Enzymatic kinetics of β-conglycinin using alkaline protease from Bacillus subtilis ACCC 01746 and analysis of antigenicity of hydrolyzed peptide

Haicheng Yin*,Xinrui Zhang,Zhixiang Yang,Jin Huang

College of Biological Engineering,Henan University of Technology,Zhengzhou 450001,China

ABSTRACT β-Conglycinin, the main protein of soybean, is a key allergen that causes soybean allergies, and hydrolysis is usually applied to lower its antigenicity.We evaluated the enzymolysis characters of βconglycinin from the perspective of enzymolysis kinetics using alkaline protease from B.subtilis ACCC 01746.A dynamic model describing the hydrolysis of β-conglycinin was proposed using the initial substrate concentration,enzyme dosage(enzyme to substrate ratio)and hydrolysis time as variables to illustrate the kinetic behavior of enzymatic hydrolysis.The hydrolysis of soybean βconglycinin was carried out at 60 g/L protein concentration, 0.6% enzyme dosage, 55 °C and pH 8.5 to observe the peptides with anti-enzymatic activities.The hydrolysates were gradually fractionated by ultrafiltration through cut-off membranes with molecular weights of 40, 30, 20, and 10 kDa, and their antigenicities were evaluated using indirect competitive enzyme-linked immunosorbent assay.The results showed that the degree of hydrolysis (DH) of β-conglycinin decreased as the β-conglycinin concentration (S0) increased, but increased with enzyme dosage (E0) increasing.Thus, the enzymatic hydrolysis of β-conglycinin followed the first-order kinetics model.The hydrolysis rate (V) was (527.89CE0–2.5533CS0) exp (–0.022DH), the DH-hydrolysis time was 45.454ln[1 + (11.614CE0/CS0–0.0562)t], and the correlated kinetic constants k2 and kd were 527.89 min−1 and 8.6126 min−1, respectively.The hydrolysis behavior of β-conglycinin varied considerably among the α', α, and β subunits.Faster hydrolysis rates were observed for the α'and α subunits compared to the β subunit.The relative molecular weights of the intercepted peptides from the hydrolysates were 14.8–40.1 kDa, and the antigenicity of the peptides with smaller molecular weight was reduced, but not removed completely.However, the alkaline protease from the strain appeared to effectively reduce the allergenicity of β-conglycinin.Therefore,it is possible to produce less allergenic soybean proteins using enzymatic hydrolysis.Additionally, the microbial alkaline protease may serve as a potential novel food enzyme and should be evaluated for the development of hypoallergenic foods.

Keywords:B.subtilis ACCC 01746 Alkaline protease β-Conglycinin Enzymatic kinetics Antigenicity

1.Introduction

Soybean,one of the most important cash crops,has been used in human and animal food industries because of its high protein content [1].However, ingestion of soy or foods containing soy protein can often cause allergic reactions such as skin reactions(skin burning,itching,and tingling), hives, and anaphylactic shock [2].Studies conducted on soybean proteins have shown 21 major allergens[3].Among them,β-conglycinin,as the main allergen accounting for 30%of the soybean protein,causes a wide range of allergic reactions in humans and animals,and is a globulin presented as a trimer (67 kDa of the α subunit,71 kDa of the α’ subunit, and 50 kDa of the β subunit)with a molecular mass of approximately 180 kDa[4].However,because of the extensive utilization of soy protein,it is very difficult to remove allergens from food or feed.Currently,different attempts have been made to reduce the antigenicity of the protein, and various technologies have been employed to control and optimize the processing conditions[5].Because of the close relationship between protein structure and functional properties, enzymatic hydrolysis has been the most powerful method to reduce antigenicity and improve the functional properties of proteins[6].Enzymatic hydrolysis has already been successfully utilized in many studies using acidic and alkaline proteases from plants,animals,or microorganisms[3].

Microbial proteases are a widely studied class of hydrolysis enzymes.They have gained popularity because they can be used to overcome many of the disadvantages of traditional animal/plant proteases,such as low degree of hydrolysis, long time required, and high cost [7].The alkaline proteases from microorganisms were initially sold as detergents and mass produced starting in the late 1960s[8].To date,they are one of the most studied enzymes because of their popular applications in the detergent,leather,textile,and food/feed industries [9,10].Bacillus species (such as B.subtilis)are the vital strains used in enzyme production[11].Wang et al.[5] found that alkaline proteasecatalyzed hydrolysis significantly decreased the antigenicity of soybean meal.Sung et al.[12]demonstrated that fermentation of soybean meal with B.subtilis produced weak immunoreactive products.Therefore, controlled enzymatic hydrolysis can be applied to produce low-antigenic soybean peptides.For the enzymolysis of soybean allergens,it is necessary to control the hydrolysis process reasonably so that a kinetic model which can best describe the reaction is required.The reaction parameters of classical Michaelis–Menten modeling(such as reaction rate constant,order of reaction,kmand Vmax)have been investigated for enzymatic kinetics of proteins(gluten protein,buckwheat protein,pea protein isolate and pecan meal)with enzymes from different sources[13,14].However,there have been no experimental studies on the kinetic model of β-conglycinin hydrolysis.

In a previous study[15],an alkaline protease with molecular weight of 18.3 kDa and the activity of 430 U/g was obtained from crude cultures of B.subtilis ACCC 01746.The effects of temperature, enzyme-to-substrate ratio,time,and pH on the antigenicity of β-conglycinin hydrolysates were investigated.The hydrolysis conditions were optimized by an orthogonal test to reduce antigenicity.The lowest antigenic inhibition rate of 27.03% was achieved under optimized conditions (pH 8.5, substrate concentration 30 mg/mL,alkaline protease dose 0.7%,temperature 55°C,and time 50 min).In this study,alkaline protease was used to digest soybean β-conglycinin samples under the following experimental conditions:temperature 55°C,pH 8.5,and time 160 min.In this paper,we studied the kinetics of alkaline protease-catalyzed hydrolysis of βconglycinin within the Michaelis–Menten framework,and further analyzed the antigenicity of peptides with antienzymatic activities in hydrolysates by indirect competitive enzyme-linked immunosorbent assay(ic-ELISA),aiming to provide a quantifiable controlled method for the preparation of low antigenic peptides.

2.Methods

2.1.Preparation and purification of soybean β-conglycinin

Soybean β-conglycinin (85.40% purity) was prepared from soybean protein isolate(SPI,Henan Kunhua Biological Technology Group Co.Ltd.,Anyang,China)according to the method described by Nagano et al.[16].Highly pure β-conglycinin was produced by combining saturated ammonium sulfate (Analytical grade, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) and Sepharose CL-6B(Beijing Dingguo Biotechnology Co.,Ltd.,China),and analyzed by Quantity One software, showing a purity of 94.5%±1.66%.

2.2.Enzymatic hydrolysis of the purified β-conglycinin

On the basis of previous study [15], the effects of pH,temperature and ratio of enzyme to substrate (E/S ratio)on the hydrolysis of β-conglycinin with alkaline protease from B.subtilis ACCC 01746 were investigated by single factors and L9(34)orthogonal tests,and the optimal pH,temperature and E/S ratio (W/V) were 8.5, 55 °C and 0.6%,respectively.As the understanding of the enzymolysis kinetics,it is important to understand the ability of alkaline protease in enzymatic hydrolysis of soybean β-conglycinin.In this study,the hydrolysis of β-conglycinin was performed under the above optimal conditions to mainly investigate the effects of substrate concentration(20,40,60,80,and 100 g/L) and enzyme dosage (0.2%, 0.4%, 0.6%, 0.8%,and 1.0%)on the degree of hydrolysis(DH).The DH at a given period of time was calculated to explore the hydrolysis dynamics of soybean β-conglycinin.After hydrolysis,the hydrolysates were immediately treated at 95°C for 10 min to terminate the enzymatic reaction.

2.3.Determination of the degree of hydrolysis

Sample preparation,determination of enzymatic hydrolysis conditions, and DH evaluation were carried out as same as the described methods in our previous study[15].The DH of all samples was evaluated(n=3)by calculating the content of free α-amino groups with o-phthaldialdehyde using serine as a standard at different time scales according to Nielsen et al.[17].

2.4.Hydrolysis kinetic model of soybean β-conglycinin

Alkaline protease is a serine protease exhibiting saturation kinetics that can be rationalized based on the Michaelis–Menten model shown in the scheme below:

where an enzyme(E)binds a substrate(S)with combination and dissociation rate constants k1and k−1, respectively, to form the enzyme-substrate complex (Michaelis complex,ES),and k2corresponds to the rate constant of S to produce the product(P).

In this process,the reaction rate(V)of enzymatic hydrolysis is the most important element that defines the catalysis reaction.Therefore,the following equation can be applied.

where V is reaction rate,CS0,t and CESrepresent initial substrate concentration,reaction time,and enzyme-substrate complex concentration,respectively.

However, the reaction could be inhibited by substrate and product in the process of hydrolysis, resulting in enzyme inactivation, and the mechanism and the dynamic model were expressed as follows:

where kdis the inhibition constant,Eais the active enzyme,Eiis the inactive enzyme,Ce,CEand CSrepresent the total enzyme concentration,enzyme concentration and substrate concentration,respectively.

Based on eqs.(2,4),eq.(5)can be deduced as follows:

During enzymatic hydrolysis,the total protease exists in both free and bound forms.Therefore, the relation Ce=CE+ CESwas obtained.In accordance with the Briggs–Haldane hypothesis,the influence of ES decomposition on the enzymatic reaction rate cannot be ignored.The dynamic equilibrium of the enzyme-substrate complexes and the decomposition of the complexes were maintained;for eq.(1),it could be assumed that k1×CE×CS=(k−1+k2)×CES,and then the equations were derived as:

When the CSwas much higher than the CE,it could be considered that CS≈CS0,and eq.(8)could be obtained.

The factor km,known as the Michaelis constant,indicates the substrate concentration at which the enzymatic hydrolysis rate reaches Vmax/2 with Vmaxindicating the maximum rate.This reflects the apparent affinity of the enzyme for its substrate, with a negative correlation.Therefore, the smaller the value of km, the higher the affinity, and vice versa.In the process of normal enzymatic hydrolysis,the kmwas far smaller than CS0.Therefore, eq.(9) can be obtained.

Combining Eqs.(5,9)yields Eq.(10)as follows:

Eq.(10)is integrated in the interval of DH to 0 and Ceto Ce0(initial concentration of total enzyme).Eq.(11)can be expressed as follows:

Therefore,the dynamics model of enzymatic hydrolysis of soybean β-conglycinin by alkaline protease is expressed as follows:

Let α1= (k2× CE0)/CS0and β1= (kd× km)/k2.The Eqs.(12)and(2)could be simplified as:

2.5.Preparation and SDS-PAGE determination of hydrolyzed peptides

2.5.1.Preparation

The above-mentioned hydrolysate obtained under the optimal kinetic conditions (β-conglycinin concentration 60 g/L,enzyme-to-substrate ratio 0.6%,hydrolysis temperature 55°C,and pH 8.5)at given points in time was heated to 55°C for 5 min,then injected to a Sephadex G-100 column(100 cm×2 cm i.d.),equilibrated,and eluted with pH 7.6 PBS(0.4 mol NaCl,32.5 mmol K2HPO4,2.6 mmol KH2PO4, 10 mmol 2-ME) at a flow rate of 1 mL/min to obtain eluate,the hydrolyzed peptides.The absorbance of peptides read at 210 nm.The mixture of peptides derived from dispersive peaks was isolated using an ultrafiltration device (Millipore, Billerica, MA, USA) with molecular weight cut-offs of 40,30,20,and 10 kDa(Merck Millipore Ltd.,Tullagreen,Ireland).

2.5.2.SDS-PAGE determination

The molecular weight distribution of the ultrafiltered peptide samples was assessed using SDS-PAGE.Electrophoresis was carried out according to the method of Laemmli[18]in a discontinuous buffered system using 12.5%separating gels and 5%stacking gels.The hydrolyzed proteins(10 μL, 2 mg/mL) obtained at given points in time were added to 10 μL Tris-HCl buffer(containing 1%SDS,20%glycerol,and 4%2-ME;pH 6.8;Sinopharm Chemical Reagent Co.,Ltd.,Shanghai,China)and heated at 95°C for 5 min before electrophoresis.A mixed sample(10 μL)was loaded onto the gels.A low-molecular-weight standard(Sigma-Aldrich Co., St.Louis, MO, USA) as a marker(14.4–97.4 kDa) was also run along with the samples.SDS-PAGE was performed at a constant 125 V(20 mA stacking gels and 40 mA separating gels)for 90 min.The gel was stained with Coomassie brilliant blue (R-250, Sigma-Aldrich, St.Louis, MO, USA) stain solution (10% acetic acid,25%methanol and 0.25%R-250)and de-stained in 10%acetic acid and 25%methanol in water until the separated protein bands clearly appeared.

2.6.Ultrafiltration and antigenicity assay

Here,antigenicity was expressed in terms of antigen inhibition rate.The antigen inhibition ratio refers to the magnitude of the antigen-inhibiting rabbit antiserum binding ability to the standard antigen in the sample,and a smaller antigen inhibition ratio indicates a lower antigenicity in the sample.

Four fiber membranes used were made of polyethersulfone and polystyrene and produced by Shanghai Motimo Membrane Technology(Shanghai,China)with cut-offs of 10,20,30 and 40 kDa.During the ultrafiltration,the fiber membranes were bundled in a module which contained a peristaltic pump to provide the desired feed pressure and flow velocity.The hydrolysate was divided into four fractions by ultrafiltration with gradient decreasing membrane cut-off.To achieve a relatively complete separation of different molecular weight peptide, the hydrolysate was added with distilled water and filtered until the color of penetrants was clear.The different obtained fractions(10,20,30 and 40 kDa)were characterized by determining antigenicity.The antigenicity of ultrafiltered peptide samples was analyzed using an ic-ELISA as described in our previous study[15].The antigenicity of β-conglycinin was expressed as inhibition rate(%),which was calculated as:

Inhibition rate （% ）= （B0−B）/B0×100

where B represents the absorbance of hydrolyzed βconglycinin samples and B0represents the absorbance of the uninhibited serum sample control.

2.7.Statistical analysis

All data analysis involving one-way analysis of variance(ANOVA)with P<0.05 and means separation(mean values±standard deviation)by Duncan's multiple range testing was performed using SPSS 20.0(SPSS Statistical Software,Inc.,Chicago,IL,USA).

3.Results and discussion

3.1.Effects of substrate concentration and enzyme dosage on DH

The DH,which represents the percentage of cleaved peptide bonds,is one of the basic parameters that describes the properties of the hydrolysates and needs to be controlled during protein hydrolysis[19].This is because several properties of protein hydrolysates are closely related to DH,while substrate concentration and protease dosage are the most important parameters affecting proteolysis[7].Our previous study[15]has demonstrated that initial substrate concentration and alkaline protease dosage influenced DH considerably,we,hereby,explored the effects of the two parameters on DH.Fig.1A showed the changes of DH with hydrolysis time under the initial β-conglycinin concentration of 20,40,60,80 and 100 g/L.We could clearly observed from Fig.1A that the DH at higher β-conglycinin concentrations(80 and 100 g/L)were significantly lower than those at lower β-conglycinin concentrations(20 to 60 g/L).This is most likely because the viscosity of the reaction system increases with the concentration increasing,which further slows down the diffusion of the substrate to the enzyme active center,resulting in a loss of enzymatic activity[20].

Fig.1B showed the evolution of DH with time under different dosages(0.2%to 1.0%)of initial alkaline protease.The results clearly showed that DH increased significantly with alkaline protease dosage increasing from 0.2% to 0.6%, and then the increasing amplitude slowed down.The increasing DH of β-conglycinin hydrolysates with the increase in initial enzyme dosage may be attributable to the greater disaggregation and unfolding of protein and greater exposure of the contact area between enzymes and substrates,thereby improving the extent of subsequent proteolysis.In addition, studies have shown that the production of digestive resistant peptides leads to a decrease in DH [7].A similar conclusion could be reached for soy protein concentrate[21].

Fig.1.Changes of DH with initial substrate concentration (A) and enzyme dosages(B).

3.2.The enzymolysis kinetic model of β-conglycinin

To explain the effect of different initial β-conglycinin concentration CS0and protease dosage CE0on enzymolysis in more detail,the kinetic constants of enzymatic hydrolysis were determined by fitting the data to eq.(13)using linear regression analysis.The α1and β1values obtained by data fitting were listed in Table 1.As shown in Table 1,when the temperature(55°C)was constant,the gradual increase in the dosage of alkaline protease resulted in gradual increase of the α1value.In contrast,the α1value experienced an opposite change with an increase in β-conglycinin concentration,which coincided with the presumed mathematical formula(α1=(k2×CE0)/CS0).That is to say,the α1value was directly proportional to the dosage of alkaline protease but inversely proportional to the concentration of β-conglycinin.At the same time, it can be observed from Table 1 that the value of β1varied little with the dosage of alkaline protease added or the concentration of β-conglycinin.In other words, the β1was not affected by the CS0and CE0(β=(kd×km)/k2),and its values fluctuated over a narrow range and were almost close to one constant,the mean value is 0.022.Therefore,β1can be regarded as a constant under the premise of a temperature-stable system[22].According to the formulas of α1and β1,it can be observed that both α1and α1β1are linearly related to CE0/CS0,and the fitting curves were shown in Figs.2 and 3.The linear regression equations were y = 527.89x −2.5533 (R2=0.9354),and y=8.6126x −0.0315(R2=0.9487)respectively.The results showed that the inactivation constant of the alkaline protease was 8.6126 min−1and that of k2was 527.89 min−1.

Table 1 Kinetic parameters of enzymatic process.

Fig.2.The relationship between α1 and CE0/CS0.

Fig.3.The relationship between α1β1 and CE0/CS0.

Import the expressions of α1(α1= 527.89CE0/CS0−2.5533)and β1(0.022)into eq.(13),and the hydrolysis time(t)-dependent expression for DH and the correlation between V and DH are expressed as follows:

DH=45.454ln[1+(11.614CE0/CS0–0.0562)t]

V=(527.89CE0−2.5533CS0)exp(−0.022DH).

3.3.Verification of the kinetic model

To confirm the legitimacy of the developed model,we performed triplicate tests at an initial substrate concentration of 60 g/L and protease dosage of 0.6% at 55 °C for 160 min.The difference in the DH of the experimental and predicted values was shown in Fig.4A.Both the predicted and the experimental values of DH rose rapidly in the first 80 min and then increased slightly from 80 to 160 min, wherein the experimental value was slightly higher than the predicted value from the 10 min to the 80 min.These results indicated that,at the initial 80 min,the alkaline protease was in full contact with the protein,but after that the hydrolysis rate decreased due to continuous heavy consumption of the protein.Similar resultswere reported by Hu et al.[23].However,the predicted DH values were higher than the experimental values from 80 to 160 min,which due to the fact of product inhibition(Fig.1 A).A similar phenomena at prolonging hydrolysis time has been observed in pecan meal using alcalase[24].Fig.4B indicated the linear relationship between the experimental DH values and predicted results,as evidenced by the high correlation coefficient(R2)value of 0.971.It can be seen that the DH values fitted well with the model with a low average relative error, indicating that the model was relatively reliable and could describe the enzymatic hydrolysis process accurately.

Fig.4.Tendency (A) and correlation (B) between experimental and predicted DHs.

3.4.SDS-PAGE results of hydrolysates

The molecular weights and their distribution of hydrolyzed β-conglycinin at different hydrolysis time were determined using SDS-PAGE, as shown in Fig.5.As expected,the protein was progressively degraded to a variety of micromolecular peptides with the hydrolysis time extending,which could be considered as a releasing component of all subunits in hydrolysis products.As shown in Fig.5,the density of electrophoresis bands of the three subunits(α',α and β)gradually decreased with an increase in enzymolysis time.It was worth noting that the α'subunit disappeared in 20 min, followed by the α subunit in 40 min and the β subunit in 60 min,indicating that the α'and α subunits are easier to be hydrolyzed than the β subunit.A similar view was proposed by Wang et al.[4], in which the β-subunit of β-conglycinin was more stable during enzymatic hydrolysis compared to its α'- and αsubunits.This phenomenon may be due to the differences in amino acid composition and sequence of the three subunits and their locations in the whole molecule [21].At the same time, different molecular weight components were produced(Fig.5).From the released peptides observations,we can approximately see that there are mainly six new peptides (a, b, c, d, e and f).The formation of these bands can be observed in about 10 min of hydrolysis time,and the density of these lines increased with the increase of hydrolysis time.At the same time,the individual bands of βconglycinin were processed by the Gel Pro analyzer 4.0(Media Cybernetics,Inc.,USA).The molecular weight and distribution of the samples treated at different times ranged from 14.8 to 40.1 kDa.The results indicated that there were significant changes in the protein polypeptide structure upon treatment with alkaline protease.The steric conformations of proteins can be influenced by the cleavage of proteins to produce smaller subunits and/or peptides[25].

Fig.5.SDS-PAGE profiles of β-conglycinin at different hydrolysis time.

3.5.The antigenicity determination of ultrafiltered peptide

The effect of enzymatic hydrolysis on the inhibition rate of ultrafiltered peptide was investigated under an initial substrate concentration of 60 g/L and protease dosage of 0.6%at 55°C for 60 min(Fig.6).At the same time,the unpretreated β-conglycinin (60 g/L) was also investigated as control under the same conditions.Compared with the untreated β-conglycinin, enzymatic hydrolysis caused significant decrease of the inhibition rate of ultrafiltered peptide of β-conglycinin hydrolysates, and with the decrease of molecular weight, the inhibition rate of peptide decreased (P < 0.05).It is well known that the antigenicity of a protein depends on its microstructure, while the antigenic epitope (conformational and sequential) plays a vital role in the microstructure evolution process [3].Enzymatic hydrolysis is considered as an effective method to destroy the epitope and eliminate the antigenicity of proteins [5].Antigenic proteins are hydrophilic proteins with a size of 10–70 kDa.Therefore, proteins capable of inducing allergic reactions should have a molecular weight larger than 10 kDa.The results showed that alkaline protease from B.subtilis was highly effective in reducing the antigenicity of β-conglycinin by degrading the βconglycinin and its subunits to micromolecular proteins(less than 40 kDa).Previous studies have shown that enzymes in B.subtilis, such as alkaline protease and Protamex, destroy most antigenic proteins and reduce and remove allergenicity [26].This can be partly explained by the altered molecular weight and epitope of peptides in the entire hydrolysate mixture, where especially in smaller molecular weight, more antigen sites are destroyed [27].The findings once again showed that enzymatic hydrolysis was an ideal way to reduce antigenicity but not eliminate it completely.

Fig.6.The antigen inhibition rate of fractions after ultrafiltration.

4.Conclusion

The influence of enzymatic hydrolysis for 160 min at various substrate concentrations and enzyme dosages resulted in hydrolytic degree changes of β-conglycinin and antigenicity changes of hydrolysates.The results demonstrated that enzymatic hydrolysis with alkaline protease from B.subtilis ACCC 01746 could significantly decrease the antigenicity of β-conglycinin.With an increase in the enzymatic hydrolysis time,the DH increased but the molecular weight decreased,suggesting that the extent of enzymatic hydrolysis might be an important factor in the decrease of antigenicity.Furthermore, with the increase in enzymatic hydrolysis time,the different molecular weight components were produced in the hydrolysates.The changes in these molecular weight may have altered the steric conformations and allergic epitopes of protein.Therefore,enzymatic hydrolysis is an effective method to reduce the antigenicity of β-conglycinin.In addition, based on DH, the kinetic model of enzymatic hydrolysis of β-conglycinin was simulated,and a perfect description of the enzymatic behavior of β-conglycinin by the protease from B.subtilis ACCC 01746 followed first-order kinetics.The kinetic parameters of the enzymolysis for β-conglycinin based on the two models were obtained that the kinetic model of the DH was 45.454ln[1+(11.614CE0/CS0– 0.0562)t] and V was(527.89CE0−2.5533CS0)exp(−0.022 DH),at the same time the correlated kinetic constants k2and kdwere 527.89 min−1and 8.6126 min−1,respectively.Therefore,the DH expression based on experiments would be feasible and convenient in controlling enzymatic hydrolysis, and be helpful to obtain bioactive peptides with low or no antigenicity from soybean protein.The hypoallergenic products in the enzymolysis might be correlated with the decrease of molecular weight by cleaving the peptide bond thereby destroying the epitopes.

Author Contributions

Haicheng Yin:Conceptualization,methodology,validation,writing–review&editing;Xinrui Zhang:Conceptualization;Zhixiang Yang: Conceptualization,investigation;Jin Huang:Conceptualization,data curation.

Conflicts of Interest

Authors declare that there are no conflicts of interest.

Acknowledgement

Authors wish to thank to Grain & Corn Engineering Technology Research Center,State Administration of Grain(GA2017004)for funding support.