Effects of enzymatic treatments on the hydrolysis and antigenicity reduction of natural cow milk

Abstract Cow milk (CM) allergy is one of the most common food allergies worldwide; the most abundant CM proteins, such as casein (CN), β‐lactoglobulin (β‐LG), and ɑ‐lactalbumin (ɑ‐LA), are all potentially allergenic. Reducing the antigenicity of CM continues to be a major challenge. However, previous studies have focused on the antigenicity of individual allergic CM proteins. Thus, in the present study, we aimed to evaluate the effects of different food‐grade enzymes on the antigenicity of CN, β‐LG, ɑ‐LA in natural CM. The degree of hydrolysis (DH) and molecular mass (MW) distribution of CM hydrolysates were assessed. Additionally, the residual antigenicity of CM hydrolysates was evaluated through enzyme‐linked immunosorbent assay and Western blotting with anti‐CN, anti‐β‐LG, and anti‐ɑ‐LA rabbit polyclonal antibodies. The results showed that Alcalase‐ and Protamex‐mediated hydrolysis could efficiently reduce the antigenicity of CN, β‐LG, and ɑ‐LA, inducing a higher DH, the loss of density of CM proteins, and the increasing levels of low MW (<3 kDa) peptides in CM hydrolysates. Further, Protamex and Alcalase could more efficiently hydrolyze the major allergenic components of CM than the other enzymes, which could represent an advantage for the development of hypoallergenic CM. These findings add further knowledge about the study and development of hypoallergenic CM.

In some parts of the world, CM is one of the most important dietary components for humans and is considered a rich source of proteins, lipids, lactose, and minerals (Liang et al., 2018). Because of its high nutritional value, diverse mineral composition, proper calcium to phosphorus ratio, and numerous sources, it is regarded as the best choice for infant formula and dairy products. However, allergy caused by CM is still a major concern in the dairy industry .
Previous research has revealed approximately 30 potentially allergenic proteins in CM that can induce immune responses in infants and young children. Moreover, large scale studies on allergic patients have suggested that the most abundant proteins in CM, such as caseins (CN, 78%), β-lactoglobulin (β-LG, 9.7%), and ɑ-lactalbumin (ɑ-LA, 3.6%), are all major allergens (Aliaga et al., 2020). Even proteins present in low amounts, such as bovine serum (BSA), lactoferrin (LF), and immunoglobulins (Igs), have shown the capacity to induce milk-related allergies (Kawamoto et al., 2020). Several studies have estimated the incidence of CM protein allergy as approximately 2%-6% (Piippo et al., 2020). Therefore, reducing the antigenicity of CM is a major challenge by milk manufacturers.
Researchers have developed a variety of methods to reduce sensitization to CN, β-LG, and ɑ-LA, or other milk components, by optimizing and improving the CM processing conditions (Bu et al., 2013), which including heat treatment, glycation, and enzymatic hydrolysis, and others. Among these, heat treatment is the most commonly used method to prevent pathogens contamination, but it remains a controversial method for reducing the risk of allergies (Fiocchi et al., 2004). Glycation is one of the most frequent chemical modifications during industrial production and processing of milk, but due to the complexity of the product, its safety still warrants evaluation (Taheri-Kafrani et al., 2009).
However, enzymatic hydrolysis, which uses digestive enzymes to alter the immunoreactivity of allergenic proteins, is the most effective method to modify proteins (Wróblewska et al., 2004).
During enzymatic hydrolysis, some peptide or disulfide bonds are damaged, leading to the collapse of conformational or linear epitopes; thus, it can eliminate or reduce the allergenicity and antigenicity of milk proteins. Furthermore, enzymatic hydrolysis can yield a variety of new peptides, preserve the nutritional value of milk proteins, and also offer many physiological benefits for infants and young children (Fiocchi et al., 2018).
Enzymatic hydrolysis is a well-known, safe, and effective processing technology to reduce the allergenicity of milk proteins.
Recently, it has been widely used to produce high quality and hypoallergenic protein hydrolysates. Several studies have shown that whey (which contains β-LG and ɑ-LA) antigenicity could be reduced by hydrolysis with Alcalase (Docena et al., 2002). Additionally, combined microwave and enzymatic hydrolysis of a whey protein concentrate (WPC) hydrolysate by Pronase, Chymotrypsin, and other five different food-grade enzymes demonstrated effective proteolysis of whey proteins by either of the enzymes in combination with these treatments (Izquierdo et al., 2006). Some researchers have evaluated the antigenicity of whey proteins hydrolysates obtained through the combination of enzymatic hydrolysis with high-pressure treatment, suggesting that this method could decrease the immunoreactivity of whey protein hydrolysates (Penas et al., 2006). However, previous studies have focused on the enzymatic hydrolysis of individual proteins or the antigenicity of individual allergic CM proteins. Thus, in the present study, the effects of different food-grade enzymes on the major allergenic proteins present in natural CM were assessed.
Briefly, degree of hydrolysis (DH), molecular weight (MW) distribution, and residual antigenicity of CM hydrolysates were evaluated.
This study aimed to lay a solid theoretical foundation for the production of CM-based hypoallergenic dairy products.

| Sample collection
Fresh CM was collected at the local farm (Shenyang, China) from 60 healthy cows (1-6 years of age) fed on grass. The CM was mainly composed of protein (3.05 g/100 ml), fat (3.54 g/100 ml), ash (1.12 g/100 ml), moisture (87.80 g/100 ml), lactose (4.98 g/100 ml), and dry matter (12.35 g/100 ml). Fat was removed from the milk by high speed centrifugation at 15,000 g for 30 min at 4°C. Skim milk was placed into bottles, transported to the laboratory in the Shenyang Agricultural University, and stored at 4°C.

| Cow milk hydrolysis experiments
The enzyme solutions were prepared by dissolving each enzyme in distilled water (100 mg/ml) at room temperature. The CM samples and enzyme solutions were preheated separately with stirring (20 min) at suitable temperature (Alcalase: 55 ± 5°C; Neutrase: 50 ± 5°C; Flavourzyme: 50 ± 5°C; Protamex: 50 ± 5°C; Papain: 20 ± 5°C; and Pepsin: 30 ± 5°C). Then, each enzyme was added to the CM in an enzyme activity-to-substrate ratio ranging between 2,000 and 10,000 U/g. The mixture was incubated for 120 min at their suitable temperature, and the enzymatic hydrolysis reaction was stopped by heating the mixture at 90°C for 10 min, followed by immediate cooling in ice. Afterward, the supernatant of hydrolysates was centrifuged at 5,000 g for 10 min at 4°C and stored at − 80°C for further study. Untreated samples were used as control.

| Determination of the DH
The DH of the hydrolysates was evaluated using the OPA method, as previously described (Church et al., 1983), with some modifications. Briefly, the method quantified the amount of hydrolyzed peptide bonds using OPA. The OPA solution was prepared by dissolving sodium dodecyl sulfate (SDS), sodium tetraborate decahydrate, 97% OPA, and 99% DTT. A solution of serine (100 μg/ml) was used as standard control. The OPA reagent (400 μl) was added to the hydrolysates (3 ml), swirled by inversion, incubated for 2 min in the dark, and the absorbance was measured at 340 nm. Each hydrolysate was analyzed in triplicate.

| Determination of the molecular weight (MW) distribution
The MW distribution of the CM hydrolysates was assessed by gel permeation chromatography (GPC) using an Agilent PL aquagel-OH 10 × 300 mm column (LC1260; Agilent) with UV detection at 214 nm under the following conditions: 100 μl injection volume, 30 min analysis time, 30°C column temperature, 0.1 mol/L sodium nitrate, and 500 mg/L sodium azide aqueous solution mobile phase.

| SDS-polyacrylamide gel electrophoresis (PAGE) and Western blotting
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was used to evaluate the hydrolysis of CN and whey (β-LG, ɑ-LA) in CM hydrolysates, depending on the method described by Laemmli (1970). The separating and stacking gels were prepared by using 15% and 3% of acrylamide concentration, respectively. Prior to electrophoresis, the CM hydrolysates plus loading buffer were heated in boiling water for 5 min. Approximately 15 μg of the protein samples was transferred into each well and the total proteins in the gel were stained with Coomassie Brilliant Blue G-250.
Next, the gels with the separated proteins were submitted to electroblotting using a miniVE blotter (Bio-Rad) at 80 V for 120 min, and the proteins were transferred onto a 0.45 μm nitrocellulose membrane (Bio-Rad). The membrane was incubated with 10 ml blocking buffer (1% BSA (w/v) in phosphate-buffered saline solution (PBS, pH 8.0) with 0.1% Tween20). Afterward, it was incubated with polyclonal primary antibodies (1:1,000) for 1 hr at 37°C, washed and incubated again with secondary alkaline phosphatase-conjugated goat anti-rabbit IgG antibodies (1:5,000 in blocking buffer) for 1 hr at room temperature. Image analysis of the membranes and the proteins was performed with a gel scanner (Amersham Pharmacia Biotech) and the Gel-pro Analyzer software (Media Cybernetics).

| Determination of IgG-binding ability
The IgG-binding of CM hydrolysate was quantitatively analyzed by indirect competitive enzyme-linked immunosorbent assay (ELISA), according to the method by Huang et al. (2019), with some modifications. Microtiter plates (96-well) were coated with 30 µg/ml CN (or 5 µg/ml ɑ-LA, or 0.5 µg/ml β-LG) diluted in PBS and incubated overnight at 4°C. The next day, the plates were washed thrice with 300 µl PBS with 0.05% Tween20 (PBST). Then, 0.5% gelatin in PBS was used to block residual-free binding sites and the plates were where B and B 0 represent the absorbance measured in the presence and absence of CM hydrolysate, respectively.

| Statistical analysis
The results are expressed as the mean ± standard deviation of three independent assays. Analysis of variance (ANOVA) was used to assess the effects of the treatments and differences between samples.
The analyses were performed using SPSS Statistics for Windows software version 17.0 (SPSS, Inc.). Differences were considered significant at p < .05.

| The degree of hydrolysis
The DH of CM hydrolysates obtained with different enzymes was quantified using the OPA method, as shown in Figure 1.

| SDS-PAGE
The Moreover, the density of the protein bands was also altered. As seen in Figure 2, Alcalase, Protamex, and Flavourzyme showed strong hydrolysis potential for those major allergenic proteins, among which Alcalase and Flavourzyme extensively hydrolyzed CN and whey (β-LG, ɑ-LA) even at the lowest concentration assessed (enzyme activity-to-substrate ratio 2,000 U/g). However, increased enzyme activity-to-substrate ratio had no significant impact on the protein bands. In addition, Protamex was associated with a concentration-dependent hydrolysis, in which the density of the protein bands gradually reduced with increasing enzyme activity-to-substrate ratio. Notably, Pepsin, Papain, and F I G U R E 1 The degree of hydrolysis (DH) of cow milk with different enzyme activity-to-substrate ratio (U/g). Each value represents the mean of three independent experiments ± SD. Different letters indicate significant differences among groups (p < .05)

F I G U R E 2 SDS-PAGE analysis. (a-f)
Neutrase were unable to completely hydrolyze CN and whey (β-LG, ɑ-LA). Therefore, among all the different enzymes used herein, Alcalase, Protamex, Flavourzyme were superior to the other enzymes in their ability to degrade antigenic proteins of CM.
Enzymatic hydrolysis can breakdown the peptide bonds of milk proteins and convert the whole proteins into smaller peptide fragments. Due to differences in enzymes species, their degradation ability for targeting proteins may be different. Smyth and Fitzgerald (1998) hydrolyzed a WPC preparation using Alcalase 0.6 L. After 30 min of enzymatic hydrolysis, they observed that the presence of a protein fraction of MW below 30 kDa. In addition, Ena et al. (2010) studied the protein conformation changes in WPC preparation during hydrolysis with Pepsin and Corolase PP. In this study, after 30 min of hydrolysis, small fractions containing proteins with MW of over 16 kDa were observed, suggesting different proteolytic enzyme specificities.

| IgG reactivity reduction analysis
To identify the IgG reactivity reduction of samples obtained by the different enzymes, rabbit polyclonal antibody and indirect competitive ELISA were performed. As shown in Figure 3, the IgG reactivity reduction of CM hydrolysates ranged from 4.02% up to 81.27%.

The IgG reactivity reduction of CM hydrolysates obtained with
Pepsin was the lowest (4.02%-12.02%). In turn, Protamex-derived IgG reactivity reduction increased gradually with increasing enzyme activity-to-substrate ratio, reaching a maximum of 72.25% when the enzyme activity-to-substrate ratio was at 8,000 U/g. In addition, it could be observed that CM subjected to Flavourzyme, Papain, or Alcalase hydrolysis exhibited higher IgG reactivity reduction (69.00%-81.27%), which was in agreements with the DH results (Section 3.1, Figure 1). The IgG reactivity reduction indicated that the CM antigenic epitopes were altered upon the enzymatic hydrolysis, although with slightly variable outcomes. Additionally, it also suggested that enzymatic hydrolysis can change the structure of the allergens in CM, interfere with the antigen-antibody complex, and thus reduce the IgG reactivity. Some studies have demonstrated that the residual antigenicity and IgE-binding ability of ɑ-LA, β-LG, α-CN, β-CN in CM with reduction rate of approximately 15%-90% (Shi et al., 2014).

| Molecular weight distribution
Enzymatic hydrolysis could break down milk proteins into short peptides and change the MW distribution, which was evaluated by GPC. The MW distribution was calculated assuming an exponential relationship between MW and elution time. After enzymatic hydrolysis with different enzymes, the MW distribution of the CM hydrolysates was significantly different, which reflected the differences observed in the peptide chain lengths. The MW distribution of the CM and CM hydrolysates is presented in Figure 4. It was clear that the MW of CM hydrolysates was mostly composed of small peptides (<3 kDa). After enzymatic hydrolysis, the percentage of milk proteins with MW distribution of 10-30 kDa, 5-10 kDa, 3-5 kDa significantly reduced (p < .05), suggesting that the cleavage reaction triggered by the different enzymes led to the fragmentation of the CM proteins into low MW peptides. Furthermore, the MW distribution of CM hydrolysates showed considerable differences. For the CM hydrolysate obtained with Flavourzyme, the relative percentage of milk proteins with a MW distribution of 3-5 kDa was approximately 2.05%, whereas for CM hydrolysate treated with Alcalase was of 0.76%. These results in combination with DH and SDS-PAGE data suggest that enzymatic hydrolysis with Alcalase and Protamex were much more effective in producing smaller peptides from CM.
Several studies proved that the most effective strategy to reduce the allergenicity of CM is by decreasing the MW of the principal CM allergens, namely CN, β-LG, and ɑ-LA, through enzymatic proteolysis (Asselin et al., 2010). Furthermore, some investigations have shown that peptides with MW between 1.6 and 3.5 KDa, prepared from either CN or whey, were unable to elicit an IgE-mediated allergic response (Otani et al., 1989). Deeslie and Cheryan (2010) reported that peptide MW was also one of the key factors regulating the functional properties of CM hydrolysates.

F I G U R E 3
IgG reactivity of cow milk hydrolyzed with different enzymes and enzyme activity-to-substrate ratio (U/g). Each value represents the mean of three independent experiments ± SD. Different letters indicate significant differences among groups (p < .05) F I G U R E 4 Molecular weight distribution of cow milk hydrolysates determined by gel permeation chromatography. Each value represents the mean of three independent experiments ± SD. Different letters indicate significant differences among groups (p < .05)

| Western blotting
To study the changes in allergenicity of the main CM allergenic proteins after enzymatic hydrolysis, Western blotting analysis using anti-CN, anti-β-LG, anti-ɑ-LA polyclonal antibodies was performed.
The CM proteins before and after the enzymatic hydrolysis had an immune binding reaction with the IgG antibody of the anti-CN, antiβ-LG, and anti-ɑ-LA rabbit serum ( Figure 5), indicating that the CM proteins before and after the enzymatic hydrolysis were immunoreactive. Compared with CM, the IgG-binding capacity of the antigen proteins (CN, β-LG, ɑ-LA) in CM hydrolysate was significantly reduced; however, the IgG-binding capacity of CN in CM hydrolysate was stronger than that of β-LG and ɑ-LA. Similar results demonstrated that Alcalase was more effective at reducing antigenicity of milk proteins, as it significantly reduced in the IgG-binding capacity of α-LA and β-LG (Yu et al., 2019). Carvalho et al. (2017) reported that Alcalase hydrolysis had similar inhibitory effect on the IgE-or IgG-binding capacity of α-LA and β-LG in WPC.

| IgG-binding ability
To determine the antigenicity of the CM hydrolysates, polyclonal antibodies were prepared, and competitive inhibition ELISA was used to evaluate the residual antigenicity of CN, β-LG, and ɑ-LA.
Overall, enzymatic hydrolysis could reduce the antigenicity of CM proteins, in which the antigenicity reduction rate of β-LG and CN was much higher than that of ɑ-LA (p < .05). Furthermore, Alcalase-and Protamex-mediated enzymatic hydrolysis of CM led to an ɑ-LA IgG-binding reduction rate of 11.28% and 6.10%, respectively, whereas a significantly higher residual antigenicity was obtained with Flavourzyme (−6.09%) (p < .05, Figure 6). In addition, the IgG-binding ability of β-LG during enzymatic hydrolysis with Flavourzyme and Protamex was significantly reduced (59.09% and 79.37%, respectively), but the enzymatic hydrolysis with Alcalase led to a dramatic reduction of 90.25% (p < .05, Figure 6). A marked decrease in IgG-binding potential of CN during the enzymatic hydrolysis process could also be observed, with a significant reduction rate of 69.14%-91.21%. It is worth noting that Alcalase and Protamex played an important role in reducing the antigenicity of CN, β-LG, and ɑ-LA; therefore, it could be concluded that they were superior to the other enzymes tested. These findings are in agreement with previously reported data. Wróblewska and Troszyñska (2005) showed that the lowest immunoreactivity to anti-ɑ-LA antibodies was found for whey protein hydrolysate obtained using Alcalase and its double dose. Quintieri et al. (2017) also reported that the antigenicity of whey was slightly reduced when it was incubated with fungal proteinases and pancreatic extracts.
In the present study, the amount of low MW peptides was significantly increased and the IgG-binding ability was considerably reduced in CM hydrolysates. This combined effect can be explained by the underlying enzymatic reaction, which partially breakdowns large proteins into low MW peptides, concomitantly, it may also split the sequence of epitopes, thereby resulting in reduced antigenicity. The allergic epitopes could be damaged or destroyed by the degradation of milk proteins or occurring conformational changes, which resulted in reduced reactivity and IgG-binding ability. In addition, the enzyme specificity or types can also affect the degradation of allergic epi-

| CON CLUS IONS
In this study, natural CM was first subjected to enzymatic hydrolysis by different food-grade enzymes, and the effects on the major al-

ACK N OWLED G M ENTS
This study was supported by the "Thirteenth Five Year" National Science and Technology Plan Project, (grant number 2018YFC1604302-03).

CO N FLI C T S O F I NTE R E S T
The authors confirm that the contents of this article pose no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data available on request from the authors-The data that support the findings of this study are available from the corresponding author upon reasonable request.

R E FE R E N C E S
F I G U R E 6 Residual antigenicity of cow milk hydrolysates determined by ELISA using polyclonal antibodies against ɑ-LA, β-LG, and CN. Each value represents the mean of three independent experiments ± SD. Different letters indicate significant differences among groups (p < .05)