Chitosan Complex Behavior and Selective

Jul 7, 2014 - The maximum protein recovery (32%) was obtained at RSWP/Ch = 4:1 and pH 6.3, whereas at RSWP/Ch = 20:1 and pH 5.5, chitosan consumption ...
0 downloads 9 Views 3MB Size
Article pubs.acs.org/JAFC

Soybean Whey Protein/Chitosan Complex Behavior and Selective Recovery of Kunitz Trypsin Inhibitor Xingfei Li, Die Dong, Yufei Hua,* Yeming Chen, Xiangzhen Kong, and Caimeng Zhang State Key Laboratory of Food Science and Technology, School of Food Science and Technology, Jiangnan University, 1800 Lihu Avenue, Wuxi, Jiangsu Province 214122, People’s Republic of China ABSTRACT: Proteins in soybean whey were separated by Tricine−SDS-PAGE and identified by MALDI-TOF/TOF-MS. In addition to β-amylase, soybean agglutinin (SBA), and Kunitz trypsin inhibitor (KTI), a 12 kDa band was found to have an amino acid sequence similar to that of Bowman−Birk protease inhibitor (BBI) and showed both trypsin and chymotrypsin inhibitor activities. The complex behavior of soybean whey proteins (SWP) with chitosan (Ch) as a function of pH and protein to polysaccharide ratio (RSWP/Ch) was studied by turbidimetric titration and SDS-PAGE. During pH titration, the ratio of zeta potentials (absolute values) for proteins to chitosan (|ZSWP|/ZCh) at the initial point of phase separation (pHφ1) was equal to the reciprocal of their mass ratio (SWP/Ch), revealing that the electric neutrality conditions were fulfilled. The maximum protein recovery (32%) was obtained at RSWP/Ch = 4:1 and pH 6.3, whereas at RSWP/Ch = 20:1 and pH 5.5, chitosan consumption was the lowest (0.196 g Ch/g recovered proteins). In the protein−chitosan complex, KTI and the 12 kDa protein were higher in content than SBA and β-amylase. However, if soybean whey was precentrifuged to remove aggregated proteins and interacted with chitosan at the conditions of SWP/Ch = 100:1, pH 4.8, and low ionic strength, KTI was found to be selectively complexed. After removal of chitosan at pH 10, a high-purity KTI (90% by SEC-HPLC) could be obtained. KEYWORDS: complex coacervation, soybean whey protein, chitosan, turbidity, KTI, 12 kDa protein



INTRODUCTION The preparation of soy protein isolate (SPI) includes the acidification of whole aqueous extractable soybean proteins to pH 4.5−4.8 and the centrifugal separation to obtain storage globulin and whey fractions, respectively. In China, millions of tons of soybean whey are produced every year. To recover proteins from soybean whey is valuable economically and environmentally. The acid-precipitable fraction includes the two major soybean storage proteins, glycinin (11S) and βconglycinin (7S). Soybean whey consists of acid-soluble proteins, soybean oligosaccharides, organic acids, and other low molecular weight substances. Soybean whey proteins (SWPs), which make up 9−15.3% of soybean seed protein,1 are mainly composed of lipoxygenase (102 kDa), β-amylase (61.7 kDa), soybean agglutinin (SBA, 33 kDa), Kunitz trypsin inhibitor (KTI, 20 kDa), and Bowman−Birk trypsin inhibitor (BBI, 7.9 kDa).2−4 Due to its high solubility in the acid pH range, soybean whey proteins are a suitable nutritional and functional ingredient for many food formulations as beverages, if they are previously thermally inactivated. In particular, bioactive proteins isolated from soybean whey, such as the protease inhibitor BBI concentrate (BBIC) and purified BBI (PBBI), are now being intensively studied as cancer-preventive agents.5−7 Smith et al.8 showed that the negatively charged colloids, such as alginic acid gum karaya, could be used to recover soybean whey protein efficiently. Gillberg and Tornell9 described that in the presence of sodium alginate and carrageenans, around 90% of proteins could be recovered in the process of preparing rapeseed protein isolates. Protein− polysaccharide systems are also of special interest to the food industry, such as mixtures of whey proteins or β-lactoglobulin © 2014 American Chemical Society

with carrageenan, gum arabic, xanthan, and acacia gum reviewed by Dickinson,10 due to the unique properties for these new protein−polysaccharide systems, such as control or stabilization of enzyme activity by complexation.11,12 Selectivity during complexation means either that a protein preferentially binds one polyelectrolyte or subpopulation thereof among several of similar structure or that a polyelectrolyte preferentially binds one protein among several. The selective recovery of protein or the purification of protein from the mixed proteins solution has been studied for years and is still in development.13−15 Hidalgo 16 found that βlactoglobulin can be precipitated completely by compounding with carboxyl methyl cellulose (CMC), at pH 4.0, with the αlactalbumin remaining in the supernatant. By careful optimization of variables such as pH, ionic strength, and solute concentrations for appropriate protein−polyelectrolyte ratios, proteins were recovered via a well-designed polyelectrolyte coacervation with relatively high efficiency.17 Xu et al.18 found that bovine serum albumin (BSA) and β-lactoglobulin (BLG) showed complex salt dependence so that the choice of ionic strength determines the order of coacervation, whereas the choice of pH controlled the yield of the target protein. Coacervation at I = 100 mM and pH 7.0 of BLG from a 1:1 (w/w) mixture with BSA was shown by SEC to provide 90% purity of BLG with a 20-fold increase in concentration. A similar principle had been used by Du et al.15 to separate BSA Received: Revised: Accepted: Published: 7279

April 22, 2014 June 23, 2014 July 7, 2014 July 7, 2014 dx.doi.org/10.1021/jf501904g | J. Agric. Food Chem. 2014, 62, 7279−7286

Journal of Agricultural and Food Chemistry

Article

Measurement of Turbidity. Both stock solutions were adjusted to a pH of 4.0 separately. Turbidity titration of the SWP/Ch mixture was conducted by first mixing two solutions at the desired ratio (from 100:1 to 1:1) and then adjusting the pH from 4.0 to 10.0. Turbidity was measured using an UV−vis spectrophotometer (Shimadzu, Kyoto, Japan) with a 1 cm cuvette at 600 nm. Turbidimetric titration was carried out by the addition of graded NaOH (1, 0.5, 0.1 M) in increments of ∼0.1 pH unit, to a final pH of 10.0, with gentle stirring and simultaneous monitoring of pH and turbidity. In this way, the added volume could be controlled in the range of 1.5−3% of total volume; thus, the dilution effect was not significant. Titrations were carried out at 25.0 ± 1.0 °C, and the pH (±0.01 pH unit) was monitored with a Mettler Toledo Delta 320 pH-meter that had been carefully calibrated. Homogenous SWP and Ch solutions were used as blanks at their respective concentrations. Homogenous SWP turbidity titration was conducted over the pH range from 3.5 to 7.5. Critical transition points, pHc, pHφ1, pHmax, and pHφ2, were obtained according to methods reported by Mattison.30 All measurements were made in triplicate, using separate stock solutions. Zeta Potential Analyses. Both stock solutions were adjusted to a pH of 4.0 separately. The zeta-potentials (mV) of SWP (0.1%) and chitosan (0.1%) in solution at different pH values ranging from 4.0 to 10.0 were calculated separately from the electrophoretic mobility, at 25.0 ± 0.1 °C, using a Zetasizer Nano ZS instrument (Malvern, UK). All measurements were made in duplicate. Protein Content Determination and SDS-PAGE Analysis. The phase-separated SWP/Ch mixtures were divided into precipitate and supernatant by centrifugation (4000 rpm, 15 min). Quantitation of protein in the supernatant was determined by using the bicinchoninic acid (BCA) method,31 and the protein content of the precipitate was calculated by difference from that of the supernatant. BCA reagent was prepared by mixing 50 volumes reagent A (containing BCA (1%), Na2CO3 (2%), Na2C4H4O6 (0.16%), NaOH (0.4%), NaHCO3 (0.95%), respectively) with 1 volume reagent B (4% CuSO4) when it will be used. Then, 100 μL of the sample and 2 mL of BCA reagent were pipetted into a test tube and mixed thoroughly. After the mixture had stood for 30 min at 37 °C and then the temperature allowed to drop to room temperature, the change in absorbance was monitored at 562 nm, using 100 μL of water as a control. All measurements were made in triplicate. Protein compositions of both parts were analyzed using a Tricine− SDS-PAGE method, according to the method of Schagger.32 Concentrations of stacking and separating gels were 4 and 16%, respectively. 2-Mercaptoethanol was added into the samples to the concentration of 2% (v/v), and samples were heated for 5 min in a boiling water bath. Then 10 μL of each sample was loaded into a sample well, and the samples were electrophoresed at a constant voltage of 30 V until all of protein sample entered into the stacking gel and then at constant voltage of 100 V until the end. The gel was stained using Coomassie Brilliant Blue G-250. SEC-HPLC Analysis. Contents of major protein component in the complex, supernatant, and soybean whey was determined by SECHPLC, using a TSK gel 2000 SWXL (300 mm × 7.8 mm, Tosoh, Tokyo, Japan) column on an Aglient Technologies (USA) 1260 Infinity system equipped with a UV detector, with 20 μL injections. The column was equilibrated and run with ultrapure water containing 45% (v/v) acetonitrile and 0.1% (v/v) trifluoroacetic acid (TFA). The flow rate was 0.5 mL/min, and the wavelength of the UV detector was at 214 nm. Initial sample solutions were directly analyzed after filtration (0.45 μm). MALDI-TOF/TOF-MS. The verification of reported protein components and the identification of unknown proteins in soybean whey were conducted by the MALDI-TOF/TOF-MS analysis. The MALDI-TOF/TOF-MS analysis of sample was conducted according to the methods of Zhao.33 The protein spots were excised from a Coomassie Brilliant Blue G-250 stained gel, washed in distilled water, and destained completely in a 30% acetonitrile (v/v) containing 100 mM NH4HCO3. After a 100% acetonitrile washing, the destained gel piece was placed into reducing solution (10 μL of 100 mM DTT, 90 μL of 100 mM NH4HCO3) for 30 min at 56 °C. After a 100%

from the two protein mixtures, with 90% purity using hyaluronic acid. As biodegradable, biocompatible, and environmentally friendly materials, anionic and cationic polysaccharides offer better methods of coacervation-based protein purification. Chitosan (Ch) is the deacetylated form of chitin and is considered to be the most widely distributed cationic biopolymer. Several complex coacervation pairs based on chitosan have been reported, including α-lactalbumin−chitosan and β-lactoglobulin−chitosan,19 soy globulin−chitosan,20 pea protein isolate−chitosan,21 xanthan−chitosan,22 caseinophosphopeptides−chitosan,23 and gum arabic−chitosan.24,25 However, relatively little is known about the interaction between soybean whey proteins and chitosan. Three types of turbidity titrations had been reported by Dubin et al.26 The experiment that led to the pH dependence of turbidity at constant ionic strength and protein and polymer concentration was called a “type 1” titration or pH titration. One may also follow the evolution of the turbidity upon addition of either protein to polymer (“type 2”) or polymer to protein (“type 3”). In our experiment, type 1 titration was used to study SWP/Ch complexing behavior. Four critical pH values can be observed, which indicated the phase transitions upon changing pH of the system. An initial slight increase in turbidity occurred at pHc, which corresponded to the soluble complex formation. A sharp elevation in turbidity was found at pHφ1, signifying the onset of phase separation. At pHmax, maximum turbidity was obtained, whereas at pHφ2 the coacervate or precipitate started to redissolove.27−29 The aim of this study is to obtain knowledge about the interaction between soybean whey protein and chitosan and to recover proteins from soybean whey. To achieve the goal, protein components in soybean whey were first re-examined. Then, soybean whey protein/chitosan mixtures were brought to complexing conditions by adding NaOH slowly, and effects of pH and protein/chitosan ratio on turbidity and protein recovery were investigated. Finally, protein composition and purity as affected by complexation conditions were studied.



MATERIALS AND METHODS

Materials. Hexane-defatted and flush-desolventized soy flake, provided by Shandong Wonderful Industrial and Commercial Co. Ltd., had a protein content of 52.4% (N × 6.25, dry basis) and a nitrogen solubility index (NSI) of 85%. Food grade chitosan (degree of deacetylation = 80−95%; viscosity average molecular weight = 620 kDa) was purchased from the Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All other reagents were of analytical grade. Preparation of SWP and Ch Stock Solutions. To obtain soybean whey, defatted soybean flake was extracted with distilled water (pH 7.0) with a water-to-meal ratio of 10:1 and blended at room temperature (25 ± 2 °C) for 30 min. After centrifugation of the suspension to remove insoluble precipitate, the supernatant was adjusted to pH 4.5 with hydrochloric acid to precipitate acid-insoluble soy proteins. The supernatant was adjusted to pH 8.0 with 1.0 M NaOH and centrifuged again. The solution was dialyzed for 48 h at 4 °C against water containing 0.02% sodium azide (to prevent microbial growth) and finally preserved at 4 °C. The protein concentration of the laboratory-prepared soybean whey was approximately 0.4% determined by Kjeldahl analysis. The stock solution of 0.1% chitosan (w/v) was prepared by dispersing chitosan powder in 0.1% acetic acid solution (v/v) with magnetic stirring for 3 h at room temperature (25 ± 1 °C). As for the SWP stock solution, soybean whey was diluted with distilled water to a final concentration 0.1% (calculated by protein content). 7280

dx.doi.org/10.1021/jf501904g | J. Agric. Food Chem. 2014, 62, 7279−7286

Journal of Agricultural and Food Chemistry

Article

acetonitrile washing, 70 μL of 100 mM NH4HCO3 and 30 μL of 200 mM iodoacetic acid were added, and the mixture was incubated for 20 min in the dark. After a 100% acetonitrile washing again, 5 μL (10 μg/ mL) of trypsin (Promega, V5111) was added, and the mixture was put into a 4 °C refrigerator for 30 min. Then 25 mM NH4HCO3 solution was added. The peptides were generated after 20 h of incubation at 37 °C. The solution containing peptides was collected and concentrated at 30 °C to get a dry powder. Three microliters of 0.1% TFA (v/v) was used to dissolve the dry powder. In the final step before MALDITOF/TOF analysis, the sample was prepared by mixing 0.7 μL of the sample and 0.7 μL of 4-hydroxy-α-cyanocinnamic acid (used as a matrix) on a MALDI target and air-dried. All mass spectra were obtained with an ultrafleXtreme MALDI-TOF/TOF (Bruker Daltonics, Germany). All mass spectra were analyzed by flexAnalysis software provided by Bruker Daltonics Corp. Protein identification was performed by searching plant proteins in the latest version of the NCBInr database. Trypsin and Chymotrypsin Inhibitor Activity Assays. Trypsin inhibitor activity (TIA) was measured with 0.04% (w/v) Nα-benzoyl-Larginine 4-nitroanilide hydrochloride (BAPA) as the substrate, which dissolved in 0.05 M, pH 8.2, Tris-HCl buffer (1% dimethyl sulfoxide, v/v; 0.02 M CaCl2). Assay solution was zeroed, and then 2 mL of 0.01% (w/v) trypsin in 0.04 mM HCl solution was added to begin the reaction.34 Chymotrypsin inhibitor activity (CIA) was determined by using a method modified from that of Tan et al.35 One milliliter of a diluted sample and 1 mL of deionization water were pipetted into a test tube. Then 4.2 mL of 0.05 M, pH 8.2, Tris-HCl buffer (0.01 M CaCl2) and 2 mL of 0.02% (w/v) α-chymotrypsin in 0.002 M HCl solution were added, followed by 0.8 mL of 0.04% (w/v) N-benzoyl-Ltyrosine p-nitroanilide (BTpNA) in acetone solution and thorough mixing. Exactly 10 min later, the reaction was terminated by adding 1 mL of 30% (w/v) acetic acid. The change in absorbance was monitored at 410 nm for at least 3 min. Trypsin inhibitor units (TIU) or chymotrypsin inhibitor units (CIU) were calculated as the amount of inhibitor that reduced the absorbance per minute of the standard reaction by 0.01. For accuracy, the reaction was measured in the linear portion in the 40−60% inhibition range.

Figure 1. Tricine-SDS-PAGE for SWPs and precipitate proteins: (a) SWP treated with 2-mercaptoethanol (2.0%); (b) precipitate treated with DTT. Lanes 1 and 2 are standards of KTI and BBI; lanes 4−8 represent 200, 300, 600, 800, and 1000 mM DTT treatment, respectively. The DTT treatment samples were chosen as RSWP/Ch = 8:1, which contained the relatively higher content of bands 3−5; all DTT treatment samples were heated for 5 min in a boiling water bath. Band 4′ was derived from bands 3 and 4, and bands 5′ and 5″ were derived from band 5.

Major protein components in soybean whey as separated by Tricine−SDS-PAGE were identified by MALDI-TOF/TOFMS. As shown in Table 1, bands 1 and 2 in Figure 1a were matched to β-amylase and soybean agglutinin, respectively. As expected, bands 3 and 4 in Figure 1a, as well as band 4′ in Figure 1b, were characterized as the same protein: soybean Kunitz trypsin inhibitor (KTI). Bands 5′ and 5″, which derive from the 12 kDa protein, were matched to uncharacterized protein LOC100306176 precursor and unnamed protein product, respectively. The latter had a theoretical nominal mass (Mr) of 8807 and a theoretical calculated pI of 4.51. The amino acid sequence of the latter was further compared with that of Bowman−Birk protease inhibitor (BBI). It can be found in Table 2 that two specific binding domains existed in soybean BBI (IBBD1_SOYBN), which offers the binding sites for trypsin inhibitors,36 that is, CTKSNPPQ and CALSYPAQ. Alfonso37 reported another type of BBI: IBBD2_SOYBN. For this type, CTKSNP was replaced by CTRSMP. It was interesting to find that band 5″ showed a combination of the above: CTRSNP. In addition, there was an extra M residue at the N-terminal. Considering about 30% protein sequence coverage, band 5″ (BBI-U) could be a new member of the BBI family. Trypsin and Chymotrypsin Inhibitor Activity of the 12 kDa Proteins. At RSWP/Ch = 4:1 and pH 6.39, the 12 kDa proteins distribute in both supernatant and precipitate, and the TIA and CIA activities in both parts (Figure 2) were investigated. It can be seen that there were 31.0% of TIA activity and 41.5% of CIA found in supernatants, due to three proteins in supernatants at the ratio of 4:1: SBA, β-amylase, and the 12 kDa proteins (illustrated under Protein Components of SWP/Ch Complexes). SBA and β-amylase showed no TIA activity; thus, the residual TIA was derived from 12 kDa proteins. As one of the 12 kDa proteins, BBI-U had a structure similar to that of BBI and should contribute part of the inhibitor activities. SWP/Ch Complexing Behavior as a Function of pH and Mixing Ratio. Critical Points of SWP/Ch Complexing. pH affected the charge densities and thus the complexing



RESULTS AND DISCUSSION Protein Components in Soybean Whey. The dialyzed soybean whey without further treatment to increase concentration was first analyzed by Tricine−SDS-PAGE, using 2% 2ME as disulfide bond reductant, so as to gain an overview of the protein components in the sample. As shown in Figure 1a, five major bands, band 1−5, were obtained. From the molecular marker, it was known that bands 1 and 2 had molecular weights of 56 and 30 kDa, respectively, whereas twin bands 3 and 4 were centered around 20 kDa. Band 5 displayed as a broad smear at a molecular weight around 12 kDa. To improve the electrophoretic resolution of bands 3, 4, and 5, proteins were enriched by mixing dialyzed soybean whey with chitosan at a ratio of 8:1 and alkalinized to pH 6 (at this condition, there were relatively high amounts of complex). The centrifuging precipitate was recovered and analyzed by Tricine−SDS-PAGE using DTT as disulfide bond reductant. As shown in Figure 1b, bands 1 and 2 were substantially eliminated, but proteins with molecular weight similar to those of bands 3, 4, and 5 were enriched by chitosan complexing. Only one band (4′) was found to correspond with the molecular weight of 20 kDa when using DTT as reductant. Therefore, bands 3 and 4 were actually the same protein showing different electrophoretic mobilities, possibly because of the insufficient reduction of disulfide bonds resulting in inhomogeneous molecular size. In Figure 1b, the original band 5 split in two bands (band 5′ and 5″), and the resolution further improved with the increase of DTT concentration. 7281

dx.doi.org/10.1021/jf501904g | J. Agric. Food Chem. 2014, 62, 7279−7286

Journal of Agricultural and Food Chemistry

Article

Table 1. Identification of Protein Bands in Figure 1 by MALDI-TOF/TOF-MS banda

protein ID

protein score

Mrb

1 2 3 4 5 4′ 5′ 5″

β-amylase (Glycine max) soybean agglutinin complexed with 2,6-pentasaccharide Kunitz trypsin inhibitor (Glycine max) Kunitz trypsin inhibitor (Glycine max) unmatched Kunitz trypsin inhibitor (Glycine max) protein precursor LOC100306176 (Glycine max) unnamed protein

396 378 342 154

5 4 5 2

NBCInr accession no.

(3) (4) (2) (1)

53983 27555 24419 24419

gi|169913 gi|6729836 gi|13375349 gi|13375349

278 115 62

5 (3) 4 (2) 2 (0)

24134 19015 8807

gi|13375349 gi|351722713 gi|58059

peptides matched

a

Band identification numbers correspond to those proteins labeled in Figure 1. bMolecular weights (Mr) are given as theoretical values. Searches were performed via Mascot. Peptides matched and accession numbers within NBCI database are given.

Table 2. Amino Acid Sequence Alignment of the 12 kDa Proteina 5″

1 51

IBBD1_SOYBN

IBBD2_SOYBN

MDDESSKPCC DQCACTRSNP PQCRCSDIRL NSCHSACKSC ICALSYPAQC FCVDITDFCY EPCKPSEDDK EN DDESSKPCC DQCACTKSNP PQCRCSDMRL NSCHSACKSC ICALSYPAQC FCVDITDFCY EPCKPSEDDK EN CTRSMP PQC C MCTRSQPGQC

Amino acid sequences of inhibitory domains are underlined. P1−P′1 are the reactive peptide bond sites, in bold text. Either K or R at position P1 determines specificity for trypsin, whereas L determines specificity against chymotrypsin. The peptides that contributed to protein identification are indicated in italics. a

Figure 3. Turbidity curves for the SWP/Ch system (Cp = 0.10%) and homologous SWP solution as a function of RSWP/Ch in the absence of NaCl. The values of pHc, pHφ1, and pHφ2 were determined graphically as the intersection point of two tangents to the curve, whereas pHmax corresponded to the maximum optical density at 600 nm. (Inset) pH values (pHc, pHφ1, pHmax, and pHφ2) of SWP/Ch system for different ratios.

amine being 6.3. This made it soluble in acid solution, and deprotonation began at pH >6.3. Therefore, the turbidity curve (Figure 3) for chitosan displayed an abrupt increase in turbidity at pKa and then reached a maximum at pH around 7.0, followed by the plateaus of turbidity, corresponding to the insolubility of chitosan. Figure 3 shows that the complexing behavior of SWP/Ch was also substantially affected by the mixing ratio, and the shifts of pH values (pHc, pHφ1, and pHφ2) for SWP/Ch system at different ratios are inset in this figure. It was found that a decrease in SWP/Ch ratio resulted in a shift of the turbidity curves to higher pH ranges. Another study done in our laboratory also found the shift of turbidity curves of soy protein−gum arabic mixtures when the mixing ratio changed.38 The authors noted that, with the increase of soy protein/gum arabic ratio, critical pH values for the formation of insoluble complex shifted to higher values. At the same time, the pH value at which soy proteins and gum arabic carried equal amount but opposite charges also changed in the same way. Therefore, the movement of the turbidity curves seemed to be related to changes of conditions at which the protein/ polysaccharide charge compensation was achieved. To verify the above speculation, a more detailed analysis of electric charge relationships between two biomacromolecules

Figure 2. Trypsin and chymotrypsin inhibitor activity of SWP/Ch system at RSWP/Ch = 4:1. At RSWP/Ch = 4:1, the 12 kDa proteins distribute in both supernatant and precipitate. The activity is expressed as inhibitor units and presented as the mean ± SE (n = 3). Lanes: 1, total protein; 2, proteins remained in the supernatant; 3, precipitated proteins.

behavior of soybean whey proteins and chitosan, as shown in Figure 3. On addition of NaOH to SWP/Ch mixtures that had been adjusted to a pH lower than the isoelectric point of SWP (pH 4.3), we first observed the formation of a soluble complex at pHc, below which electrostatic repulsive forces prohibited the formation of complexes, and the turbidity remained at the baseline. Upon further elevation in alkalinity to pHφ1, an abrupt increase in turbidity was observed, indicating the formation of insoluble complex particles. Turbidity rose to the maximum at pHmax and decreased rapidly as the pH continued to increase, and a turbidity level-off was found at pHφ2. As a control, chitosan was a primary aliphatic amine that can be protonated by selected acids (such as acetic acid), the pK of the chitosan 7282

dx.doi.org/10.1021/jf501904g | J. Agric. Food Chem. 2014, 62, 7279−7286

Journal of Agricultural and Food Chemistry

Article

Figure 4. Zeta potential (mV) values for homogeneous SWP (0.1%) and chitosan (0.1%) in the absence of NaCl (a) as a function of pH (n = 2) and the relationship between [Ch]/[SWP] and |ZSWP|/ZCh as a function of pH (b).

Figure 5. Relationship between turbidity (OD600) and protein recovery (a) and between protein recovery and chitosan consumption (b) for the SWP/Ch (Cp = 0.10%) system. OD600 represented the maximum turbidity of given ratios; the protein recovery corresponded to maximum turbidity.

rapidly at higher pH and more complexes were formed in a limited period of time. Therefore, the position and height of maximum turbidity were affected both by thermodynamic and by kinetic factors. Maximum Turbidity and Protein Recovery. Figure 5a shows the effect of chitosan level on maximum turbidity. It can be observed that turbidity increased rapidly when the amounts of chitosan, which was added into 100 g of soybean whey proteins, increased from 1 to 8.3 g. Further increase in chitosan level resulted in a decrease of turbidity. On the other hand, percent recovery of proteins from soybean whey by centrifugation precipitation of the SWP/Ch complex reflected not only the amount of complex formed but also the deposit ability of the complex. As shown in Figure 5a, protein recovery increased along with turbidity when the chitosan level increased from 1 to 8.33 g/100 g SWP. However, protein recovery kept increasing, although at a slower pace, at higher chitosan level and reached a maximum at 25 g/100 g SWP. A chitosan consumption was defined here as the amount of chitosan used to recover 1 g of proteins in soybean whey. Our experiment (Figure 5b) revealed that although the protein recovery was not very high at a SWP/ Ch ratio of 20:1 and pH 5.5, chitosan consumption was lowest (0.196 g Ch/g recovered proteins) under this condition. Protein Components of SWP/Ch Complexes. Dialyzed soybean whey was mixed at different mixing ratios, and SWP/

was conducted to get a more quantitative understanding of the effect of mixing ratio on the complexing behavior of SWP/ chitosan mixtures. Figure 4a shows that chitosan gave positive zeta potential below about pH 8.3, whereas SWP presented negative zeta potential at pH >4.3. In addition to the nature of electric charges, the absolute value of zeta potentials revealed that the quantities of charges on SWP and Ch, respectively, were also pH dependent. Figure 4b shows the plot of ln(|ZSWP|/ ZCh) as a function of pH. At the same time, ln([Ch]/[SWP]) versus pHφ1 was also plotted in the same coordinate system. It was interesting to find that two sets of data can be described by essentially the same curve. We knew that a mixture with a given Ch/SWP ratio started phase separation at pHφ1. If we looked at the Ch/SWP mixing ratio of 1:8, or ln([Ch]/[SWP]) = −2, the corresponding pHφ1 value was found to be 5.1. At around the same pH (5.05), the natural logarithm of |ZSWP|/ZCh had almost the same value. Thus, our study indicated that at pHφ1, [SWP] × |ZSWP| = [Ch] × ZCh or charge compensation was established. For other mixing ratios, the curve told us that charge compensations were also achieved. In Figure 3, the maximum turbidity occurred at the pH higher than pHφ1, indicating that electric charges on soybean whey proteins and chitosan were not balanced. Most probably, the dissociation of carboxyl groups of proteins and the deprotonation of amino groups of chitosan proceeded more 7283

dx.doi.org/10.1021/jf501904g | J. Agric. Food Chem. 2014, 62, 7279−7286

Journal of Agricultural and Food Chemistry

Article

Figure 6. Tricine−SDS-PAGE for proteins in supernatant (S) and precipitate (P). Lanes: 1, marker; 2, SWP; 3, 5, 7, and 9, supernatants for RSWP/Ch 1:1, 20:1, 40:1, 100:1; 4, 6, 8, and 10, precipitates for the same ratios.

Figure 7 shows that the protein composition of precipitates formed from 4:1 to 40:1. Soybean whey/chitosan mixtures

Ch complexes were formed by pH titrating to the maximum turbidity using NaOH solution, followed by centrifuging to obtain precipitates and supernatants. Protein compositions of complexes and supernatants were analyzed by SDS-PAGE, and the results are shown in Figure 6. In the case of the SWP/Ch ratio range from 1:1 to 20:1 (the other four ratios 16:1−4:1 are not shown in the figure), a negligible content of KTI in the supernatant and low contents of SBA and β-amylase in the precipitates were observed. For the mixture with a SWP/Ch ratio of 40:1, it was estimated that >80% of KTI transferred into the SWP−Ch complex, whereas >70% of SBA and βamylase remained in the supernatants. The inability of SBA and β-amylase to form a complex with chitosan could partly be attributed to their higher isoelectric points (pH 5.8 and 5.6). Chitosan started to deprotonate at a pH higher than the pKa (pH 6.3), at which point SBA and β-amylase were marginally negatively charged. Thus, the electric interaction between proteins and chitosan was weak. From the above, we knew that if the SWP/Ch mixing ratio was less than 20:1, all of the KTI of soybean whey would form a complex with chitosan. Quite unexpectedly, the precipitate that was formed at very low chitosan level (SWP/Ch =100) and low pH (pH 4.8) displayed a very different protein composition, SBA and 12 kDa protein being the major proteins in addition to a small amount of KTI. The precipitate could not be formed by complexing because both SBA and Ch carried the same positive charge at this pH range. The precipitation of SBA in dialyzed soybean was also found at pH around 4.8. Most possibly, this precipitate was caused by the low solubility of SBA in very low salt concentration. Additionally, it was also remarkable to find that all precipitates seemed to contain essentially the same amount of 12 kDa proteins, regardless of the difference in pH and SWP/Ch ratio. Selective Complexing of KTI. Protein separation had been achieved by Xu et al.18 as a consequence of selective protein− polyelectrolyte complexation. This technique may afford the possibility of large-scale protein purification that is fast, efficient, and inexpensive compared to traditional protein separation methods such as chromatography, membrane separation, or centrifugation.39 In this experiment, dialyzed soybean whey was adjusted to pH 4.0 and centrifuged at 4000 rpm to remove self-aggregated SBA and 12 kDa proteins to increase the selectivity of KTI. Protein content in the supernatant was determined and mixed with chitosan according to different SWP/Ch ratios.

Figure 7. Tricine−SDS-PAGE for proteins in precipitates. Lanes: 1, marker; 2−6, precipitates for RSWP/Ch 100:1 (pH 4.86), 40:1 (pH 4.99), 20:1 (pH 5.52), 12:1 (pH 5.88), and 4:1 (pH 6.39).

were similar to the respective protein composition when soybean whey was not precentrifuged, KTI being the major component together with a small amount of 12 kDa proteins, whereas SBA and β-amylase showed only very faint bands. However, the protein composition for 100:1 precipitate was totally different from the above experiment: before precentrifugation, SBA and 12 kDa proteins were the main components, whereas after precentrifugation, KTI became the major protein. Therefore, the removal of self-aggregated SBA and 12 kDa proteins could increase the availability of chitosan for complexing with KTI. As a lectin, SBA is a nonenzyme, nonantibody, and carbohydrate-binding protein40 and is wellknown for its highly specific interactions with galactose/Nacetylgalactosamine. 41 On the other hand, chitosan is composed of randomly distributed β-(1−4)-linked D-glucosamine (deacetylated unit) and N-acetyl-D-glucosamine (acetylated unit), which can provide binding sites for SBA. In contrast to protein−polysaccharide complexing, the specific binding of SBA with chitosan at acidic pH range actually stabilized proteins against precipitation. This was confirmed by the fact that, when the isolated SBA was mixed with chitosan and the pH adjusted to the range of phase separation (pH >6.3), the 7284

dx.doi.org/10.1021/jf501904g | J. Agric. Food Chem. 2014, 62, 7279−7286

Journal of Agricultural and Food Chemistry

Article

the average purity of these three prepared samples was also >90%. Therefore, further studies will be conducted to consistently improve the KTI recovery. In summary, a newly identified protein (12 kDa), which had a structure similar to that of BBI and showed CIA and TIA activities, was found in soybean whey by SDS-PAGE and MALDI-TOF/TOF-MS. The KTI and the 12 kDa protein were the main components to form complexes with chitosan. For different SWP/Ch ratios, charge compensation of system was found to be established at pHφ1, at which phase separation began to occur. Selective complexation between KTI and chitosan was observed only at the condition of SWP/Ch = 100:1, pH 4.8 and low ionic strength. After removing chitosan, a high purity KTI (more than 90%) could be obtained.

obtained mixture gave turbidity lower than that of chitosan at the same condition. Therefore, by partly removing SBA through precentrifugation to reduce the consumption of chitosan, KTI was provided the opportunity to be complexed. As to the behavior of 12 kDa proteins, experimental results suggested that they displayed self-aggregating ability at the same pH as that for SBA and the complexing ability with chitosan under the same conditions for KTI complexing. However, it seemed that interaction between 12 kDa proteins and chitosan was weaker than that between KTI and chitosan, which was possibly because 12 kDa proteins had higher isoelectric points than that for KTI. This can explain why these proteins always coexisted with KTI in precipitates but were less enriched. Kaibara et al.42 reported that ionic strength, polyelectrolyte charge, and protein charge densities affected the protein purification. In their experiment, the selective coacervation of BSA from the BLG was achieved by the nonmonotonic control of ionic strength to enhance protein charge anisotropy. However, in our experiment, with an environment lacking ionic strength, the selectivity of coacervation depended on the charge of protein and chitosan pairs. The stronger binding affinity to chitosan of KTI than of the 12 kDa protein offered us the possibility to separate them by changing the ratio of protein/polysaccharide. As expected, in the low presence of chitosan, the two proteins showed a significant difference in binding ability, giving the priority of KTI to 12 kDa protein binding to chitosan, and a purified KTI was obtained. This KTI−Ch complex obtained was adjusted to pH 10 to precipitate chitosan, followed by centrifugation (4000 rpm, 15 min) to collect the supernatant, which was rich in KTI. The purified KTI contained 95.89% protein, 2.02% polysaccharide, 0.87% ash, and 2.2% water. The TIA of the purified KTI was 5901.3 U/mg, higher than that of standard KTI (Sigma) with 5058.08 U/mg, showing that the presence of chitosan may have no effect on the natural activity of KTI. The purifying effect of KTI by selective complexing with chitosan after precentrifugation was further verified by SECHPLC (Figure 8). Several peaks were observed in the dialyzed



AUTHOR INFORMATION

Corresponding Author

*(Y.H.) Phone: 0510-85917812. Fax: 0510-85329091. E-mail: [email protected]. Funding

The work received financial support from the National Natural Science Foundation of China (No. 21276107), The National Great Project of Scientific and Technical Supporting Programs funded by the Ministry of Science and Technology of China during the 12th 5-year plan (No. 2012BAD34B04-1), and the 863 Program (Hi-tech research and development program of China, No. 2013AA102204-3). Notes

The authors declare no competing financial interest.



REFERENCES

(1) Smith, A. K.; Rackis, J. J.; Isnardi, P.; Cartter, J. L.; Krober, O. A. Nitrogen solubility index, isolated protein yield, and whey nitrogen content of several soybean strains. Cereal Chem. 1966, 43, 261−270. (2) Iwabuchi, S.; Yamauchi, F. Electrophoretic analysis of whey proteins present in soybean globulin fractions. J. Agric. Food Chem. 1987, 35 (2), 205−209. (3) Koshiyama, I.; Kikuchi, M.; Fukushima, D. 2S globulins of soybean seeds. II. Physicochemical and biological properties of protease inhibitors in 2S globulins. J. Agric. Food Chem. 1981, 29 (2), 340−343. (4) Rackis, J. J.; Wolf, W. J.; Baker, E. C. Protease inhibitors in plant foods: content and inactivation. In Nutritional and Toxicological Significance of Enzyme Inhibitors in Foods; Friedman, M., Ed.; Plenum: New York, 1986 (5) Kennedy, A. R. Chemopreventive agents: protease inhibitors. Pharmacol Ther. 1998, 78 (3), 167−209. (6) Armstrong, W. B.; Kennedy, A. R.; Wan, X. S.; Atiba, J.; McLaren, C. E.; Meyskens, F. L. Single-dose administration of Bowman-Birk inhibitor concentrate in patients with oral leukoplakia. Cancer Epidemiol. Biomarkers Prev. 2000, 9 (1), 43−47. (7) Wang, J.; Shen, W. C. Gastric retention and stability of lipidized Bowman-Birk protease inhibitor in mice. Int. J. Pharm. 2000, 204 (1− 2), 111−116. (8) Smith, A. K.; Nash, A. M.; Eldridge, A. C.; Wolf, W. J. Recovery of soybean whey protein with edible gums and detergents. J. Agric. Food Chem. 1962, 10 (4), 302−304. (9) Gillberg, L; Tornell, B. Preparation of rapeseed protein isolates. J. Food Sci. 1976, 41, 1070−1075. (10) Dickinson, E. Stability and rheological implications of electrostatic milk protein−polysaccharide interactions. Trends Food Sci. Technol. 1998, 9, 347−354. (11) Morawetz, H.; Sage, H. The effect of poly(acrylic acid) on the tryptic digestion of hemoglobin. Arch. Biochem. Biophys. 1955, 56, 103−109.

Figure 8. SEC-HPLC analysis for SWP and purified KTI.

soybean whey, and the content of KTI was 33.17% calculated from peak area. After complexing with chitosan, other proteins were substantially removed and the purity of KTI was 92.5%. The problem with the present result was the low (about 10%) recovery of KTI at one time. Two further experiments were also needed to extract the residual KTI in the supernatant, and 7285

dx.doi.org/10.1021/jf501904g | J. Agric. Food Chem. 2014, 62, 7279−7286

Journal of Agricultural and Food Chemistry

Article

(12) Larionova, N. I.; Unksova, L. Y.; Mironov, V. A.; Sakharov, I. Y.; Kazanskaya, N. F.; Berezin, I. V. Study of the complex-formation between soluble carboxymethyl ethers of polysaccharides and proteins. Vysokomol. Soedin., Ser. A 1981, 23, 1823−1829. (13) Izumrudov, V. A.; Galaev, I. Y.; Mattiasson, B. Polycomplexes − potential for bioseparation. Bioseparation 1998, 7, 207−220. (14) Strege, M. A.; Dubin, P. L.; West, J. S.; Flinta, D. In Protein Purification: From Molecular Mechanisms to Large-scale Processes; Ladisch, M., et al., Eds.; American Chemical Society: Washington, DC, USA, 1990; Chapter 5. (15) Du, X. S.; Dubin, P. L.; Hoagland, D. A. Protein-selective coacervation with hyaluronic acid. Biomacromolecules 2014, 15 (3), 726−734. (16) Hidalgo, J.; Hansen, P. M. T. Selective precipitation of whey proteins with carboxymethylcllulose. J. Dairy Sci. 1971, 54, 1270− 1274. (17) Dubin, P. L.; Gao, J.; Mattison, K. W. Protein purification by selective phase separation with polyelectrolytes. Sep. Purif. Methods 1994, 23, 1−16. (18) Xu, Y.; Mazzawi, M.; Chen, K. Protein purification by polyelectrolyte coacervation: influence of protein charge anisotropy on selectivity. Biomacromolecules 2011, 12, 1512−1522. (19) Lee, A. C.; Hong, Y. H. Coacervate formation of α-lactalbumin− chitosan and β-lactoglobulin−chitosan complexes. Food Res. Int. 2009, 42 (5−6), 733−738. (20) Liu, C.; Yang, X. Q.; Lin, M. G.; Zhao, R. Y.; Tang, C. H.; Luo, L.; Liu, L. Complex coacervation of chitosan and soy globulins in aqueous solution: a electrophoretic mobility and light scattering study. Int. J. Food Sci. Technol. 2011, 46 (7), 1363−1369. (21) Elmer, C.; Karaca, A. C.; Low, N. H.; Nickerson, M. T. Complex coacervation in pea protein isolate−chitosan mixtures. Food Res. Int. 2011, 44 (5), 1441−1446. (22) Argin-Soysal, S.; Kofinas, P.; Lo, Y. M. Effect of complexation conditions on xanthan−chitosan polyelectrolyte complex gels. Food Hydrocolloids 2009, 23 (1), 202−209. (23) Hu, B.; Wang, S. S.; Li, J.; Zeng, X. X.; Huang, Q. R. Assembly of bioactive peptide chitosan nanocomplexes. J. Phys. Chem. B 2011, 115, 7515−7523. (24) Coelho, S.; Flores, S. M.; Herrera, J. L. T.; Coelho, M. A. N.; Pereira, M. C.; Rocha, S. Nanostructure of polysaccharide complexes. J. Colloid Interface Sci. 2011, 363 (2), 450−455. (25) Andrews, H. E.; González, J. G. B.; Sosa, F. C.; Carter, E. J. V. Gum arabic−chitosan complex coacervation. Biomacromolecules 2007, 8 (4), 1313−1318. (26) Yoshida, K.; Dubin, P. L. Complex formation between polyacrylic acid and cationic: nonionic mixed micelles: effect of pH on electrostatic interaction and hydrogen bonding. Colloids and Surfaces A: Physicochem. Eng. Aspects 1999, 147, 161−167. (27) Ru, Q.; Wang, Y.; Lee, J.; Ding, Y.; Huang, Q. Turbidity and rheological properties of bovine serum albumin/pectin coacervates: effect of salt concentration and initial protein/polysaccharide ratio. Carbohydr. Polym. 2012, 88, 838−846. (28) Liu, S.; Cao, Y. L.; Ghosh, S. Intermolecular interactions during complex coacervation of pea protein isolate and gum arabic. J. Agric. Food Chem. 2010, 58, 552−556. (29) Weinbreck, F.; Nieuwenhuijse, H.; Robijn, G. W.; de Kruif, C. G. Complexation of whey proteins with carrageenan. J. Agric. Food Chem. 2004, 52 (11), 3550−3555. (30) Mattison, K. W.; Dubin, P. L.; Brittain, I. J. Complex formation between bovine serum albumin and strong polyelectrolytes: effect of polymer charge density. J. Phys. Chem. B 1998, 102, 3830−3836. (31) Smith, P. K.; Krohn, R. I.; Hermanson, G. T.; Mallia, A. K.; Gartner, F. H.; Provenzano, M. D.; Fujimoto, E. K.; Goeke, N. M.; Olson, B. J.; Klenk, D. C. Measurement of protein using bicinchoninic acid. Anal. Biochem 1985, 150, 76−85. (32) Schagger, H. Tricine-SDS-PAGE. Nat. Protoc. 2006, 1, 16−22. (33) Zhao, L. P.; Chen, Y. M.; Cao, Y. Y.; Kong, X. Z.; Hua, Y. F. The integral and extrinsic bioactive proteins in the aqueous extracted soybean oil bodies. J. Agric. Food Chem. 2013, 61, 9727−9733.

(34) Liu, K. S.; Markakis, P. An improved colorimetric method for determining antitryptic activity in soybean products. Cereal Chem. 1989, 66 (5), 415−422. (35) Tan, N. H.; Zubaidah, H. A.; Rahim, H. T. K.; Wong, K. C. Chymotrypsin inhibitor activity in winged beans (Psophocarpus tetragonolobus). J. Agric. Food Chem. 1984, 32 (1), 163−166. (36) Yehudith, B. The Bowman-Birk inhibitor trypsin- and chymotrypsin-inhibitor from soybeans. Int. J. Pept. Protein Res. 1985, 25, 113−131. (37) Alfonso, C.; Francisco, J. M.; Manzano, M. C. M.; Jiménez, E.; Domoney, C. The cytotoxic effect of Bowman−Birk isoinhibitors, IBB1 and IBBD2, from soybean (Glycine max) on HT29 human colorectal cancer cells is related to their intrinsic ability to inhibit serine proteases. Mol. Nutr. Food Res. 2010, 54, 396−405. (38) Dong, D; Hua, Y. F.; Chen, Y. M.; Kong, X. Z.; Zhang, C. M.; Wang, Q. Charge compensation, phase diagram, and protein aggregation in soy protein−gum arabic complex formation. J. Agric. Food Chem. 2013, 61, 3934−3940. (39) Ladisch, M. R.; Willson, R. C.; Painton, C.; Builder, S. E. In Protein Purification: From Molecular Mechanisms to Large Scale Processes; American Chemical Society: Washington, DC, USA, 1990; Chapter 1. (40) Liener, I. E.; Sharon, N.; Goldstein, I. J. The Lectins: Properties, Functions, Applications in Biology and Medicine; Academic: Orlando, FL, 1986; 600 pp. (41) Sharon, N.; Lis, H. How proteins bind carbohydrate: lessons from legume lectins. J. Agric. Food Chem. 2002, 50, 6586−6591. (42) Kaibara, K.; Okazaki, T.; Bohidar, H. B.; Dubin, P. L. pHinduced coacervation in complexes of bovine serum albumin and cationic polyelectrolytes. Biomacromolecules 2000, 1, 100−107.

7286

dx.doi.org/10.1021/jf501904g | J. Agric. Food Chem. 2014, 62, 7279−7286