher VD₃ retention observed in dry glycosylation product emulsions after simulated digestion demonstrates their superior structural stability and protective capacity. This enhanced performance suggests the dry glycosylation process creates a more robust network architecture that effectively shields encapsulated VD₃ from digestive degradation.

Table 5.

Digestive stability of VD₃ encapsulated in glycosylated protein-stabilized emulsions.

		Simulates stomach stage digestion	Simulates intestinal stage digestion
Emulsion embedding of VD3 in W-BP-FEP	Retention rate(%)	49.901 ± 0.098⁎⁎	9.784 ± 0.032⁎⁎
	Zeta potential(mV)	−13.920 ± 0.982⁎	−23.000 ± 1.023⁎
	Particle size(μm)	8.012 ± 0.823	0.561 ± 0.281
	PDI index	0.300 ± 0.021	0.842 ± 0.023
Emulsion embedding of VD3 in D-BP-FEP	Retention rate(%)	56.823 ± 0.035⁎⁎	12.733 ± 0.087⁎⁎
	Zeta potential(mV)	−12.350 ± 0.921⁎	−17.980 ± 1.002⁎
	Particle size(μm)	6.726 ± 0.212	0.394 ± 0.281
	PDI index	0.627 ± 0.031	0.621 ± 0.072

Data are mean ± standard deviation. ** indicates highly significant difference (P < 0.01), * indicates significant difference (P < 0.05).

4. Conclusion

The experimental results demonstrate distinct functional enhancements through wet and dry glycosylation modifications of BP. Wet-processing glycosylation achieved a grafting efficiency of 17.72 %, accompanied by exceptional radical scavenging capacities: 96.40 % for DPPH, 74.59 % for hydroxyl, and 90.12 % for ABTS radicals. This modification concurrently enhanced fat-binding capacity by 75.07 %, reduced glucose availability by 55.72 %, and increased in vitro digestible peptide yield by 32.51 %. In contrast, dry-process glycosylation, with a grafting efficiency of 16.53 %, exhibited superior hydroxyl radical scavenging (77.67 %) and fat-binding capacity improvement (77.33 %), along with a 46.28 % reduction in glucose availability. Both modification protocols significantly enhanced the rheological performance of protein gels through the effective reduction of free water content. Most notably, glycosylated emulsion systems demonstrated statistically significant improvement in  VD₃ encapsulation stability. This study elucidates the impact of glycation on buckwheat proteins, providing a basis for their functional optimization.

CRediT authorship contribution statement

Qian Luo: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Yan Cui: Software, Investigation, Formal analysis. Weifan Gao: Validation, Software, Investigation. Wei Wang: Validation, Software, Investigation. Mingzhu Zheng: Writing – review & editing, Supervision, Project administration, Funding acquisition, Conceptualization. Jingsheng Liu: Supervision, Project administration, 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.