WP was provided by Changbai Fairy Biotechnology Co., Ltd. (Liaoning, China), which contained 75% protein (dry weight). Alcalase (350,507 U g⁻¹), neutrase (230,453 U g⁻¹) and pancreatin (109,000 U g⁻¹) were obtained from Novozymes Co., Ltd. (Nanjing, China). Pepsin (32,600 U g⁻¹), papain (21,413 U g⁻¹), flavourzyme (64,144 U g⁻¹), protamex (226,186 U g⁻¹), DPPH and D-galactose were purchased from Sigma-Aldrich (St. Louis, MO, USA).

The WPH was prepared by different proteases according to our previous study (10). Briefly, WP was mixed with distilled water in a ratio of 1:20 (w/v), and then the mixture was stirred and adjusted to the optimum temperature and pH of proteases. The WP was hydrolyzed under the catalysis of different protease (5,000 U/g, of protein), and the pH of the reaction system was maintained constant with 0.5 M NaOH. Enzyme was inactivated by keeping the reaction system in boiling water for 10 min after hydrolysis. The supernatant of the hydrolysate was collected after centrifuged (TGL18M, Kate Laboratory Instrument Co., Ltd, Yancheng, China) at 11,000 rpm for 10 min.

The DH was measured according to the pH-stat method (11). The DH was calculated following the equations below:

where B is the consumed volume of 0.5 M NaOH (mL); N is the concentration of NaOH (M); α is the average degree of dissociation of the α-NH₂ groups; m is the mass of hydrolyzed protein (g); and h_tot is the total number of peptide bonds per gram of protein, which is 8.0 mmol g⁻¹ for WP (12).

Simulated in vitro gastrointestinal digestion was conducted as described in our previous work (13). Briefly, the samples were mixed with pepsin (2%, w/w of protein) and incubated under the gastric condition (37°C, pH 1.5) for 2 h. Subsequently, after the simulated gastric digestion, pancreatin was added (2%, w/w of protein) to the mixture to continue the simulated intestinal digestion under the intestinal condition (37°C, pH 7.5) for 4 h. Finally, the supernatants of the WP and WPH digests were collected after the reaction system was inactivated.

Peptide contents were determined by the Folin phenol method as described by Yang et al. (11).

Data were expressed as the means ± standard deviation (SD). The differences were evaluated using the one-way analysis of variance (ANOVA). Statistical analysis was accomplished using SPSS 19.0 (SPSS Inc., USA) software.

After limited enzymatic hydrolysis of walnut protein with five different commercial proteases in vitro, the peptide content in the hydrolysate increased significantly with the increase of DH (as shown in Figure 2). Among them, the enzymatic hydrolysis efficiency of alcalase (Figure 2A) and papain (Figure 2D) was relatively high, and higher peptide content could be obtained in the hydrolysate under the same DH compared with other proteases. Papain hydrolysate with 16% DH produced the highest content (13.55 mg/mL) of the peptides in the WPH. After the simulated gastrointestinal digestion in vitro, the content of peptides in the digests of WP and WPH increased significantly compared to WP and WPH, which reached 20.504 mg/mL in the WP digest and ranged from 17.744 to 21.804 mg/mL in WPH digest. The results illustrate that the peptides in WP can be released by the enzymatic hydrolysis of commercial protease and the digestion of pepsin and pancreatin, and the digestive effect of pepsin and pancreatin on WP was even better than that of commercial protease. This is because hydrophobic amino acids and acidic amino acids are the main components of WP, while chymotrypsin (in pancreatin) decomposes mostly hydrophobic amino acids and pepsin catalyzes primarily the decomposition of acidic amino acids (9). This is also the main reason for the high digestibility of WP. Therefore, the gastrointestinal digestive system should be considered in the preparation and activity evaluation of walnut bioactive peptides.

The DPPH radical-scavenging activities of WP (DH = 0%) and WPH with different DHs after in vitro limited enzymatic hydrolysis and their simulated gastrointestinal digests are shown in Figure 3. It can be observed in Figure 3A that, alcalase hydrolysis could enhanced the DPPH radical-scavenging activity of WP. The DPPH scavenging-activity of WPH increased rapidly as the DH changed, reaching the maximum value of 69.39% at DH of 8%, and then decreased. After the simulated digestion in vitro, the scavenging-activity of WPH gastrointestinal digestive fluid (WPH-GI) with the DH of 4% reached the highest value of 67.99%, however, which showed no significant difference (p > 0.05) when compared to WPH-GI with the DH of 0% (that is WP gastrointestinal digestive fluid, WP-GI). Similar to the results of alcalase hydrolysis, the DPPH scavenging-activity of the neutrase hydrolysate (Figure 3B) increased rapidly as the DH increased, reaching the maximum (71.42%) at the DH of 8%, and then decreased slightly with the continuous increase of DH. When the WPH was digested by pepsin and pancreatin in vitro, the DPPH scavenging activity of WPH-GI at DH of 4, 12, and 16% showed no striking difference in comparison with the WP-GI (p > 0.05). The enzymatic hydrolysis efficiency of neutrase was high, and it could hydrolyze WP into peptides with high DPPH-scavenging activity under low DH, and the scavenging-activity did not appear to be change significantly with the increase of DH. Meanwhile, the simulated gastrointestinal digestion did not significantly change the DPPH scavenging-activity of WPH, indicating that the peptides with DPPH scavenging activity produced by neutrase hydrolysis had good digestion stability. In the process of protamex hydrolysis (Figure 3C), the DPPH scavenging-activities of WPH showed an upgrade trend as the DH increased (p < 0.05). However, the DPPH radical-scavenging activity of protamex hydrolysate was relatively lower than that of alcalase and neutrase hydrolysate at the same DH. After the in vitro simulated gastrointestinal digestion, the DPPH radical-scavenging activity of WPH with different DH was improved to varying degrees, and no significant difference (p > 0.05) was found between WPH-GI with the DH of 0, 4, 12, and 16%, and slightly higher than WPH-GI with the DH of 8% (p < 0.05). Although the DPPH radical-scavenging activity of the protamex hydrolysates was relatively lower than that of other protease hydrolysates under the same DH, both WP and WPH could further improve DPPH radical-scavenging activity to achieve the same effect after the simulated gastrointestinal digestion. Papain had a higher enzymolysis efficiency compared with other proteases used in this work in terms of the DPPH radical-scavenging activity, which was higher than that of other protease hydrolysates under the same DH. The scavenging activity of WPH produced by papain (Figure 3D) reached the maximum value of 78.75% at the DH of 4%, and then decreased slightly as the DH continuously increasing. During the simulated digestion in vitro, the peptide bonds and disulphide linkages in WP were broken by pepsin and pancreatin, and resulting a rapid increase in the DPPH scavenging activity. However, the high scavenging activity of papain hydrolysates decreased after in vitro simulated gastrointestinal digestion. The highest value of the DPPH scavenging activity (67.11%) was observed in WPH with the DH of 0%, additionally, no significant difference was found between WPH-GI with the DH of 0, 4, 12, and 16% (p > 0.05). This is because the WP was enzymatically hydrolyzed into peptides with high DPPH radical-scavenging activity under the catalysis of papain. After the further digestion by pepsin and pancreatin, the original antioxidant characteristic structure of peptides was destroyed and the scavenging activity of DPPH was decreased. The DPPH radical-scavenging activity of WPH was partially improved by flavourzyme hydrolysis (Figure 3E) under the low DH (DH ≤ 12%), and the scavenging activity reached a maximum value of 59.27% when the DH reached 16%. After the in vitro simulated gastrointestinal digestion, the DPPH radical-scavenging activity of WPH was further increased (p < 0.05), especially the scavenging activity of the WPH-GI with the DH of 16%, which reached 73.01%.

The reducing power characterizes the ability of a compound to reduce ferric ion (Fe³⁺) to ferrous ions (Fe²⁺) through electrons or hydrogen donation (15). The reducing powers of WPHs produced by different proteases and their in vitro simulated gastrointestinal digests are shown in Figure 4. As depicted in Figure 4A, the reducing power of the WPH hydrolyzed with alcalase increased, then dropped a little as DH increase, with the highest value 10.68 μmoL TE/mL at DH of 8%. However, after the in vitro simulated gastrointestinal digestion, the reducing power of WPH-GI with the DH of 8% is reduced to 8.48 μmoL TE/mL, which is not significantly different (p > 0.05) from that of WP-GI (8.55 μmoL TE/mL). The maximum value (9.02 μmoL TE/mL) of the reducing power occurred in the WPH-GI with the DH of 4%. According to the traditional enzymatic hydrolysis method without considering the function of gastrointestinal digestion, the WPH with the DH of 8% may be selected for further production. However, the WPH with the DH of 4% was the best choice after the evaluation of the two-stages enzymatic hydrolysis method. Similar with the result of the DPPH radical-scavenging activity, the reducing power of neutrase hydrolysate (Figure 4B) increased rapidly with the increase of the DH and reached the highest value of 10.48 μmoL TE/mL at the DH of 8%. After the WPH being digested by pepsin and pancreatin, the reducing power of WPH-GI decreased partly compared with WPH. Finally, the WPH-GI at DH of 4% showed the highest reducing power (8.65 μmoL TE/mL), which had no significant difference (p > 0.05) compared with WP-GI (8.55 μmoL TE/mL). The digestion of pepsin and pancreatin could not only hydrolyze WP into WP-GI with high reducing power, but also destroy the active peptides that have been already formed in the WPH, resulting in a decrease in reducing power. When the WP was hydrolyzed by the protamex (Figure 4C), the reducing power of the WPH showed a continuous upward trend as the DH increase, and further increased after the gastrointestinal digestion. Finally, the reducing power of WPH-GI at DH of 4% (8.80 moL TE/mL) was highest, and the reducing power of the WPH-GI with the rest DH showed no significant difference (p > 0.05) with that of WP-GI. The reducing power of papain hydrolysates with different DH, and their simulated gastrointestinal digests are depicted in Figure 4D. Due to the high enzymatic hydrolysis efficiency, the hydrolysis of papain could significantly increase the reducing power of WP, and as the DH increases, the reducing power increased slowly. After the in vitro simulated gastrointestinal digestion, the reducing power of WPH-GI showed no significant change compared to WPH, indicating that the papain hydrolysate has good digestion stability. Moreover, no significant difference (p > 0.05) was observed on the reducing power of the WPH-GI with the DH of 4 and 8% compared with that of WP-GI. The reducing power of WPH-GI produced by flavourzyme (Figure 4E) was limited due to the low enzymatic hydrolysis efficiency. After the simulated digestion in vitro, the reducing power of WPH-GI was greatly improved, but it was still lower than that of WP-GI (p < 0.05).

In summary, the results of enzymatic hydrolysis of the five different proteases showed that all of the proteases could hydrolyze WP into WPH with DPPH radical-scavenging activity and reducing power, but the enzymatic hydrolysis efficiency of different proteases and the antioxidant activities of WPH at the same DH are different. In addition, WP could be hydrolyzed by pepsin and pancreatin in the simulated digestion system to exert antioxidant activities. Although the process of enzymatic hydrolysis could obtain bioactive peptides with high antioxidant capacity properties, their activities not increase significantly or even decrease after in vitro simulated digestion. This is due to the unique cleavage sites of gastrointestinal digestive enzymes. As a kind of endopeptidase, pepsin mainly catalyzes the decomposition of acidic amino acid carboxyl terminal peptide bonds to form peptides, changes the natural conformation of the intact proteins, and facilitates the further digestion of pancreatin (16). Pancreatin is a mixture of multiple enzymes, including trypsin, chymotrypsin, elastase and carboxypeptidase, which function jointly to cleave peptides into oligopeptides and amino acids with certain biological activity (17). Trypsin mainly catalyzes the decomposition of the C-terminal peptide bonds of basic amino acids residues, especially the peptide bonds adjacent to Arg and Lys. Chymotrypsin decomposes peptide bonds adjacent to hydrophobic amino acids and aromatic amino acid residues (14). Elastase catalyzes the decomposition of peptide bonds at the C-terminus of short-chain fatty chains, including Ala, Gly, Ser etc. Additionally, carboxypeptidase is a kind of exopeptidase that systematically removes amino acids from the C-terminus of peptides produced by digestion with other endopeptidases (18), amongst, carboxypeptidase A preferentially releases Val, Leu, Ile, Ala, and carboxypeptidase B releases the basic amino acids Arg and Lys (17). Our previous study has found that WP has a high content of Arg (14.27%), Leu (6.75%), and Vel (4.55%), as well as acidic amino acids (33.66%) and hydrophobic amino acids (27.18%), which results in a good digestibility of WP (10). Antioxidant capacity properties of peptides are partly attributed to the amino acid composition (19). He et al. (20) reported that hydrophobic amino acids (such as Leu, Ile, Pro, Tyr and Met) endow protein hydrolysates with antioxidant capacity due to their abundant electrons can be donated to quench free radicals. Aromatic amino acids (Tyr, Phe) have been considered to enhance the potency of radical scavenging ability by donating protons to stabilize electron-deficient radicals, and maintain their stability through resonance structures (21). Therefore, after the evaluation of antioxidant activity in two-stages of in vitro limited enzymolysis and in vitro simulated gastrointestinal digestion, it was found that the DH of 0% for limited enzymolysis might be the best choice for exerting its antioxidant activity. That means, WP can exert its potential antioxidant activities after being digested by pepsin and pancreatin in human body without additional enzymatic hydrolysis by commercial protease. To further confirm this result, WP (WPH with the DH of 0%) and its alcalase hydrolysate with DH of 8% was selected for in vivo verification experiments.

The weight growth rate and organ index of SD rats reflect the degree of oxidative damage caused by D-galactose to the rat body, and also reflect the protective effect of WP and WPH on oxidative damage. As depicted in Table 2, the weight growth rate of the D-Gal group was significantly (p < 0.01) lower than that of the control group, which is consistent with the findings of Li et al. (22). The weight growth rate of rats in the WP-I, WP-II and WPH groups was higher than that of the D-Gal group, but still significantly lower than that of the control group (p < 0.05), especially the WP-II group (p < 0.01). The WPH group showed the most significant improvement in weight loss caused by oxidative damage. The results of weight growth rate indicated that the injection of D-galactose had a certain inhibitory effect on the weight gain of SD rats, while the intragastric administration of WP and WPH could alleviate the slow weight gain of rats caused by oxidative damage.

The organ index of SD rats is shown in Table 2. It can be seen that the organ index of the liver, heart and spleen tissue of rats in the D-Gal group was significantly lower than that of the control group (p < 0.05), especially the liver tissue (p < 0.01), whereas no significant change in any organ index was observed in the WP-I, WP-II and WPH groups compared with the control group (p > 0.05). This results indicated that the injection of D-galactose had certain damage to the liver, heart and spleen of SD rats, while administration of WP and WPH could effective relieve the oxidative stress and enhance antioxidant defense in the body. Li et al. (22) also found that the injection of D-galactose caused liver and kidney atrophy, and tilapia skin collagen polypeptides treatment could improve the decline of organ function in mice caused by D-galactose.

An oversupply of D-galactose can cause the tissue damage in rats, especially in the liver (22). The histopathology of liver tissue was observed and the results are showed in Figure 5. In the control group, the hepatocytes of rats were in normal shape with regular cell shapes and structurally sound hepatic lobule, radially arranged hepatic cords around a central vein, prominent nuclei and nucleoli, and uniform and dense arrangement. Conversely, compared with the control group, the rats in the D-Gal group had lesions around the central vein area of the liver, the liver cells were enlarged with unclear borders, the cords were arranged disorderly, and the hepatic lobule structure was destroyed. Nevertheless, compared with the D-Gal group, the structure and morphology of hepatocytes in the three treatment groups were ameliorated gradually, the hepatic cords were arranged relatively neatly. Both WP and WPH exhibited a preventive effect against D-galactose-induced pathologic damage to the liver.

To further explore the antioxidant effect of WP and WPH in the body, the MDA and GSH contents and activities of antioxidant enzymes SOD and T-AOC in serum of rats were measured, and the results are presented in Figure 6. As a product of polyunsaturated fatty acid peroxidation, MDA is one of an reactive carbonyl compound decomposed by the unstable lipid peroxides (23). MDA content often reflects the lipid peroxidation level in vivo, and indirectly reflects the degree of oxidative damage of cells (24). As depicted in Figure 6A, the content of MDA in the serum of rats in the D-Gal group was 2.07 times higher than that of the control group (p < 0.01), indicating that the D-galactose-induced oxidative stress model of SD rats was successfully established. The treatment with WP and WPH could significantly reduce the MDA level in serum, and there was no significant difference in MDA level among WP-I, WP-II and WPH groups (p > 0.05). These results indicate that administration of WP and WPH can effective reduce lipid peroxidation as seen in lowered MDA content, thereby reducing the oxidative damage caused by D-galactose. In addition, there was no significant difference (p > 0.05) in the inhibition of lipid peroxidation between WP and WPH. GSH is an important non-enzymatic antioxidant in the body (25). From Figure 6B, it can be seen that the GSH level was significantly (p < 0.01) decreased by D-Gal group (54.53%) compared with the control group, indicating that the oxidative stress model was successfully constructed. In comparison with the control group, the GSH levels of WP-I, WP-II and WPH groups also showed a significant (p < 0.01) reduction by 27.39, 29.30, and 29.30%, respectively, while the GSH levels of these three treatment groups were significantly higher (p < 0.01) than the D-Gal group. Moreover, no significant difference was observed on the GSH level amongst the WP-I, WP-II and WPH groups (p > 0.05). These results suggest that both WP and WPH can effectively alleviate the decrease of GSH level caused by D-galactose-induced oxidative damage, thereby enhancing antioxidant defense, reducing oxidative stress and protect rats from oxidative damages.

Similar to the MDA and GSH level, D-galactose treatment also successfully and significantly reduced (p < 0.01) the activities of the innate antioxidant enzymes SOD (Figure 6D) and non-enzymatic antioxidant marker T-AOC (Figure 6C) in the serum of rats in the D-Gal group when compared to the control group. However, the SOD activities of the three treatment groups were not significantly different from that of the D-Gal group (p > 0.05), and was significantly lower than that of the control group (p < 0.01). T-AOC is an index that comprehensively react the antioxidant enzyme system of the body to resist external oxidative damage and scavenge radicals, reflecting the overall antioxidant capacity (23). As shown in Figure 6C, in comparison with the D-Gal group, the low-dose and high-dose WP treatment showed a significant improvement (p < 0.01) on T-AOC level in the serum of rats, as well as the WPH treatment (p < 0.05). Taken together, the results of in vivo antioxidant tests demonstrated that D-galactose could disrupt the antioxidant defense system of rats, cause oxidative damage and promote lipid peroxidation, while WP and WPH treatment could prevent these damages, and the antioxidant effect of WP was more significant than that of WPH. The protective effect of WP and WPH against oxidative damage might be rationalized due to the impairment of antioxidant defenses systems damage.