PEDS Advance Access first published online on August 24, 2007
This version published online on October 4, 2007
Protein Engineering Design and Selection, doi:10.1093/protein/gzm039
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
C-terminus engineering of soybean proglycinin: improvement of emulsifying properties
Laboratory of Food Quality Design and Development, Graduate School of Agriculture, Kyoto University, Uji, Kyoto 611-0011, Japan
* To whom correspondence should be addressed. E-mail: sutsumi{at}kais.kyoto-u.ac.jp
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
Introduction of the extension region of ß-conglycinin
' subunit at the C-terminus of proglycinin A1aB1b results in the improvement of its emulsifying properties. To understand the basic for such improvement, we introduced the ' and extension regions to the A2B1a C-terminus, and the extension and A5A4B3 hypervariable regions, and an oligopeptide composed of 20 negatively or positively charged residues to the A1aB1b C-terminus, creating A2B1a', A2B1a, and A1aB1b, A1aB1bA4IV, A1aB1bNeg and A1aB1bPos, respectively. All the modified versions were produced in Escherichia coli. Their molecular size, thermal stability, surface hydrophobicity, solubility and emulsifying ability were studied. Analyses of molecular size and thermal stability suggested that all the modified versions formed the proper conformation similar to that of the wild type (WT). Solubility was intrinsic to each mutant. At ionic strength 0.5, the emulsifying abilities of all mutants were better than that of the WT except A1aB1bPos and A1aB1bNeg, and at ionic strength 0.08, all mutants especially A1aB1bPos exhibited better emulsifying ability than did the WT. The order of stability of the emulsion at both ionic strengths (0.08 and 0.5) was A1aB1b 
A2B1a > A1aB1b' A2B1a' >> A1aB1bPos > A1aB1bA4IV A1aB1bNeg > A1aB1b, A2B1a. These results indicate that the emulsion stability of proglycinin mutants depends on length and hydropathy profile of the polypeptides added to the C-terminus of proglycinin.
Keywords: emulsion/proglycinin/protein engineering/solubility/soybean
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
There has been a growing interest in the utilization of soybean proteins in food products due to health benefits attributed to soybean proteins and the increasing costs of animal proteins. Soybean protein isolates have been used in the manufacture of yogurts, coffee creamers, whipped toppings and infant formulas to replace, totally or partially, milk proteins (Kolar et al., 1979
). However, the low functionality of commercially available products has limited the development of some types of products. Presently, there is a renewed interest in the study of soybean protein functionality, in part because of the increased availability of soybean protein ingredients with good quality and, perhaps more importantly, because of the increased consumer demand for novel products containing soybean proteins. This demand is at least in part caused by the health claims related to the consumption of products containing sufficient amounts of soybean protein (Food and Drug Administration, 1999).
Soybean (Glycine max L.) protein is composed of two major components, glycinin (11S globulin) and ß-conglycinin (7S globulin), accounting for 40 and 30% of the total seed proteins, respectively (Utsumi, 1992a; Utsumi et al., 1997). ß-Conglycinin is a trimeric protein composed of three subunits: (
67 kDa), ' (71 kDa) and ß (50 kDa). According to the amino acid sequences deduced from the nucleotide sequences, the and ' subunits contain extension regions, 125 and 141 amino acid residues for and ', respectively, in addition to the core regions (414–418 residues) which are common to all three subunits (Maruyama et al., 1998). The homology between the core regions of the subunits are 71–87%, and that of the and ' extension regions is 57.3% (Maruyama et al., 1998). The extension regions are rich in acidic amino acid residues. On the other hand, glycinin is a hexameric protein composed of five major subunits (A1aB1b, A1bB2, A2B1a, A3B4 and A5A4B3), each of which consists of an acidic (30 kDa) and a basic (20 kDa) polypeptide linked by a single disulfide bond, except for the acidic polypeptide A4 of A5A4B3 (Staswick et al., 1984). The five subunits have been classified into two groups based on homology in their sequences. Group I consists of A1aB1b (53.6 kDa), A1bB2 (52.2 kDa) and A2B1a (52.4 kDa), and group II consists of A3B4 (55.4 kDa) and A5A4B3 (61.2 kDa). The homology of each subunit is more than 84% within a group and 45–49% between groups (Utsumi et al., 1997; Nielsen et al., 1989). According to the amino acid sequences deduced from the nucleotide sequences of the five subunits, the main difference in the subunits is due to the hypervariable regions which are at the C-termini of their acidic polypeptides and consist of 43, 29, 35, 70 and 103 amino acid residues for A1aB1b, A1bB2, A2B1a, A2B1a, A3B4 and A5A4B3, respectively (Nielsen et al., 1989; Lawrence et al., 1994; Adachi et al., 2001).
In developing seeds, the constituent subunits of 11S globulin are synthesized as a single polypeptide precursor, preproprotein, the signal sequence of which is removed cotranslationally. The resultant proproteins assemble into trimers of 8S in the endoplasmic reticulum. The proprotein trimers are transported from the endoplasmic reticulum to protein storage vacuoles (PSVs), where they then are cleaved to form acidic and basic polypeptides that are linked by a disulfide bond (Dickinson et al., 1989). Finally, the mature proteins assemble into hexamers. The protein reserves are stored in the dormant seed until its germination.
Studies on glycinin crystal structures indicated that the two proglycinin molecules (trimers) combine to form the mature glycinin (hexamer) between IE-faces after processing (Adachi et al., 2001, 2003a). This means that the recombinant protein (proglycinin) instead of the mature protein should be engineered to assess modifications before utilization in genetic crop improvement.
The emulsifying property of a protein is one of its important functional properties in relation to its application in food systems (Dickinson, 1992). Proteins with both hydrophilic and hydrophobic regions can exhibit emulsifying property (Utsumi, 1992b). In oil–water emulsion system, proteins migrate to and interassociate with the oil–water interface to be absorbed and be partially unfolded there, depending on their structure (Kinsella et al., 1985). The relationship between physicochemical characteristics (solubility and surface hydrophobicity) and emulsifying property of proteins was extensively investigated (Graham and Phillips, 1976; Phillips, 1981; Dickinson et al., 1988; Damodaran, 1997). It was suggested that the balance of surface hydrophobicity and hydrophilicity is one of the characteristics of the protein most likely to define its surface behaviors and consequently its emulsifying property. Though such theories have been suggested to explain the emulsifying properties of proteins, improvement of the emulsifying property of a protein has been limited. The emulsifying property of soybean proteins has been extensively studied (Wagner and Guéguen, 1995, 1999; Maruyama et al., 1999, 2002, 2004; Liu et al., 1999; Palazolo et al., 2003; Prak et al., 2005; Roudsari et al., 2006). Soybean protein may be used to stabilize oil/water emulsion because of the surface properties of its constituent proteins (Palazolo et al., 2003). Its solubility and surface-active properties were improved when its oligomeric structure was appropriately dissociated with simultaneous unfolding of the acidic and basic polypeptide chains (Kim and Kinsella, 1987). It was later confirmed that the acidic polypeptide of soybean glycinin has good emulsifying ability (Liu et al., 1999). On the other hand, analysis of the individual ß-conglycinin subunits showed that the extension regions of the and ' subunits that are rich in negatively charged residues confer better solubility and emulsifying ability to the subunits while the core domains determine their thermal stability (Maruyama et al., 1998, 1999). This is consistent with the observation that the negatively charged hypervariable region also contributes to the solubility and emulsifying ability of glycinin, although the extent is lower than that of the extension region (Maruyama et al., 2004). However, previously it was found that the addition of hypervariable region (42 aa) of proglycinin A1aB1b to its variable regions II and III and C-terminus, and the ' extension region (141 aa) to proglycinin A1aB1b hypervariable region did not improve the emulsifying property of the protein. On the other hand, addition of the ' extension region to the A1aB1b C-terminus improved remarkably its emulsifying ability and emulsion stability (Tandang et al., 2005). In addition, changing the position of the extension region from the N- to C-terminus of ß-conglycinin ' subunit resulted in improved emulsion stability. The 30 amino acid residues (residues 1–30) at the N-terminus of the ' extension region are less hydrophilic than the 30 amino acid residues (residues 111–141) at its C-terminus. Tandang et al. (2005) suggested that introduction of the ' extension polypeptide to the C-terminus of proglycinin could be a good strategy to improve the emulsifying property of proglycinin. However, the structural characteristics of the extension region such as the length, and the ratio and the distribution of hydrophilic residues being important for the improvement are not known.
It is noted that the extension (125 aa) and A5A4B3 hypervariable (103 aa) regions are also long and rich in charged residues. Moreover, the hydropathy profiles of the ' and extension and A5A4B3 hypervariable regions are different (Fig. 1). In the present study, to elucidate the structural characteristics of the extension region required for improving the emulsifying properties of proglycinin, we introduced the ' and extension regions to the A2B1a C-terminus, and the extension and A5A4B3 hypervariable (A4IV) regions to the A1aB1b C-terminus. To evaluate the influence of the length and the charge of a polypeptide introduced at the C-terminus, we also introduced oligopeptides composed of 20 negatively (DDDDDDDDDDEEEEEEEEEE) or positively charged residues (RKKKKRKRKKRRKRKRRRRK) to the A1aB1b C-terminus (Fig. 1). All the mutants were produced in Escherichia coli, and their structures were characterized according to molecular size and thermal stability, and their physicochemical properties such as surface hydrophobicity, solubility and emulsifying property were studied.
|
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
Construction of expression plasmids for mutants
Schematic representations of proglycinin wild types (WTs) and their mutants are shown in Fig. 2. To construct the expression plasmids for the mutants, the expression plasmids pEA1aB1b (Katsube et al., 1999), pEA2B1a, pEA5A4B3 (Prak et al., 2005), pEC and pEC' (Maruyama et al., 1998) were used as templates for PCR. The different primers used in amplifying the desired mutant cDNAs by PCR using KOD plus (TOYOBO) are as follows: A1aB1b'; pEA1aB1b as a template, 5'-TAG AAT TCC GGA TCC GAA TTC GAG CTC-3' and 5'-AGC CAC AGC TCT CTT CTG AGA CTC C-3', pEC' as a template, 5'-GTG GAG GAA GAA GAA GAA TGC GAA GAA GGT C-3' and 5'-TGG TTC TCT TTG AGA CTC AGA ACC TTC-3', A1aB1b; pEA1aB1b as a template, 5'-TAG AAT TCC GGA TCC GAA TTC GAG CTC-3' and 5'-AGC CAC AGC TCT CTT CTG AGA CTC C-3', pEC as a template, 5'-GTG GAG AAA GAA GAA TGT G-3' and 5'-TAA CTC AGA ATC TTC ACT TTC TTC GCT-3', A1aB1bA4IV; pEA1aB1b as a template, 5'-TAG AAT TCC GGA TCC GAA TTC GAG CTC-3' and 5'-AGC CAC AGC TCT CTT CTG AGA CTC C-3', pEA5A4B3 as a template, 5'-AAG TGG CAA GAA CAA CAA GAT GAA G-3' and 5'-TCT TGT CTC GCA TCC TCT TTC ACG T-3', A1aB1bNeg; pEA1aB1b as a template, 5'-GAA GAA GAA GAA GAA GAG GAG GAG GAG GAG TAG AGC CCT TTT TGT ATG GAT CCG-3' and 5'-GTC GTC GTC ATC ATC ATC ATC ATC ATC ATC AGC CAC AGC TCT CTT CTG AGA CTC-3', A1aB1bPos; pEA1aB1b as a template, 5'-CGC CGT AAG CGT AAA CGT CGT CGT CGT AAA TAG AGC CCT TTT TGT ATC GAT CCG-3' and 5'-TTT CTT CCG CTT GCG CTT CTT CTT CTT GCG AGC CAC AGC TCT CTT CTG AGA CTC-3', A2B1a'; pEA2B1a as a template, 5'-TAG AAG CTT GCG GCC GCA CTC GAG CAC-3' and 5'-AGC CAC AGC TCT CCT CTG AGA CTC-3', pEC'as a template, 5'-GTG GAG GAA GAA GAA GAA TGC GAA GAA GGT C-3' and 5'-TGG TTC TCT TTG AGA CTC AGA ACC TTC-3', A2B1a; pEA2B1a as a template, 5'-TAG AAG CTT GCG GCC GCA CTC GAG CAC-3' and 5'-AGC CAC AGC TCT CCT CTG AGA CTC-3', pEC as a template, 5'-GTG GAG AAA GAA GAA TGT G-3' and 5'-TAA CTC AGA ATC TTC ACT TTC TTC GCT-3' (italic and bold letters are a stop codon and codons for the insertion of charged amino acid residues, respectively).
|
For amplifying DNAs encoding pEA1aB1b and pEA2B1a, 30 cycles of denaturation at 94°C for 15 s, annealing at 60°C for 30 s and elongation at 68°C for 10 min were used. The cDNAs encoding the
and ' extension and A5A4B3 hypervariable regions were amplified by 25 cycles of 94°C for 15 s, annealing at 60°C for 30 s and elongation at 68°C for 3 min. The resulting fragments corresponding to the and ' extension and A5A4B3 hypervariable regions were phosphorylated before ligations with the corresponding vectors to construct the expression plasmids pEA1aB1b', pEA1aB1b, pEA1aB1bA4IV, pEA2B1a' and pEA2B1a. For construction of pEA1aB1bNeg and pEA1aB1bPos, PCR reactions were 30 cycles of denaturation at 94°C for 15 s, annealing at 65°C for 40 s and elongation at 72°C for 7 min, and extended for 10 min at 72°C, and the resulting fragments were phosphorylated and ligated.
The expression plasmids were transformed into E.coli expression hosts HMS174(DE3), BL21(DE3), AD494(DE3) and Origami(DE3). Culture and expression conditions of A1aB1b and A2B1a WTs were as described previously (Prak et al., 2005). The cells containing individual expression plasmids were grown at 37°C. When A600 reached 0.4 to 0.6, expression was induced with 1 mM isopropyl-1-thio-ß-D-galactoside (IPTG) at 18 and 20°C. After cultivation, cells were harvested by centrifugation at 9000 xg for 15 min at 4°C and stored in –20°C until used. Proteins in aliquots of the cells were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) (Laemmli, 1970) using 11% gel as described previously (Kim et al., 1990). Expressed recombinant proteins were identified based on their expected sizes and confirmed by western blotting (Prak et al., 2005) using anti-glycinin antibody followed by goat-rabbit IgG-alkaline phosphatase conjugate (Promega).
Purification of wild types and their modified versions
All purification steps were carried out at 4°C and centrifugation was at 9000 xg for 20 min unless otherwise stated. The basic buffer for all purification steps was buffer A (35 mM potassium phosphate, pH 7.6, 1 mM EDTA, 10 mM 2-mercaptoethanol, 0.1 mM (p-amidinophenyl)-methylsulfonyl fluoride, 1 µg/ml pepstatin A, 1 µg/ml leupeptin). Ammonium sulfate fractionation followed the procedure of Green and Hughes (1955).
A1aB1b and A2B1a WTs were purified as described previously (Prak et al., 2005). Frozen cells containing A1aB1b' and A1aB1b were resuspended in buffer B (buffer A containing 0.4 M NaCl); cells containing A1aB1bA4IV, A1aB1bNeg, A1aB1bPos, A2B1a' and A2B1a were resuspended in buffer C (buffer A containing 1.0 M NaCl) at a density of 40 g/l buffer and lysed by sonication on an ice bath. Insoluble matter was removed by centrifugation. Expressed modified proteins were fractionated using ammonium sulfate: 30% for A1aB1b', A1aB1b, A2B1a', A2B1a and A1aB1bNeg; 35% for A1aB1bA4IV and A1aB1bPos. The precipitate was removed by centrifugation, and the soluble fraction containing recombinant proteins was applied onto a Toyopearl (Butyl-650M) (TOSOH, Japan) column (2.6 cm x 20 cm) equilibrated with buffer B or C containing 30% ammonium sulfate. Elution was carried out with a linear gradient (800 ml) from 30 to 0% of ammonium sulfate in buffer B for A1aB1b' and A1aB1b, and in buffer C for the other modified proteins. Fractions containing A1aB1b' and A1aB1b were dialyzed against buffer D (buffer A containing 0.15 M NaCl) and clarified by centrifugation. The dialysates were applied onto Mono Q HR 10/10 column (Pharmacia Biotech) equilibrated with buffer C. Elution was performed with a linear gradient from 0.15 to 0.6 M of NaCl in buffer A over a period of 120 min at 2 ml/min. Fractions containing A1aB1bA4IV, A1aB1bNeg, A2B1a' and A2B1a were pooled, and concentrated by VIVASPIN 20 (VIVASCIEN, Japan) and applied on a gel filtration column (Hi-Prep 26/60 Sephacryl S-300 HR) using buffer C as a mobile phase. However, fractions containing A1aB1bPos were concentrated by dialysis in buffer C containing 75% ammonium sulfate. The precipitate containing A1aB1bPos was dissolved in 2 ml buffer C and then it was applied on the same gel filtration column.
The levels of protein expression and purity of the protein samples were analyzed based on densitometric scan estimated by analyzing the gel image with ImageMaster 1D Elite, version 3.0 (Amersham-Pharmachia Biotech, Uppsala, Sweden).
The amount of protein in the samples was determined using a Protein Assay Rapid Kit (Wako) with bovine serum albumin as a standard.
Analysis of self-assembly into trimers
Self-assembly of individual modified proteins was analyzed using Hi-Prep 16/60 Sephacryl S-300 HR column (Pharmacia Biotech) as described previously (Prak et al., 2005). All the samples used were 500 µl of 0.25 mg/ml in buffer E [35 mM sodium phosphate, pH 7.6, 0.4 M NaCl, 1 mM EDTA, 0.1 mM (p-amidinophenyl)-methylsulfonyl fluoride, 1 µg/ml pepstatin A, 1 µg/ml leupeptin, 0.02% NaN3, 10 mM 2-mercaptoethanol] except for A1aB1bPos, which was in buffer F (identical to buffer E but containing 1.0 M NaCl). The mobile phase used for all the samples was buffer E and the flow rate was 0.5 ml/min.
Differential scanning calorimetry measurement
Differential scanning calorimetry (DSC) measurement of the samples was carried out as described previously (Prak et al., 2005) using 1 mg/ml of the sample in buffer E. Scanning was recorded using Microcal MC-2 Ultra Sensitive Microcalorimeter (Micro Cal Inc., Northampton, MA) at a rate of 1°C/min.
Surface hydrophobicities of the samples were analyzed as described previously (Prak et al., 2005) using butyl and phenyl sepharose columns (Amersham Bioscience, Sweden), and 500 µl of the samples (0.25 mg/ml) in buffer G (buffer E containing 35% saturation of ammonium sulfate). The proteins were eluted with a linear gradient from 35 to 0% of ammonium sulfate over a period of 55 min and further with buffer E for 45 min at a flow rate of 0.25 ml/min.
Solubility analysis as a function of pH
Except A1aB1bPos, all samples were dialyzed against buffer H [10 mM sodium phosphate, pH 7.6, 0.5 M NaCl, 1 mM EDTA, 0.1 mM (p-amidinophenyl)-methylsulfonyl fluoride, 1 µg/ml pepstatin A, 1 µg/ml leupeptin, 0.02% NaN3, 10 mM 2-mercaptoethanol]. Experimental conditions were the same as described previously (Prak et al., 2005). A1aB1bPos precipitated in buffer H. To avoid the precipitation before analysis of its solubility, A1aB1bPos was dialyzed against buffer I (5 mM sodium phosphate, pH 7.6, µ = 0.014). For the analysis at the same ionic strength and pH as the other samples, the buffer at different pHs used for the other samples were used, and the final ionic strength 0.08 and 0.5 of the sample was adjusted by adding appropriate amount of buffer J (buffer I containing 3.42 M NaCl). The sample (0.8 mg/ml) was then analyzed by the same procedure as the other samples.
The emulsifying properties of samples were analyzed as described previously (Tandang et al., 2005) using 0.5 mg/ml of proteins in buffer E and buffer K (identical to buffer H but containing 0.05 M NaCl) for µ = 0.5 and 0.08, respectively.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
Expression and purification of individual modified proteins
AD494(DE3) was a good expression host for A1aB1b and A2B1a WTs. The expression levels of these proglycinins as soluble proteins were estimated to be about 15% of the total proteins (Prak et al., 2005). However, the performance of the expression host differed depending on the individual modified proteins. After assessment of the expression level and solubility of the recombinant proteins, the most suitable E.coli strain for each plasmid was determined to be as follows: HMS174(DE3) for A1aB1b', Origami(DE3) for A1aB1bNeg, BL21(DE3) for A1aB1bA4IV and A1aB1bPos, and AD494(DE3) for A1aB1b, A2B1a' and A2B1a. To achieve a high level expression of a soluble modified protein, LB medium containing NaCl from 0.4 to 0.6 M was used and the most suitable E.coli strain harboring each individual expression plasmid was cultured at 18–20°C for 20–40 h (Table I). In optimizing expression conditions, we found that in any condition used in this study the expression level of the modified versions of A2B1a were much lower (<5% of the total proteins) than that of the modified versions of A1aB1b. This expression behavior was consistent with that of their WTs (Prak et al., 2005). This is the reason why A1aB1b was more preferentially modified than A2B1a in this study.
|
The solubility of recombinant proteins from individual expression plasmids in E.coli cells was analyzed by SDS–PAGE after sonication (Fig. 3). On the basis of band intensity, the solubility of all the modified proteins was estimated to be >80% based on densitometric scan estimated by analyzing the gel image with ImageMaster 1D Elite, version 3.0 (Amersham-Pharmachia Biotech, Uppsala, Sweden). These were confirmed by western blotting (Prak et al., 2005
) using purified anti-glycinin serum (data not shown).
|
After purification, the purity of the proteins was assessed by SDS–PAGE (Fig. 4) and was found to be >95% based on the densitometric scan. We found that the mobilities of all modified proteins corresponded to their expected molecular sizes. To use these modified proteins for further analysis, it was necessary to confirm that the modified proteins could form proper conformation similar to those of the WTs. To investigate this point, at first, the individual purified WTs and their modified versions were subjected to gel filtration chromatography using Hi-Prep 16/60 Sephacryl S-300 HR column at pH 7.6 and µ = 0.5. Proglycinin A1aB1b and A2B1a WTs can self-assemble into trimers (Prak et al., 2005
). This was confirmed as shown in Fig. 5. From the elution time and the molecular masses of the modified proteins and WTs, we can say that most of all modified versions eluted from a gel filtration column according to their molecular masses except for A1aB1bPos (169.5 kDa) and A1aB1bNeg (168.3 kDa), which eluted slower (122.7 min) and faster (112.6 min) than the times expected from their molecular sizes, respectively. This phenomenon may be due to the fact that in the form of trimers the C-terminus of a monomer is located near the disordered regions II and IV of the other monomer (Fig. 6) (Adachi et al., 2001). The two disordered regions, especially disordered region IV, are rich in negatively charged residues (Nielsen et al., 1989; Lawrence et al., 1994; Adachi et al., 2001). Therefore, there could be interaction between the negatively charged disordered region IV and the positively charged 20 amino acid residues that were introduced to the C-terminus, resulting in a molecular dimension smaller than that expected from its molecular size. In the case of A1aB1bNeg, the negatively charged 20 amino acid residues at the C-terminus might be repulsed by the negatively charged residues in its variable regions II and IV, resulting in a molecular dimension bigger than that expected from its molecular mass. A2B1a' containing seven amino acid residues more than A1aB1b eluted later than A1aB1b. However, this phenomenon is not surprising since A2B1a WT had the most compact structure among five proglycinins (Prak et al., 2005). The folding property of the modified versions of A2B1a should be close to that of A2B1a WT and should exhibit similar elution behaviors. Thus, all the modified versions of A1aB1b and A2B1a were assessed to be trimers similar to their WTs.
|
|
|
DSC analysis for comparison of thermal stability of WTs and their modified versions (Table II) showed that Tm values of all A1aB1b and A2B1a modified versions were slightly (0.4–3.4°C) lower or higher than those of the WTs. The present thermal stability analysis data are consistent with our previous study on the analogous modified versions of A1aB1bs (Tandang et al., 2005
). Previously, we found that the mutations of soybean proglycinin A1aB1b to delete one of the two disulfide bonds resulted in lower Tm values of the protein for 2.3 and 3.4°C, respectively. However, their crystal structures determined by X-ray crystallography showed that the mutants form correct conformations identical to that of the WT except the vicinities of the mutation sites (Adachi et al., 2003b). Therefore, A1aB1bA4IV, A1aB1bPos, A2B1a'and A2B1a which exhibited a little lower Tm values (3.4, 2.9, 3.0 and 2.3°C, respectively) than those of the WTs probably formed conformation similar to those of the WTs like A1aB1b', A1aB1b and A1aB1bNeg which exhibited Tm values close to that of the WT.
|
Surface hydrophobicity
The surface hydrophobicity of food proteins is related to some of their physicochemical properties such as emulsification and solubility (Nakai, 1983; Nakai and Li-Chan, 1988). Kato and Nakai (1980) described that cis-parinaric acid (CPA) is more suitable for the measurement of surface hydrophobicity than 1-anilino-8-naphthalenesulfonate (ANS). Hayakawa and Nakai (1985) reported that hydrophobicity measured by phenyl sepharose column chromatography and ANS correlated well with the protein insolubility, whereas no significant correlation was observed between CPA and the protein insolubility. They suggested that aromatic hydrophobicity might play a more important role in protein solubility than aliphatic hydrophobicity. Therefore, we employed two columns of phenyl and butyl sepharose for the measurement of surface hydrophobicity. With this analysis, the longer the elution time of the sample, the greater the surface hydrophobicity. A2B1a WT was insoluble at the concentration (30%) of ammonium sulfate used for hydrophobic columns, but its modified versions were soluble at the same concentration, indicating that the surface hydrophobicity of the modified versions was lower than that of A2B1a WT. The surface hydrophobicities of all the modified versions accessed using the two columns were lower than those of the WTs except that of A1aB1bPos (Table III). This might be due to the covering of hydrophobic regions by highly hydrophilic polypeptides introduced to the C-terminus, which contain 79.4, 74.4, 80.6 and 100% hydrophilic residues for the 'extension region, extension region, A5A4B3 hypervariable region and twenty negatively charged residues, respectively. However, A1aB1bPos eluted at 7 and 5.5 min later than A1aB1b WT from butyl and phenyl sepharose, respectively, indicating that introduction of a fully positively charged oligopeptide did not reduce the surface hydrophobicity but increased it instead. Increased surface hydrophobicity of A1aB1bPos might be due to the interaction between positively charged twenty amino acid residues and the negatively charged amino acids residues in disorder regions II and IV, which were at an IE face (Adachi et al., 2001), contributing to the increase in the surface hydrophobicity at the IE face. Hence, the resulting surface hydrophobicity of the A1aB1bPos was higher than the WT. These discussions are consistent with those for the behavior of the modified proteins on gel filtration.
|
Solubility as a function of pH
Solubility is a fundamental physicochemical property of food proteins (Kinsella, 1979; Peng et al., 1984; Bilgi and Çelik, 2004). We measured the solubility of A1aB1b WT and its modified versions (Fig. 7A) and A2B1a WT and its modified versions (Fig. 7B) at high (µ = 0.5) and low (µ = 0.08) ionic strengths. At µ = 0.5, the solubility of A1aB1b' and A1aB1bA4IV was similar to that of A1aB1b WT. A1aB1b and A1aB1bNeg exhibited similar solubility profiles. They exhibited low and high solubilities at pH 
4 and pH 5, respectively. At this ionic strength, A1aB1bPos exhibited unpredicted solubility profile and was mostly insoluble in all pHs. The solubility of this modified protein was never beyond 40%. At µ = 0.08, the solubility profiles of A1aB1b' and A1aB1bA4IV shifted to lower pHs than that of A1aB1b WT. At this ionic strength, A1aB1bA4IV exhibited solubility similar to that of A5A4B3 rather than that of A1aB1b WT (Maruyama et al., 2004). A1aB1b and A1aB1bNeg exhibited similar solubility profiles and their solubilities were < 30% at pH < 4.5. The solubility of A1aB1bPos increased from 0% to the level of 70% when the pH was increased from 3.5 to 8.0. Since histidine residue is charged at pH < 4.5, the number of negatively and positively charged residues in the introduced peptide ' extension, extension, A4IV and 20 negatively charged residues were 53 and 35, 52 and 24, 41 and 24, and 20 and 0 residues, respectively. The positively charged residues occupy 39.8, 31.6, 36.9, and 0% of the total charged residues, respectively. The solubility of the mutants at pH < 4.5 were A1aB1b'> A1aB1bA4IV > A1aB1b A1aB1bNeg. This order is proportional to that of the occupancy of the positively charged residues, suggesting that the occupancy of the positively charged residues is an important factor for the solubility at pH <4.5.
|
, A1aB1b WT; •, A1aB1b
, A1aB1b
, A1aB1bA4IV; 
, A1aB1bNeg; 
, A1aB1bPos. (B) A2B1a and its modified versions; 
, A2B1a WT; 
, A2B1aThe solubilities of A2B1a
' and A2B1a were similar to that of A2B1a WT at µ = 0.5. At µ = 0.08, A2B1a showed clear isoelectric precipitation at pH 4.0 to 5.2 which were not similar to A2B1a WT. The solubility behavior of all the modified proteins might be due to the combinations of the effects of (i) change of electrostatic potential of charged residues by pH (Nagano et al., 1994, Grimsley et al., 1999; Lindman et al., 2006), (ii) increase of hydrophobic interaction by salt (Lee et al., 2002), (iii) neutralization by salt (Utsumi and Kinsella, 1985, Lee et al., 2002; Lindman et al., 2006), (iv) increase of negatively or positively charged residues (Takano et al., 2003), and (v) neutralization by electrostatic interaction of positively charged residues at the C-terminus with the negatively charged residues at the disordered regions II and IV, in addition to distribution of charged residues (Grimsley et al., 1999; Loladze et al., 1999).
The emulsifying property of WTs and the modified versions were studied at pH 7.6 and at high (µ = 0.5) and low (µ =0.08) ionic strengths. The investigation was based on two criteria: emulsifying ability (Fig. 8) and emulsion stability (Fig. 9). The emulsifying ability of the proteins was analyzed by measuring the particle size distribution and calculation of mean droplet diameter of the emulsion samples using a light scattering instrument (Fig. 8). The smaller the particle size of the emulsion droplet, the better the emulsion. At µ = 0.5, the emulsions of all A1aB1b modified versions were similar in average particle size to that of the A1aB1b WT (3.9 µm) except those of A1aB1b'(2.4 µm) and A1aB1bNeg (6.8 µm), which were the best and the poorest among all A1aB1b modified versions at the high ionic strength. At µ = 0.08, the emulsions of A1aB1b', A1aB1b and A1aB1bPos were much smaller in average particle size than that of A1aB1b WT. However, the average particle sizes of the emulsions of A1aB1bNeg and A1aB1bA4IV were close to that of the A1aB1b WT. On the other hand, the emulsifying ability of A2B1a'and A2B1a were much better than A2B1a WT at both ionic strengths. These results indicate that addition of and 'extension regions and 20 positively charged amino acids to the C-terminal of proglycinin is a good method for improving the emulsifying ability of proglycinin, but addition of A5A4B3 hypervariable region and 20 negatively charged amino acids residues does not give similar beneficial effects on said property.
|
|
The stability of the emulsion of the WTs and modified versions was analyzed by sealing and keeping the test tube containing the emulsions at room temperature without agitation and was visually observed after 1/24, 20/24, 2, 5, 7 and 14 days (Fig. 9). Emulsion stabilities of all modified versions were better than those of the WTs at both ionic strengths after 20 h and were noted to follow the order: A1aB1b
A2B1a > A1aB1b' A2B1a' >> A1aB1bPos > A1aB1bA4IV A1aB1bNeg > A2B1a > A1aB1b (Fig. 9A). Among the modified versions in this study, those resulting from the introduction of the ' and extension regions to the C-terminal of proglycinins exhibited very good emulsion stabilities. For further investigation, the A1aB1b and A1aB1b' emulsions were kept at room temperature and the time dependence of their emulsion stabilities was recorded. It was found that the emulsions of the modified versions at µ = 0.08 were better than that at µ = 0.5. It is known that glycinin structure is more unstable at µ = 0.08 than at µ = 0.5 when exposed to heat denaturation (Utsumi et al., 1987) and that modified proglycinin C12G/C88S lacking two disulfide bonds is much susceptible to proteinase attack at lower ionic strength (Utsumi et al., 1993). Therefore, the higher stability of the emulsions at µ = 0.08 than at µ = 0.5 is probably due to that destabilization at low ionic strength of the modified proteins enhanced more absorption in the unfolding process and remained unfold at the interface (Damodaran, 1997). At µ = 0.5, the phase separation of these emulsions was slightly observed after 2 days and clearly observed after 5 days (Fig. 9B). However, the lower phases did not become transparent completely even after 28 days in contrast to that observed with A1aB1b WT (data not shown). At µ = 0.08, the phase separation of A1aB1b' and A1aB1b emulsions became visible only after 5 and 14 days, respectively. These indicate that the ' and especially extension regions are good polypeptides for the improvement of emulsifying ability and emulsion stability of soybean proglycinins.
A1aB1b and A2B1a exhibited similar emulsion stabilities and their stabilities were better than those of A1aB1b', A2B1a' and A1aB1bA4IV. Tandang et al. (2005) found that changing the position of the ' extension region from the N- to the C-terminus of the ' subunit results in better emulsion stability of the protein, indicating that the extension region at the C-terminal side is more effective than that at the N-terminal for the emulsifying properties. The C-terminus (residue no. 141) of the extension region is connected with the N-terminus (residue no. 142) of the core region in the original protein. After conversion of the position, the C-terminal side of the extension region becomes exposed and flexible. Insertion of the ' extension region at the junction of the acidic and basic polypeptides of proglycinin A1aB1b did not improve the emulsion stability at all (Tandang et al., 2005). These suggest that the flexibility of the C-terminal side of the extension region is important. At the oil/water interface, the probability of a collision leading to adsorption should be a function of hydrophobic/hydrophilic ratio of the protein surface (Damodaran, 1997). The exposed regions at the protein C-terminus which is not important for proper folding of the protein therefore absorbed at the oil/water interface and remained unfold.
For a critical analysis of the introduction of polypeptides to the C-termini of proglycinins, the hydrophobicity profiles of the introduced polypeptides were analyzed by DNAsis program (Hitachi Software Engineering Co., Ltd, Japan) (Fig. 1). The first 20–30 amino acids of the N-terminus of the extension region were less hydrophilic than the last 20–30 amino acids of the C-terminus. It is noted that the hydrophilicity of 20–30 amino acids of the C-terminus of the polypeptides in this study was > ' > A4IV and the emulsion stability was also A1aB1b > A1aB1b' > A1aB1bA4IV. These suggest the assumption that the hydrophilicity of the last 20–30 amino acids of the polypeptide is important for the emulsion stability. Although the A1aB1bPos and A1aB1bNeg had higher hydrophilic composition than those of the others, they provided emulsion stability lesser than did A1aB1b'and A2B1a'. Therefore, it might be not only the distribution of the hydrophilicity of the 20–30 amino acid residues at the end of C-terminus region but also the length of the introduced polypeptide and the other factors that are important for emulsion stability.
All the introduced C-terminal regions of modified proteins were rich in hydrophilic residues. The percentage of hydrophilic residues in the C-terminal regions were 79.4, 74.4, 80.6, 100 and 100% for A1aB1b', A1aB1b, A1aB1bA4IV, A1aB1bNeg and A1aB1bPos, respectively. Among all mutant C-terminal regions, the extension region has two slightly hydrophobic areas corresponding to the amino acid residues number 39–49 and 91–97 of A1aB1b C-terminal region (Fig. 1). Residues 38–46 of the ' extension region and 21–26 of A4IV region were also hydrophobic areas but their hydrophobicities were poorer than those of the extension region (residues 39–49 and 91–97). It is possible that the hydrophobic areas also contribute to the outstanding emulsion stability of A1aB1b.
The information obtained in this study would be very useful for protein design studies, and can be an effective strategy to improve the emulsifying property of seed storage proteins.
|
Footnotes |
|---|
The originally published version of this paper was incorrect. Due to the resizing of figure 7, some data was lost in the original version.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
This work was supported in part by a grant to S. U. from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionAcknowledgementsReferences |
|---|
Adachi M., Takenaka Y., Gidamis A.B., Mikami B., Utsumi S. J. Mol. Biol (2001) 305:291–305.[CrossRef][Web of Science][Medline]
Adachi M., Kanamori J., Masuda T., Yagasaki K., Kitamura K., Mikami B., Utsumi S. Proc. Natl Acad. Sci. USA (2003) a 100:7395–7400.
Adachi M., Okuda E., Kaneda U., Hashimoto A., Shutov A.D., Becker C., Müntz K., Utsumi S. J. Agric. Food Chem. (2003) b 51:4633–4639.[CrossRef][Web of Science][Medline]
Bilgi B., Çelik S. Eur. Food Res. Technol. (2004) 218:437–441.[CrossRef]
Damodaran S. Food Proteins Their Applications: Protein-Stabilized Forms Emulsions.—Damodaran S., Paraf A., eds. (1997) New York: Marcel Dekker. 57–110.
Dickinson E. An Introduction of Food Colloids. (1992) Oxford, UK: Oxford Science.
Dickinson E., Murray B.S., Stainsby G. J. Chem. Soc. Faraday Trans. I (1988) 84:871–883.[CrossRef]
Dickinson C.D., Hussein E.H.A., Nielsen N.C. Plant Cell (1989) 1:459–469.
Food and Drug Administration. Federal Register. (1999) 64(206):57699–57733.
Graham D.E., Phillips M.C. Theory Practice of Emulsion Technology: The Conformation of Proteins at Interfaces Their Role in Stabilizing Emulsions.—Smith A.L., ed. (1976) London: Academic Press. 75–93.
Green A.A., Hughes W.L. Methods in Enzymology.—Colowick S.P., Kaplan N.O., eds. (1955) Vol. 1. New York: Academic Press. 67.[CrossRef][Web of Science]
Grimsley G.R., Shaw K.L., Fee L.R., Alston R.W., Huyghues-Despointes B.M.P., Thurlkill R.L., Scholtz J.M., Pace C.N. Protein Sci. (1999) 8:1843–1849.[Web of Science][Medline]
Hayakawa S., Nakai S. J. Food Sci (1985) 50:486–491.[Web of Science]
Kato A., Nakai S. Biochim. Biophys. Acta (1980) 624:13–20.[Medline]
Katsube T., Kurisaka N., Ogawa M., Maruyama N., Ohtsuka R., Utsumi S., Takaiwa F. Plant Physiol. (1999) 120:1063–1073.
Kim S.H., Kinsella J.E. J. Food Sci (1987) 52:128–131.[CrossRef][Web of Science]
Kim C.S., Kamiya S., Kanamori J., Utsumi S., Kito M. Agri. Biol. Chem. (1990) 54:1543–1550.[Web of Science]
Kinsella J.E. J. Am. Oil Chem. Soc (1979) 56:242–258.[CrossRef][Web of Science]
Kinsella J.E., Damodaran S., German B. New Protein Foods: Seed Strorage Proteins.—Altschul A.M., Wilcke H.L., eds. (1985) Academic Press Orlando, FL. 107–179.
Kolar C.W., Cho I.C., Watrous W.L. J. Am. Oil Chem. Soc (1979) 56:389–391.[CrossRef][Web of Science]
Laemmli U.K. Nature (1970) 227:680–685.[CrossRef][Medline]
Lawrence M.C., Izard T., Beuchat M., Blagrove R.J., Colman P.M. J. Mol. Biol. (1994) 238:748–776.[CrossRef][Web of Science][Medline]
Lee K.K., Fitch C.A., García-Morenoe B. Protein Sci. (2002) 11:1004–1016.[CrossRef][Web of Science][Medline]
Lindman S., Xue W-F., Szczepankiewicz O., Bauer M.C., Nilsson H., Linse S. Biophys. J (2006) 90:2911–2921.[CrossRef][Web of Science][Medline]
Liu M., Lee D.S., Damodaran S. J. Agric. Food Chem. (1999) 47:4970–4975.[CrossRef][Web of Science][Medline]
Loladze V.V., Ibarra-Molero B., Sancher-Ruiz J.M., Makhatadze G.I. Biochemistry (1999) 38:16419–16423.[CrossRef][Medline]
Maruyama N., Katsube T., Wada Y., Oh M.H., Barba de la Rosa A.P., Okuda E., Nakagawa S., Utsumi S. Eur. J. Biochem (1998) 258:854–862.[Web of Science][Medline]
Maruyama N., Mohamad Ramlan M.S., Takahashi K., Yagasaki K., Goto H., Hontani N., Nakagawa S., Utsumi S. J. Agric. Food Chem (2002) 50:4323–4326.[CrossRef][Web of Science][Medline]
Maruyama N., Prak K., Motoyama S., Choi S.K., Yagasaki K., Ishimoto M., Utsumi S. J. Agric. Food Chem. (2004) 52:8197–8201.[CrossRef][Web of Science][Medline]
Maruyama N., Sato R., Wada Y., Matsumura Y., Goto H., Okuda E., Nakagawa S., Utsumi S. J. Agric. Food Chem. (1999) 47:5278–5284.[CrossRef][Web of Science][Medline]
Nagano T., Mori H., Nishinari K. J. Agric. Food Chem. (1994) 42:1415–1419.[CrossRef][Web of Science]
Nakai S. J. Agric. Food Chem. (1983) 31:676–683.[CrossRef][Web of Science]
Nakai S., Li-Chan E. Hydrophobic Interactions in Food Systems: Hydrophobicity-Functionality Relationship of Food Proteins (1988) Boca Raton, FL: CRC Press. 23–41.
Nielsen N.C., Dickinson C.D., Cho T.J., Thanh V.H., Scallon B.J., Fischer R.L., Sims T.L., Drews G.N., Goldberg R.B. Plant Cell (1989) 1:313–328.
Palazolo G.G., Mitidieri F.E., Wagner J.R. Food Sci. Technol. Int. (2003) 9:409–411.[CrossRef]
Peng I.C., Quass D.W., Dayton W.R., Allen C.E. Cereal Chem. (1984) 61:480–490.
Phillips M.C. Food Technol (1981) 35:50–59.
Prak K., Nakatani K., Katsube T.K., Adachi M., Maruyama N., Utsumi S. J. Agric. Food Chem (2005) 53:3650–3657.[CrossRef][Web of Science][Medline]
Roudsari M., Nakamura A., Smith A., Corredig M. J. Agric. Food Chem (2006) 54:1434–1441.[CrossRef][Web of Science][Medline]
Staswick P.E., Hermodson M.A., Nielsen N.C. J. Biol. Chem. (1984) 259:13424–13430.
Takano K., Scholtz J.M., Sacchettini J.C., Pace C.N. J. Biol. Chem. (2003) 278:31790–31795.
Tandang M.R.G., Atsuta N., Maruyama N., Adachi M., Utsumi S. J. Agric. Food Chem (2005) 53:8736–8744.[CrossRef][Web of Science][Medline]
Utsumi S. Advances in Food and Nutrition Research: Plant Food Protein Engineering.—Kinsella J.E., ed. (1992) a. SanDiego, CA: Academic Press. 89–208.
Utsumi S. Adv. Food Nutr. Res. (1992) b 36:89–208.[Medline]
Utsumi S., Kinsella J.E. J. Food Sci. (1985) 50:1278–1282.[CrossRef][Web of Science]
Utsumi S., Nakamura T., Harada K., Mori T. Agric. Biol. Chem (1987) 51:2139–2144.[Web of Science]
Utsumi S., Gidamis A.B., Kanamori J., Kang I.J., Kito M. J. Agric. Food Chem (1993) 41:687–691.[CrossRef][Web of Science]
Utsumi S., Matsumura Y., Mori T. Food Proteins and Their Applications: Structure-Function Relationships of Soy Proteins.—Damodaran S., Paraf A., eds. (1997) New York NY: Marcel Dekker. 257–291.
Wagner J.R., Guéguen J. J. Agric. Food Chem. (1995) 43:1993–2000.[CrossRef][Web of Science]
Wagner J.R., Guéguen J. J. Agric. Food Chem (1999) 47:2181–2187.[CrossRef][Web of Science][Medline]
Received September 21, 2006; revised June 17, 2007; accepted June 21, 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||