Extraction optimization and screening of antioxidant peptides from grass carp meat and synergistic–antagonistic effect

Abstract Grass carp (Ctenopharyngodon idellus) is one of the three most cultivated freshwater fish around the world, but it is mainly consumed afresh, so only a small part of them are processed into salted fish or snack food. This research was performed to prepare and screen antioxidant peptides from grass carp muscle to promote its high‐value utilization. The parameters of double‐enzyme two‐step hydrolysis were optimized, the peptides with the highest ABTS.+ scavenging ability were enriched and identified by Sephadex G‐25 and LC‐Q‐Orbitrap‐MS/MS. The synergistic–antagonistic effect among identified peptides was also investigated. The optimized conditions were hydrolyzed with protamex (10,000 U/g) at pH 8.0, 50°C for 3 h, followed by hydrolysis with alcalase (6,000 U/g) at pH 9.0, 50 °C for 2 h, and the protein–liquid ratio was 4%. The hydrolysates were further fractionated to obtain five fractions, in which fraction 3 (F3) exhibited the strongest ABTS.+ and O2·‐ scavenging ability with the IC50 values of 0.11 and 0.47 mg/ml, respectively. Twelve novel antioxidant peptides were identified, in which VAGW possessed the highest activity (139.77 μmol GSH/g). Significantly synergistic effects were observed on the two and three peptides’ combination among VAGW, APPAMW, LFGY, FYYGK, and LLLYK, while the C‐terminal tryptophan (Trp) played an important role in the synergism. This study found that grass carp muscle hydrolysates can be potential natural antioxidants in functional products. The synergistic effects among peptides may provide a perspective for the combined application of peptides.

indispensable to obtain additional protection to balance the oxidation state (Jaouad & Torsten, 2010), at this point, antioxidants become the primary option. Although synthetic antioxidants such as butylated hydroxytoluene, butylated hydroxyanisole, propyl gallate, and tertiary butylhydroquinone have good antioxidant activities, their use in diet is limited because of toxic and side effects (Lobo et al., 2010). Natural antioxidants, especially antioxidant peptides derived from food, have attracted widespread attention, since they can be isolated from countless sources and have the advantages of low side effects and high absorption (Sarmadi & Ismail, 2010).
Antioxidant peptides are usually composed of 2-20 amino acid residues, which can be released by enzymatic hydrolysis during gastrointestinal digestion or food processing. Up to date, a large number of antioxidant peptides have been isolated and identified from aquatic protein, for example Raja porosa cartilage (Pan et al., 2016), Pseudosciaena crocea muscle (Chi et al., 2015), Setipinna taty (Song et al., 2015), and fish gelatin (Zamorano-Apodaca et al., 2020).
However, as the majority of reported bioactive peptides were derived from seafood proteins, researches on the antioxidant peptides from freshwater fish are much less.

Grass carp (Ctenopharyngodon idellus), belonging to the family
Cyprinidae, is not only one of the seven major freshwater fish species in China (Yang et al., 2020), but it is also one of the four most cultivated freshwater fish around the world. The annual production of cultured grass carp in China exceeded 5.50 million tons in 2018 (China, 2019). According to the abundance in bioactive proteins and unsaturated fatty acids, grass carp is a traditionally high-quality resource . Its muscle and skin hydrolysates were reported to show angiotensin-I converting enzyme (ACE) inhibition (Yi et al., 2016) and antioxidant activities (Chen et al., 2016).
The antilisterial peptides derived from grass carp proteins can efficiently inhibit the growth of L. monocytogenes in surimi noodle (Xiao & Niu, 2015). Furthermore, a novel excellent ACE inhibitory peptide Val-Ala-Pro  and a potent antioxidant peptide Pro-Ser-Lys-Tyr-Glu-Pro-Phe-Val  were isolated from the grass carp protein hydrolysates prepared with alcalase. However, reports regarding the screening and characterization of antioxidant peptides from grass carp muscle are much less than skin.
Enzymatic hydrolysis is the most common method for preparing bioactive peptides because of the milder and controllable process, which includes single-, double-, and multi-enzyme hydrolysis, the latter two hydrolyses can be further divided into step-by-step and mixed hydrolysis (Sharma et al., 2020;Liu et al., 2016). Doubleenzyme hydrolysis possesses the advantages of more cleavage sites, higher hydrolysis degree, and simpler enzymatic hydrolysis process.
For example, the degree of hydrolysis (DH) of spirulina platensis protein catalyzed by alkaline and papain was 25.47% and 21.73%, respectively. It was increased to 32.90% when the protein was hydrolyzed by alkaline and papain sequentially (Sun et al., 2016). The 2,2-diphenyl-1-picrylhydrazyl (DPPH) scavenging ability of corn protein alkaline-flavourzyme two-step hydrolysates was 2.59-fold of that hydrolyzed by flavourzyme (Jin et al., 2016).
The purpose of this work was to optimize the double-enzyme two-step hydrolysis parameters of grass carp muscle using the ABTS. + scavenging ability and degree of hydrolysis (DH) as indicators, and to screen the antioxidant peptides via chromatography separation and LC-Q-Orbitrap-MS/MS. The identified peptides were synthesized to evaluate the antioxidant activity, and to analyze the synergistic and antagonistic effects. Finally, the relationship between chemical structure and antioxidant ability of tested peptides was analyzed. This work would provide technical and theoretical support for further utilization of grass carp proteins as potential natural antioxidants in functional products.

| Optimization of hydrolysis conditions
Protamex-alcalase stepwise hydrolysis was selected as an appropriate method based on the results of our pre-experiments, which showed the strongest ABTS. + scavenging ability compared with other double-enzyme combinations of alcalase, neutrase, flavourzyme, and protamex (shown in Figure S1).
Fresh grass carp meat collected from the back was minced, and the content of crude lipid (1.68 ± 0.55% [fresh weight]) was determined by the Soxhlet extraction method. Therefore, the minces were mixed with distilled water directly without degreasing treatment. This mixture was then hydrolyzed with protamex at an enzyme/substrate [E/S] ratio of 10,000 U enzyme/g protein, pH of 7.0, and temperature of 50°C for 3 h (Tkaczewska et al., 2020). Following, the second-step hydrolysis was operated with alcalase. According to the ABTS. + scavenging ability of hydrolysates, the parameters of alcalase hydrolysis, including initial the protein-liquid ratio (1%, 2%, 3%, 4%, 5%), alcalasesubstrate ratio [E/S] (6,000, 8,000, 10,000, 12,000, 14,000 U/g), hydrolysis temperature (0, 40, 50, 60, 70 °C), hydrolysis time (1, 2, 3, 4, 5 h), and pH (6,7,8,9,10), were compared to achieve the optimal dualenzyme stepwise hydrolysis conditions. The blank samples (that contained all of the reagents without grass carp meat) were prepared in parallel. The protein-liquid and alcalase-substrate ratios (U/g protein) were calculated based on the protein content in minces detected by Kjeldahl's method (Marcia & Sebranek, 1993). After enzymatic hydrolysis, the solutions were boiled for 10 min to inactivate the enzyme and centrifuged at 602 g for 10 min, and the supernatants were gathered and used for antioxidant ability analysis.

| Determination of the degree of hydrolysis
The degree of hydrolysis (DH) of all hydrolysates was calculated from the ratio of α-amino nitrogen to total nitrogen. The amino nitrogen content (X 1 ) was determined by the formaldehyde titration method (Lin et al., 2013). The total nitrogen content (X 2 ) was measured with the Kjeldahl method (Marcia & Sebranek, 1993). The DH was calculated according to the following equation:

| Amino acid composition analysis
The grass carp hydrolysates (GCHs) prepared with the optimal hydrolysis conditions were lyophilized and subjected to amino acid composition analysis according to a reported method with slight modifications (Siswoyo et al., 2011). The GCHs were hydrolyzed with 6 mol/L HCl in a hydrolysis tube under 110°C for 24 h. Then, the volume was adjusted to 25 ml with distilled water, and 1 ml of the mixed sample was dried under reduced pressure and redissolved in sodium citrate buffer solution (1.0 ml, pH 2.2). Finally, the sample was filtered through a 0.22μm membrane and subjected to an Automatic Amino Acid Analyzer (L-8900, Hitachi, Japan).

| Determination of molecular weight distribution
The molecular weight (MW) distribution of GCHs was determined using an Agilent 1260 Infinity II LC HPLC System (Agilent, Palo Alto, CA) equipped with a Waters XBridge Protein BEH 125 Å SEC column (3.5 μm, 7.8 × 300 mm). Samples were eluted with 40% acetonitrile con-

| Separation by gel filtration chromatography
The GCHs were dissolved in distilled water and separated on a Sephadex G-25 gel filtration column (Ф 1.6 cm × 80 cm) (Haofeng et al., 2021). The sample solution was loaded onto the preequilibrated column and eluted by ultrapure water at a flow rate of 0.4 ml/min. The elution was collected at 5-min intervals by an automated fraction collector and detected at 220 nm. Ultimately, five fractions were collected and freeze-dried to evaluate the ABTS. + scavenging capacity, O ⋅− 2 scavenging ability, and Fe 2+ chelating ability.

| Antioxidant ability analysis
The ABTS. + radical scavenging assays were carried out according to the methods reported by Yang et al. (2021). Sample solutions (50 μl) at suitable concentrations were reacted with 150 μl of freshly diluted ABTS. + solution in a 96-well microplate at 25°C for 30 min.
The absorbance (A i ) at 734 nm was measured by a microplate reader (BioTek, USA). GSH was applied as positive control. The percentage inhibition was calculated using the following formula: where Ab is the absorbance of blank group, Ac is the absorbance of control group, Ai is the absorbance of sample group, and Aib is the absorbance of sample blank group with radical replaced by distilled water. The concentration required to scavenge 50% of ABTS. + (IC 50 value) was expressed as mg/ml.
The O ⋅− 2 scavenging ability and Fe 2+ chelating ability of GCHs and their fractions were measured based on the methods reported by Guo et al. (2009) andHu et al. (2012), respectively, and calculated with Equation (2). GSH was used as positive control, while the concentration required to scavenge 50% of O ⋅− 2 or chelate 50% of Fe 2+ (IC 50 value) was expressed as mg/mL and used to evaluate the activity.

| Identification of peptide sequences
The peptide fraction exhibiting the strongest ABTS. + scavenging activity was used for further identification of amino acid sequence through a Nano-LC-Orbitrap-MS/MS system (Ma et al., 2003).
The mass data were acquired on an Orbitrap Q-Exactive mass spectrometer controlled by Xcalibur 2.2 SP1 software under positive ion mode. The mass spectrometry (MS) spectra were obtained at a resolution of 70,000 with the target value of 3e6 and scan range of m/z 250-1,350. Peptide fragmentation was performed via higherenergy collision dissociation (HCD), while MS/MS spectra were acquired at a resolution of 17,500 and a target value of 5e4. PEAKS Studio 7.0 software combined with de novo sequencing was used to process the MS/MS data. The identified peptides meet the false discovery rate (FDR) ≤ 5% and the average local confidence score (ALC) ≥ 95%.

| Peptide synthesis and synergistic effect analysis
In this research, 15 identified peptides were selected based on the structure-activity relationships of antioxidant peptides and synthesized by Jier Biotechnology Co. Ltd (Shanghai, China) Liu et al., 2016;Rajapakse et al., 2005). The purity of all synthesized peptides was over 95%. The ABTS. + scavenging ability of synthesized peptides was tested according to the method described above.
The synergistic or antagonistic effect of synthesized peptides was investigated according to the method of Becker et al. (2007).
The concentration of all peptides was fixed at 2 mg/ml, then one, two, three, or four peptides with different amino acid sequences were mixed in an equal volume to obtain the combined peptides.
The experimental value (EV) was expressed as GSH equivalent value (μmol GSH/mg peptide). The calculated value (CV) of combined peptides was calculated based on the average GSH equivalent value of each single peptide. Higher EV than CV indicates a synergistic effect, while lower EV implies an antagonistic effect.

| Statistical analysis
All samples were analyzed in triplicate, and the data were expressed as mean ± standard deviation (SD). Statistical analysis was carried out by one-way analysis of variance (ANOVA) and Tukey's test with SPSS version 17.0, p < .05 was regarded as significant.

| Optimization of alcalase second-step hydrolysis parameters
The mincing of grass carp was hydrolyzed first by protamex, and the hydrolysates were then subjected to further hydrolysis with alcalase. The hydrolysis parameters of protamex were chosen based on the previous report (Tkaczewska et al., 2020). According to the ABTS. + scavenging ability and DH, the protein-liquid ratio, alcalase addition, hydrolysis temperature, hydrolysis time, and pH for the second-step hydrolysis of alcalase were optimized subsequently to obtain the most suitable parameters, and the results are presented in Figure 1. Higher IC 50 value indicates lower radical scavenging ability, whereas a higher DH suggests better hydrolysis efficacy. However, simply positive or negative correlation between DH and ABTS. + scavenging ability was not observed among all single-factor experiments, except for the protein-liquid ratio. This suggests that higher DH does not indicate stronger radical scavenging ability. Therefore, ABTS. + scavenging ability was considered as the evaluation index to optimize parameters for screening antioxidant peptides. The IC 50 value decreased gradually when the protein-liquid ratio was increased from 1% to 4%, implying that a high protein-liquid ratio results in stronger ABTS. + scavenging ability (Figure 1a). But a decreasing trend was observed when the alcalase-substrate ratio was increased from 6,000 to 14,000 U/g P ( Figure 1b). Therefore, for the next experiments, the optimal protein-liquid ratio and alcalase-substrate ratio were 4% and 6,000 U/g P, respectively.
As for the hydrolysis pH, the ABTS. + scavenging ability of hydrolysates was improved with increasing pH. The lowest IC 50 value (0.10 mg P/mL) was detected at pH 10.0, but insignificant difference was observed between the IC 50 values of hydrolysates prepared at pH 9.0 and 10.0 ( Figure 1c). Thus, 9.0 was chosen as the suitable pH for the following experiments.
As shown in Figure 1d, the second-step hydrolysis with alca- Based on the stepwise optimization results of single-factor experiments, the optimal second-step hydrolysis parameters for alcalase were: protein-liquid ratio, 4%; pH, 9.0; enzyme-substrate ratio, 6,000 U/g; hydrolysis temperature, 50°C; time, 2 h. Finally, grass carp meat was first hydrolyzed with protamex at a proteinliquid ratio of 4%, enzyme/substrate ratio of 10,000 U/g, pH of 7.0, and temperature of 50°C for 3 h. Then, the pH was adjusted to 9.0, alcalase was added at an enzyme/substrate ratio of 6,000 U/g to start the second-step hydrolysis progress at 50°C for 2 h. After centrifugation at 782 g for 10 min, the supernatant was gathered and freeze-dried to obtain the grass carp hydrolysates (GCHs). The ABTS. + scavenging ability, O ⋅− 2 scavenging ability, and Fe 2+ chelating ability of GCHs were evaluated. An obvious dose-dependent relationship was observed in the three antioxidant models (data were not shown), and the calculated IC 50 values are shown in Table 1. The IC 50 value (0.21 mg/ml) for ABTS. + scavenging ability was much lower than that without optimization (0.31 mg/ml, Figure S1), suggesting a good optimization efficacy.
In addition, it was much lower than that of collagen peptides from tilapia skin (IC 50 = 2.51 mg/ml) (Sheng et al., 2018), and similar to that of polypeptides from yellowfin tuna (Thunnus albacares) head (IC 50 = 0.24 mg/ml) (Pu et al., 2018). The IC 50 values for O ⋅− 2 scavenging ability and Fe 2+ chelating ability were 5.60 and 2.47 mg/ml, respectively, suggesting a relatively weaker activity. But, as the

| Amino acid composition of hydrolysates
The amino acid content expressed as mg/100 g GCHs is shown in Table 2. The content of hydrophobic, acidic, basic, and aromatic amino acids was 21.60, 16.66, 11.07, and 7.03 g/100 g sample, respectively. The hydrophobic amino acid accounted for 28.05% of total amino acids, among which leucine (Leu), alanine (Ala), and F I G U R E 1 Effects of the protein-liquid ratio (a), alcalase-substrate ratio (b), hydrolysis pH (c), hydrolysis time (d), and hydrolysis temperature (e) on the ABTS. + scavenging capacity and degree of hydrolysis of grass carp two-step hydrolysates

| Molecular weight distribution of hydrolysis
The size exclusion chromatogram and molecular weight (MW) distribution curve (B) of standards are shown in Figure 2a,b, respectively.
The retention time and log (lg) MW were applied to obtain calibra-

| Antioxidant ability of GCHs fractions
The GCHs were fractionated using Sephadex G-25 gel filtration column to enrich the peptides with high antioxidant ability. Totally, five fractions were collected, lyophilized, and labeled as F1, F2, F3, F4, and F5 orderly (Figure 3a). Then, all fractions were redissolved in distilled water and used to evaluate the ABTS. + scavenging ability, O ⋅− 2 scavenging ability, and Fe 2+ chelating ability, and the results are listed in Table 1.
The F3 exhibited the strongest ABTS. + and O ⋅− 2 scavenging ability with the IC 50 value of 0.11 mg/ml and 0.47 mg/ml, respectively, which was nearly 2-and 12-fold of that of GCHs. The  showed significantly stronger antioxidant activity than that with higher MW. Ren et al. (2008) also found that peptides with MW less than 3 kDa contribute more to the antioxidant activity than polypeptides. But the scavenging efficacy of F3 was also higher than those of F4 and F5, which may have resulted from the occurrence of small peptides or free amino acids with low or without antioxidant abilities. The radical scavenging ability of the fractions with MW <3 kD from duck breast hydrolysates showed a negative correlation with its molecular weight . F4 exhibited the best Fe 2+ chelating ability, but no ABTS. + and O ⋅− 2 scavenging ability was detected when the concentration was set at 10 mg/ml. Therefore, F3 was selected for further peptide identification and antioxidant peptide screening.

| Identification and screening of antioxidant peptides in F3
The amino acid sequence and MW of peptides in F3 were analyzed  In this research, the potential antioxidant peptides were selected and synthesized based on the following well-known structureactivity relationships: (1) Peptides contain hydrophobic amino acid residues, such as Ala, Pro, Leu, Ile, and phenylalanine (Phe), which can increase the accessibility of peptides in water-lipid interface and promote the quenching on free radicals (Cai et al., 2015;Zou et al., 2015). (2) The presence of aromatic amino acids of tyrosine (Tyr) and tryptophan (Trp), which can act as good hydrogen donors and exhibit strong radical scavenging activity . (3) The presence of basic or acidic amino acids of Arg, Lys, histidine (His), Asp, and Glu, which are able to chelate metal ions through the carbonyl and amino groups in the side chain (Saiga et al., 2003).
In addition, the imidazole ring in the R group of His has the ability of donating hydrogen, trapping lipid peroxyl radical, and chelating metal ion . (4) The presence of cysteine (Cys), the sulfhydryl (SH) group in the R group can act as radical scavenging . Finally, a total of 15 potential antioxidant peptides were screened for further synthesis and bioactivity evaluation, and the physical and chemical properties of these synthesized peptides are shown in Table 4.

| ABTS· + scavenging ability of synthetic peptides
To compare the antioxidant ability of synthesized peptides intuitively, the ABTS. + scavenging ability of peptides was expressed as μmol GSH equivalent per gram of peptide (μmol GSH/mg), while a higher value suggests stronger antioxidant ability. As shown in However, the ABTS. + scavenging ability of the 12 peptides with Trp and Tyr was different, indicating the importance of amino acid sequence. It was apparent that the peptides containing Trp or Tyr residue at the C-terminus had higher scavenging activity. For example, P7 (VAGW) showed the highest scavenging ability with the value of 139.77 μmol GSH/g, which was followed by P11 (APPAMW) (80.83 μmol GSH/g).. This was consistent with the results found by Saito et al. (2003). Meanwhile, there is no significant difference between P10 and P9 (p > .05), which may be due to the presence of two Tyr residues in P10 (FYYGK), enhancing its ABTS. + scavenging ability to a certain degree. The equivalent value of P12 (LGGY) was significantly lower than that of P9 (LFGY), indicating that Phe attached to the N-terminal Leu contributed more to the ABTS. + scavenging ability of L-X-GY than Gly. Among the five similar pentapeptides P1-P5,

| Synergetic effect of synthetic peptides
Usually, the isolation and purification of protein hydrolysates with antioxidant activity may result in three different results: (1) A minimum of one separated fraction or purified peptide has stronger antioxidant activity than the crude hydrolysates.
(2) A minimum of one fraction showed better bioactivity than the hydrolysates, but the purified peptides exhibited weaker activity.
(3) The separated fractions gave lower antioxidant ability than the hydrolysate (Zou et al., 2015). For example, the ABTS. + and ·OH scavenging ability, and suppressing lipid oxidation of peach protein hydrolysates (MW >5 kDa, 3-5 kDa, and <3 kDa) were all reduced after ultrafiltration (Vásquezvillanueva et al., 2016). But the antioxidant ability of fraction and peptides from duck breast protein hydrolysates was significantly enhanced after fractionation and purification .
In this research, the radical scavenging ability of F3 was much higher than those of GCHs, but the purified peptides presented much weaker ability, suggesting the presence of synergistic effect among peptides. The combination of two, three, four, and five peptides among P 7 (VAGW), P 11 (APPAMW), P 9 (LFGY), P 10 (FYYGK), and P 13 (LLLYK) (the top five activity) was designed to investigate the synergistic effect. The ABTS. + scavenging ability of the combinations with two and three peptides is shown in Figure 5b-c. Excepting for the combination of P 9 + P 10 + P 13 , no antagonism was observed. All the combinations with P 7 and/or P 11 exhibited significant synergistic effect (p < .05), indicating that P 7 and P 11 synergized greatly with other tested peptides. This could result from the role of C-terminal Trp (W) (Zou et al., 2015). Among the two peptides' combination, the P 7 + P 11 presented the highest ABTS. + scavenging ability, with the GSH equivalent of 115.36 μmol GSH/g. The strongest synergism was found in P 7 + P 13 , the EV was 38.93 μmol GSH/g higher than the CV. For three peptides' combination, P 7 + P 11 + P 13 exhibited the strongest ABTS. + scavenging ability and synergism, the EV was 132.42 μmol GSH/g, which was 59.66 μmol GSH/g higher than the CV. In addition, the combination of P 7 or P 7 + P 11 with P 13 always showed significantly higher synergism than the combination with P 9 and P 10 (p < .05). The above results indicated strong synergistic effect among P 7 , P 11 , and P 13 again, which could be due to the fact that P 13 possesses 3 Leu and 1 Lys, while P 9 and P 10 contain only Unexpectedly, except for P 7 + P 10 + P 9 + P 13 , no synergistic effect was observed on the 4 or 5 peptides' combination among P 7 , P 9 , P 10 , P 11 , and P 13 (Figure 5d). Controversially, except for P 7 + P 10 + P 9 + P 13 , all tested four or five peptides' combination exhibited different degrees of antagonism. The EV was significantly lower than the corresponding CV (p > .05). Among which, P 10 + P 9 antagonized P 11 + P 13 most, the EV was 10 μmol GSH/g lower than the CV. In addition, the ABTS. + scavenging ability of P 7 + P 11 + P 13 was greatly decreased when P 10 or P 9 was included, the GSH equivalent value was reduced by 51%~52%, suggesting that the presence of P9 or P10 could reduce the radical scavenging ability of P 7 + P 11 + P 13 greatly. This could be due to the fact that the presence of P9/P10 suppressed the proton-donating ability of P 7 + P 11 + P 13 , leading to reduced radical scavenging ability, but the detailed mechanism needs further research.

| CON CLUS IONS
In this study, the two-step enzymatic hydrolysis of grass carp muscle for preparing antioxidant peptides was optimized. The optimal conditions were: first-step hydrolysis with protamex at a protein-liquid ratio of 4%, enzyme/substrate ratio of 10,000 U/g, pH of 7.0, and temperature of 50°C for 3 h, followed by the second-step hydrolysis with alcalase at  The pI and net charge of peptides with amino acids equal to or greater than 5 were calculated with https://web.expasy.org/protp aram/, while those of peptides with amino acids less than 5 were calculated with https://pepca lc.com/ppc.php.

TA B L E 4
Physiochemical properties of the 15 synthesized peptides identified in grass carp hydrolysates (GCHs) the following conditions: pH, 9.0; enzyme/substrate ratio, 6,000 U/g; temperature, 50°C for 2 h. The obtained GCHs mainly consisted of hydrophobic and acidic amino acids with MW lower than 5.4 kDa. Twelve novel antioxidant peptides were identified from the most active fraction (F3), among which VAGW possessed the highest ABTS. + scavenging activity (139.77 μmol GSH/g). Significantly synergistic effects were observed between the combination of two and three peptides, VAGW, APPAMW, and LLLYK, which exhibited the strongest synergism with the experimental value of 59.66 μmol GSH/g higher than the calculated. The C-terminal Trp played an important role in the synergism. In conclusion, fraction 3 of GCHs and VAGW have a promising potential to serve as natural antioxidants used in functional materials for a supplement.

ACK N OWLED G M ENTS
This work was supported by the National Key R&D Program of China (2018YFD0901101) and Jiangxi Provincial Key R&D Program (20192ACB60005).

CO N FLI C T O F I NTE R E S T
The authors declare no competing financial interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are openly available