936 WANG ET AL.: JOURNAl. OF AOAC INTERNATIONAL VOL. 91, No. 4, 2008

SPECIAL GUEST EDITOR SECTION

Analysis of Protein-Derived Peptides and the Effect of Cultivar, Environmental Conditions, and Processing on Lunasin Concentration in Soybean and Soy Products

WENYI WANG, VERMONT P. DIA, MIGUEL VASCONF:Z, and ELVIRA GONZALEZ 1W MEJIA1 University of Illinois, Department of Food Science and Human Nutrition, 228 ERML, 1201 W. Gregory Dr, Urbana, IL 61801 RANDALL L NELSON U.S. Department of Agriculture, Agricultural Research Service, Soybean/Maize Gerrnplasm, Pathology, and Genetics Research Unit, Department of Crop Sciences, 1101 W. Peabody Dr. University of Illinois, Urbana, IL 61801

Soybean, an important source of food proteins, has quantified by using different techniques and received increasing interest from the public conditions. In addition, lunasin concentration in because of its reported health benefits. These health soybean depends mainly on cultivar and to some benefits are attributed to its components, including extent on environmental factors, particularly isoflavones, saponins, proteins, and peptides. temperature. Lunasin concentration in soy products Lunasin, Bowman-Birk inhibitor, , and was also affected by processing conditions. 3-conglycinin are some of the biologically active peptides and proteins found in soybean. This article provides a comprehensive review on the recently ood proteins have long been recognized as a source of used techniques in the analysis and essential amino acids required for the maintenance of characterization of food bioactive peptides, with F life. They also serve certain functions as ingredients in emphasis on soybean peptides. The methods used foods and food preparations. For the past 2 to 3 decades, to isolate and purify lunasin from defatted soybean attention has been given to other aspects of food proteins, flour were ion-exchange chromatography, such as their biologically active peptides derived either from ultrafiltration, and gel filtration chromatography. The enzymatically digested proteins or by protein fermentation. identity of lunasin was established by sodium In vitro and in vivo studies have shown that these bioactive dodecyl sulfate-polyacrylam ide gel electrophoresis, peptides can perform certain beneficial biological functions Western blot, matrix-assisted laser desorption in the body, such as antihypertensive and antioxidant ionization—time of flight, and liquid chromatography. activities, prevention, hypocholesterolemic, The results on the effect of soybean cultivar and antiobesity, and immunomodulatory actions (I). It is environmental factors on lunasin concentration are therefore important to have precise and accurate methods for also reported. The highest lunasin concentration, quantification and characterization of these bioactive 11.7 ± 0.3 mg/g flour, was found in Loda soybean peptides present in foods. The most studied food sources of cultivar grown at 23°C; the lowest concentration, 5.4 bioactive peptides are milk and soybean. Lunasin is a novel ± 0.4 mg/g flour, was found in Imari soybean cultivar bioactive peptide found in soybean. It is composed of 43 grown at 28°C. Lunasin concentration was affected amino acids with 9 aspartic acid residues on its carboxyl end. by cultivar—temperature, cultivar—soil moisture, and Its bioactive properties are attributed to its capability to cultivar—temperature—soil moisture interactions. The arrest cell division in cancer cells and to inhibit core histone variation on lunasin concentration suggests that its acetylation in mammalian cells (2). content can be improved by breeding, and by This paper reviews the methods used in the analysis of optimization of growing conditions. In summary, bioactive peptides in soybean and the effect of processing. bioactive peptides can be accurately identified and cultivar, and environmental factors on lunasin concentration in soybean and soy products. Guest edited as a special report on Accurate Methodology for Amino Acids and Bioactive Peptides in Functional Foods and Dietary Supplements for Assessing Protein Adequacy and health Effects" by (i Sarwar Gilani Techniques Used in the Analysis of Soybean and Paul J. Moughan. Peptides Mention of a trademark, proprietary product, or vendor does not constitute a guarantee or warranty of the product by the USDA or the University of Illinois and does not imply its approval to the exclusion of Protein is the most abundant component in soybean. On other products or vendors that may also be suitable. average, soybean contains 40% protein conformed by a Author to whom correspondence should be addressed; e-mail: complex mixture of different protein types (3). The major edemejiaiuiuc.edu components of so y proteins are seed storage proteins known as

WANG FT AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 91, No. 4, 2008 937

C-)a,. a a ci) N- - C C,) (N (0)N 0) 0) (. ci) (U a, — a. II 0 cc ci) (U (U ci)—..------ci) (U (U (U CU (U ci) LU ow (0 -a CU . -ccS 0) a, CU ) CU (U CU CU CU CU CU — • ci) CU a © cC) (U00 CU a. a, a, a, Cl) a, a, a, CU< a CU C.) 0) E > a- a. a. a. a. a. a. E - cca, Da aJ a, a, 0 .2 .2.2 .2 .2 .2.? CU 0 : °a ci, (U V o - t t t3t (U 0 CO CU (U CU CU CU CU CU >.9 ci a c) a CU .Lc) — D 0) 0) 0) 0) 0) 0) 0) C 3 a . . . . . 6 E ci) - 0 +1 a, ci) a a CU o . :aa.5- .0 a - ci) > — - .2 : = 0 -C 0OD oci) 0 0 0 -D 70 D D - A .2 a, a, a, (D a, a, a, a, LX) Q C)) a, Cl) (U ci) o ci) CO (U Cl) ._(U :1 < 0 D

ci) -D(D 0ci) -cc a_ a. 0 a) H 0_ a. a- a. .9 .9 .9 .9 a, 0 E .0 .0 CC)

a. • 2 ci) C ci a, UJ - - LU w LU LU W LU 2.5. 0 LU LU LU LU m 0 0 0 0 0 0 ci, 0 0 a, 0 0 0 0 < _5 < < - CL > Cl. II I I 90 a2 cr a a — Ca C aa. ci) ci) C 2 2 .2 O a, (1) 0 — CU a, CU— ..-. CUCU .0 >S a. 0) ci) 0 0 C) 0) 0) 0 ci) 0) > 0 a a, (U a. a. -.,- -- .0 .0_ 0 :6 :6 2 a^- a a a-i, >1 - 0 CD0 >O (U Oa, 2 am) M CL 0 U) U)Cl) C/) C/) a, 0) E Oo E E -J a a E< 0 a u E LL LUC E OH I— c0(N C 64 0 E E U - LO LO E C Ea N- - E C..) E 0 > a. a E DE > H E 0 0 d o a CU o + 04 1 E 0 0) + E(NE c CU cD E EZELU Z O- cc z a 2 0 CD LO 0 a 0 E 3E a 0 0 <--a ... LU EEE (Na CU 0)Q LU - C (00 LU U.) E E ,CciC) ci) ©© ,_ H Z H —2 , - OdZH + 0a ci a, (N (U LO 0 CU Z 0) Cc 0 CU N- C© -< 0 .0 0 —20 Z .2-2 .5. 3 a. 0 ci) + Z- 0- 0 Cl a LO (_) O -aCL - .© I -0 C) Z N. .— 00 —C C qzi ci) U 03 + < CD (N I LU 0 110 (0:i: (N Z 0 ) 0 H ci, LL .2) CU 0 0< QZ .6 -a CU CY5 (:)CD ci) CD 8-1 0 0 z..— 0 .011 CU C C) a 0 5. 03 ci) D a 0 Oa, . E ci) (U)< I a,a (U (N a co ? c-, 75 < - LO E 2 L 0 _ -o (N <0 c -E -2 CL 0 8. << E.ci (0.9 a. a,.a-Cl) 8 o o 2E5. C a, x-i &)0 0 -ö CL X0 CU lo Ln u 2© aD° LU = 0t = - 9. o - a 2 a a. a,© .0a, (N 0 E 0 aEE w5 a, a,o 0

1 0 0 -J 0 11 0> 03 0 I 0(Ua .0 0 LL 0 0 wO CU a, 0 u LL < 0 0 —CC)0 I— w o o 938 WANG ETAL.: JOIJRNALOF AOAC INTERNATIONAL VOL. 91, No. 4,2008

a) -conglycinin and glycinin, which account for 50-70% of total C.) C 5). 3-Conglycinin is a trimer with a molecular a) Q N (N (0 10 U) Lt) It) N. (N C)) C) N- seed proteins (4, a) weight (MW) of! 50-200 kilodaltons (kDa). It is composed of 3 subunits: a, a, and 3(6). The a and a subunits consist of core regions with high degree of homology (86.8%) and extension a) regions (a, 125 residues; a, 141 residues) exhibiting lower CO C homologies (57.3%), whereas the f3 subunit consists only of a 0 core region that has homology with the a and a core regions 0a) (75.5 and 71.4%, respectively; 7). Glycinin is a hexamer with MW of 320-375 kDa and with 5 major subunits (G I, G2, G3, :S! E .0 -J 0 0) C G4, and G5). Each subunit consists of an acidic chain (about E E C > 40 kDa) and a basic chain (about 20 kDa), joined by disulfide > bonds. G 1, G2, and G3 can be grouped as they share 90% 0 > CO .2 C C a) H sequence homologies. Similarily, G4 and G5 share 90% sequence 0 (0E C Q Lt Ii co CC) 0 LO - homologies. However, sequence homologies between these co co 2 groups (G I, G2, G3, and G4, G5) are only 50%. In addition, 0) (> " N-

co there are many enzymes (such as lipoxygenase, chalcone o synthase, catalase, urease) in soybean, but only a relatively small 2 number of them exceed 1% of total seed protein. Upon ingestion, soy proteins are digested to peptides by gastrointestinal enzymes Cv)Cq and have been found to be bioactive (I). It) Soybean has been extensively studied for the presence of

Cl) 0 biologically active peptides. Table I summarizes some of the 0 LL IL analytical techniques used recently in the analysis of bioactive C peptides from soybean and soy products. Table 2 provides >-JO) E> 0 examples of the biological activities of some of the peptides C/) I recently isolated from soybean and soy products. Combined 0 -D with mass spectrometry (MS) techniques, gel-based 0) a) separations provide valuable information on protein/peptide 0 0 0 0 composition. Natarajan et al. (8) compared the presence of D >. Kunitz tlypsin inhibitor (KTI) from wild and cultivated 00 N = () - 0 soybean using 2-dimensional polyacrylamide gel o o a) 0 0 CO C 0. 0 (0 0 electrophoresis (2D/PAGE) and matrix-assisted laser CC B CCOCO C) (0 (I) desorption ionization/time of flight/mass spectrometry Oa) a) 0 0 0. 0. 2 a) .0 0 a) CO (I) -D 0. 0) (5 a) >. 0 (MALDI/TOF/MS). The results showed that the number and C 00 CO 0 0 U) (0 a) U) (0 a) CO C intensity of the protein spots between wild and cultivated D -D — CO CO Cd) o a) 0) 0 C CL C) CC 0. CO genotypes varied even though overall distribution patterns of C OZ I- 0) a) CO > EE a) O) a) KTI protein spots appeared similar. 0 .0 10 (0 (I) G) a) >, .0 CO u- u- 0 C Due to the complexity of peptides present in protein digests, E a) 0 or fermented foods, various chromatographic techniques are I- 0 0 .4— 0 0. -C 0 often coupled together to achieve desired separations. Lo and C) 4— 0 0 CO Li-Chan (9) used anion-exchange chromatography, 0 ultrafiltration, reversed-phase high-performance liquid An C C chromatography (RP-HPLC), and immobilized metal affinity 0 > C C chromatography (IMAC) to study angiotensin converting C) C C.) -C Ca) 0 C C-) enzyme (ACE) inhibitors from in vitro pepsin-pancreatin C) C 0 a) > 0 CL C) U) C digest of soy proteins. They showed that many different C) >. 0 o LI C a) U) 0- peptides with ACE inhibitory activity were produced by in 0. vitro enzyme treatment and speculated that physiological C) 0. 0 gastrointestinal digestion could also yield ACE inhibitory C) U) > peptides from . Zhang et al. (10) used gel filtration .4- CO CO E chromatography (GFC) and HPLC to purify and separate 0 a) 0 CO C C peptides with ACE inhibitory function from , a Chinese 0 a) CO .0 fermented soybean product, and showed that a peptide C..l 00 10 .0 C 0 composed of phenylalanine/isoleucine/glycine in a 1:2:5 ratio C) 0 CO CU o ) =- .0 .^ w c_ 10 E exhibited the highest ACE inhibitory activity. Multiple HPLC, CO to > C I- < < 1< MS, and amino acid analyses were used by Gibbs et al. (11) to I WANG ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 91, No. 4,2008 939

the preparation, yield, and in vitro hypocholesterolemic A activities of low MW soy protein hydrolysates. They found that the optimal parameters for desalting soy protein alcalase hydrolysates (SPAH) were the use of macroporous adsorption resin (DA210-C) at pH 4.5, with SPAH dispersion to resin ratio of 75:100, and a loading rate of I bed volume/h with 89.7% adsorption rate of peptide. This nonpolar resin adsorbed hydrophobic peptides and separated them from the salt in solution; because of its large capacity, it is very suitable in desalting peptides (16). Zhong et al. (17) also analyzed a novel hypocholesterolemic peptide from soy protein hydrolysate using fractionation by gradient ethanol elution from DA20 1-C resin. They further characterized the peptide by GFC using Sephadex G-15 and RP-HPLC. The hypocholesterolemic peptide structure was identified by B HPLC/MS. Gianazza et al. (18) compared soy protein diets 1 10 used for hypercholesterolemia studies in Europe and the Ser-Lys-Try-GIn-His-GIn-GIn-Asp-Ser-Cys- United States by using 2D electrophoresis, and the identities 20 Arg-Lys-GIn-Leu-GIn-GIy-VaI-Asn-Leu-Thr- were established by MALDI/MS. They showed that there 30 were differences in the composition of the soy products used Pro-Cys-GIu-Lys-His-I le-Met-GIu-Lys-I le- in these clinical studies; thus a comparison was difficult 39 to establish. GIn-GIy-Arg-GIy-Asp-Asp- Asp- Asp- Asp- Lee et al. (19) developed a calcium-binding mediator using 43 peptides derived from isolated soybean protein. The MW of Asp- Asp- Asp- Asp- the peptides was determined by GFC, and the results showed the possibility that soybean phosphopeptides can be potent Figure 1. Lunasin predicted structure. (A) Helix with calcium carriers; thus, they can be used to prevent poor structural homology to a conserved region of absorption of dietary calcium in animals. chromatin-binding proteins; (B) 43 amino acid peptide that contains an arginine-glycine-aspartate motif (ref. 2). Rho et al. (20) used size-exclusion chromatography-HPLC to confirm and quantify amino acid composition of peptides from black soybean. These peptides were analyzed by GFC to determine the MW distribution, and results showed that 80% of black soybean peptides had an MW <10 kDa. This study analyze ACE inhibitory peptides from soy hydrolysate and also showed the effect of these peptides on significant fermented soy foods. The results showed that endoproteases attenuation of body, liver, and adipose tissue weight (P < 0.05) with lower specificity produced more oligopeptides with as compared to casein groups in Sprague-Dawley male rats. A higher biological activities. GFC and RP-HPLC were also detailed review on the use of HPLC and capillary used by Kuba et al. (12) to isolate 2 ACE inhibitors from the electrophoresis on the analysis of soybean proteins and tofuyo extract. RP-HPLC and ion-pair chromatography peptides has been provided by Saz and Marina (2 1). coupled with MALDI/TOF were used by Mal likarjun Gouda These published articles show that a wide variety of et al. (13) to study ACE inhibitory peptides in different techniques can be used in the isolation, purification, enzymatic hydrolysates of glycinin. A potent ACE inhibitory characterization, and identification of biologically active peptide, Val-Leu-Ile-Val-Pro, was identified from the protease peptides in soybean. Choosing suitable methods depends on P hydrolysate of glycinin. GFC and RP-HPLC with the objectives of the study and the chemical nature and electrospray ionization with tandem MS (ESJ/MS/MS) were stability of the target compound. used by Kodera and Nio (14) to study soybean protein hydrolysates produced by protease D3. Kuba et al. (15) Lunasin: A Novel Bioactive Peptide in Soybean hydrolyzed glycinin and 3-conglycinin from soybean using an acid proteinase from Monascus pupureus to produce ACE Lunasin is a soybean peptide composed of 43 amino acid inhibitory peptides. They used DIAION HP2ISS resin, residues with an MW of 5.5 kDa. It contains 9 aspartic acid ultrafiltration, and RP-HPLC to purify active fractions from residues on its carboxyl end, a cell adhesion motif composed glycinin and 3-conglycinin. They were able to isolate and of arginine-glycine-aspartic acid residues, and a predicted analyze 4 ACE inhibitors, 2 from each soybean seed helix with structural homology to a conserved region of storage protein. chromatin-binding proteins (22). Earlier studies on animals Due to the relatively small size of bioactive peptides, showed that lunasin is not fully digested in the gastrointestinal protein desalting procedures, such as dialysis, may not be system but is absorbed intact, reaching target tissues (23). The applicable for peptide concentration. Zhong et al. (16) studied biological activity of lunasin depends on its concentration in 940 WANG ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 91, No. 4, 2008

100 2.5 350 2 330

250 1.5 A 1200 150 100 0.5 EQ 0 0

3 2

Frc1ion iiunei

eiJu

400 rALJ B

200

0

8 12 18

Minutes

9 B 7 -J

C C 4

—J 2

0

N . 5 b Fraction number

Figure 2. Chromatographic profiles for the isolation, identification, and purification of lunasin from defatted soy flour. (A) Lunasin profile of ion-exchange fractions of soy extract. The extract was eluted using Tris—HCI as solvent Aand Tris—HCI plus 2 M NaCl with step gradient from 0, 5, 10, 20, and 100% solvent B. (B) HPLC profiles of synthetic lunasin (B, dash line) and lunasin purified by IEC (A and C, solid line). Profile A corresponds to fraction 104 and profile C to fraction 88. The buffers used were 5% acetonitrile + 0.08% trifluoroacetic acid (buffer A) and 95% acetonitrile + 0.1% trifluoroacetic acid (buffer B) using linear gradient elution, and were detected at 215 nm. (C) Lunasin profile of gel filtration fractions after one void volume using active pooled fraction from IEC. The pooled fraction was eluted using Tris—HCI plus 0.15 M NaCl.

the product, which in turn is affected by cultivar, 14 kDa are not in association with each other. More research is environmental factors, and processing conditions. Figure 1 needed regarding the chemical structure of this peptide. presents the predicted secondary structure of lunasin, its 43 amino acids, and the motif (2). Identification and Quantification of Lunasin by Chromatography, Western Blot, and Enzyme-Linked Studying the structure of lunasin using native PAGE, Immunosorbent Assay nonreducing and reducing denaturing PAGE revealed that In our laboratory, a defatted soybean protein extract was lunasin is a single polypeptide chain. Under those 3 different prepared by extraction at pH 8.2, loaded in an ion-exchange electrophoretic conditions, the same gel pattern was observed, chromatographic (IEC) column and run with Tris—HCI and implying that the 3 active bands corresponding to 5, 7, and Tris—HCI plus 2 M NaCI buffer. Fractions were collected A WANG El AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 91, No. 4, 2008 941

SDS-PAGE and Western blot results for the identification of 3 M 1 2 lunasin in ion-exchange and gel filtration fractions. Three active bands can be seen from ion-exchange fractions 88 and 104, and gel filtration fractions 11, 14, and 16 corresponding to 5, 7, and 14 kDa peptides. Figure 4 shows the MALDI-TOF profile of synthetic lunasin (A), lunasin purified by IEC (B),

• 78kDa14 0. and lunasin purified by GFC (C). Figure 4A and B show that 5.1. 5D 3 major peaks were found which corresponded to 7.9, and 14.1 kDa peptides. MALDI-TOF profile of active fraction from GFC showed very few peaks as compared to

= 14 kDa ion-exchange fractions, suggesting further purification of ALt - - - - the sample. _ L4 T ------The detailed procedure to quantify lunasin by ELISA in our laboratory has been described previously (24). Briefly, IOU p1 soy flour extract was plated on a 96-well plate and Figure 3. SOS-PAGE (A) and Western blot (B) of gel filtration fractions for identification of lunasin. Lanes stored overnight at 4°C; the plate was then washed and 1-5 are examples from different GF fractions. Lane M is blocked with 5% bovine serum albumin for I h. After molecular weight standard. blocking and washing, 50 p1 lunasin monoclonal antibody (1:4000) was plated and incubated for 1 h. After washing, 50 .jL antimouse immunoglobulin G (1:7000) was plated and incubated for I h. The color was developed by adding 100 p1 using step gradient elution at 10 mL/min. The lunasin profile p-nitrophenyl phosphate (pNPP) to each well. The color the ion-exchange fractions is shown in Figure 2A. From of produced after 20 min was read using an ELISA plate reader at this collection, fractions 88 and 104 were taken and analyzed 405 rim. The reaction was stopped by adding 100 .tL 3 N by HPLC due to the higher concentration of lunasin per total NaOH after 25 min and read again at 35 mm. Lunasin protein, as determined by enzyme-linked immunosorbent concentration was quantified using a standard curve from assay (ELISA) and Bradford protein assays. A 100 1.iL volume different concentrations of synthetic lunasin ranging from 8 to of filtered ion-exchange fraction was injected into an HPLC 36 ng/mL (y = 0.0291x + 0. 1533, R 2 = 0.95). system equipped with a Vydac C4 column (25 cm x 4.6 mm, 10 I.tm) after system equilibration and stabilization, using Effect of Soybean Cu/tivar and Environment on diode array detection (215 nm). A mobile phase [5% Lunasin Concentration acetonitrile and 0.08% trifluoroacetic acid (TFA). buffer A; and 95% acetonitrile and 0.1% TFA, buffer B] in linear In order to understand the effect of cultivar, temperature, gradient for 30 min at 1 mL/rnin was used. Lunasin was and soil moisture on lunasin concentration of soybean, identified by retention time using synthetic lunasin (profile B) 5 cultivars (2 French, Imari and Queen, and 3 U.S., Dwight, as shown in Figure 2B. The chromatogram shows that fraction Jack, and Loda) were studied (25). They were selected based 104 (profile A) contains fewer contaminating proteins than on similar maturity from a greenhouse experiment by fraction 88 (profile Q. Active fractions from ion exchange Lozovaya et al. (26) and Vasconez et al. (27). All entries were were run through GFC to further purify lunasin. Gel filtration grown in the Plant Sciences Laboratory greenhouse on the was performed by introducing 15 mL pooled active campus of the University of Illinois; one plant per 30 cm ion-exchange fractions in Superdex Prep Grade 75 column diameter plastic pot under intermediate night/daytime using 20 mM Tris—HCI plus 0A5 M NaCl buffer as mobile temperatures of 18/28°C (23°C mean) with high soil moisture phase at a flow rate of 4 mL/min. Fractions were collected and under a 14.5 h photoperiod. When the plants reached the every minute after the void volume. Lunasin concentration of R6 growth stage, they were moved into one of 3 different the gel filtration fractions is shown in Figure 2C. From this temperature regimens until maturity: low. 13/23°C (18C figure, one active peak corresponding to fraction 14 mean); intermediate, 18/28°C (23°C mean); and high, was detected. 23/33°C (28°C mean). In each temperature treatment, Lunasin active fractions from IEC and GFC were analyzed one-half of the plants were grown in high soil moisture by sodium dodecyl sulfate (SDS)-PAGE with high-density (approximately 70% of soil-holding capacity) and one-half in gels and trans-blotted to polyvinylidene difluoride (PVDF) low soil moisture (approximately 30% of the high treatment). membranes for Western blot analysis. The PVDF membrane, Out of a total of 5 replicates of cultivar/teniperature/soil with the transferred proteins, was blocked and incubated for moisture, a pooled sample was obtained; thus, a total of 16 h at 4°C with lunasin monoclonal antibody 30 samples was analyzed for lunasin concentration. The (1:1000 dilution) provided by Ben 0. de Lumen, University of results showed that cultivar and temperature, but not soil California, Berkeley. After washing, the membrane was moisture, significantly affected lunasin concentration in soy. incubated with an antimouse secondary antibody Significant interactions of cultivar-temperature, cultivar-soil (1:10000 dilution) for 3 h and was examined using moisture, and cultivar-temperature-soil moisture were also chemiluminescent reagent. Figure 3A and B show the detected (Table 3). 942 WANG ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 91, No. 4, 2008

A

B

Figure 4. MALDI-TOF profile of synthetic lunasin (A), lunasin purified by IEC (B), and lunasin purified by GFC (C) using the following parameters: linear mode of operation, positive polarity, and 3000-20 000 Da scanning range. Defatted soybean extract was loaded in an ion-exchange column and run with Tris—HCI and Tris—HCI plus 2 M NaCl buffer. Fractions were collected corresponding to 5-20% Tris—HCI plus 2 M NaCl using step gradient elution. Active fractions from IEC were pooled and passed through a gel filtration chromatographic system using Tris—HCI plus 0.15 M NaCl as mobile phase.

Soybean genotype had a significant effect on lunasin concentration of lunasin in soybean accessions ranged from concentration. Across all temperatures and soil moisture 1.0 to 13.3 mg/g flour (24). These genetic differences of traits, the concentration of lunasin in the 5 cultivars ranged lunasin should make it possible to modify lunasin from 7.5 to 10.4 nig/g flour. The 3 U.S. cultivars. Loda, Jack, concentration in high-yielding cultivars. Temperature also had and Dwight, had a higher concentration of lunasin than did the a significant effect on lunasin concentration (Figure 513). High 2 French cultivars, Queen and Imari (Figure 5A). Loda had a temperature led to significantly lower concentration of lunasin 38% higher ILmasin concentration than Imari, which (8.08 ± 1.62 mg!g flour) than did intermediate (9.15 ± demonstrated genetic differences in lunasin concentration. 1.64 mg/g flour) and low temperatures (8.83 ± 1.38 rng/g Analysis of lunasin concentration of 144 selected, diverse flour). This trend was not predictable by either the agronomic soybean accessions from the U.S. Department of Agriculture data (maturity date, plant height, seed yield. 100-seed weight), (USDA) Soybean Germplasm Collection showed that the or the isoflavone concentration (26). Lunasin concentration of

wANT; FT AL J0uItNALOF AOAC INTERNATIONAL VOL.91, No. 4,2008 943

Table 3. Analysis of variance of lunasin concentration of 5 soybean cultivars grown under controlled temperature and soil moisture conditions Source DFa Type I SSb Mean square F-value Probability> F

Cultivar 4 71.29 17.82 29.48 <.0001° 0.0004c Temperature 2 12.22 6.11 10.11

Soil moisture 1.25 1.25 2.07 0.1603 0.0014C Cultivar-temperature 8 21.06 2.63 4.36

Cultivar-moisture 4 20.48 5.12 8.47 0.0001°

Temperature-moisture 2 2.36 1.18 1.95 0.1594

Cultivar-temperature-moisture 8 13.35 1.67 2.76 0.0206°

a DF = Degrees of freedom. Type I SS = Sum of squares. C Significant at the 0.01 probability level.

soybean grown under interniediate temperature had the values are similar to our findings, which ranged from 10.72 to highest value, although it was not significantly different from 23.36 mg lunasinig protein (27). that grown at low temperature. A 141N) increase in lunasin In summary. we have demonstrated the effect of genetics concentration was observed when soybean was grown at and environmental conditions on lunasin concentration. The intermediate temperature in comparison to high temperature. variation in lunasin concentration of 5 soybean cultivars Soil moisture did not show a significant effect on lunasin grown in different environmental conditions indicates that the concentration; however, high soil moisture delayed maturity concentration of this important bioactive component in and increased isoflavone concentration, while the plant height soybean can be manipulated by using genetics and different and 100-seed weight remained the same (26). Considering the growing conditions. relatively large moisture difference between the 2 groups, the Effect of Processing on Lunasin Concentration results indicated that lunasin concentration is not very sensitive to soil moisture conditions. The effect of processing on lunasin concentration in A significant interaction between cultivar and temperature soybean products has been demonstrated by de Melia et was observed for lunasin concentration (Table 4). For Jack, al. (24) and Jeong Ct al. (29). de Mejia et al. (24) quantified Loda, and Queen. lunasin concentration did not change with temperature (P > 0.05); however, high temperature led to significantly lower (P< 0.05) lunasin concentration for Dwight and Imari. 14 Although soil moisture did not show a significant effect on o12- A Es lunasin concentration, significant cultivar-soil moisture 0 o 5- interactions were observed (Table 5). For Imari, high soil S -J=4 moisture led to significant higher concentration of lunasin 2 5 0 (P < 0.05) than did dry conditions, while the trend was Loda Jack Dwght Queen inri reversed for Jack. The temperature-soil moisture interaction Cultivar was not significant. Temperature and soil moisture combinations did not significantly affect lunasin concentration for all cultivars. However, for individual 12 a b cultivars, significant difference was observed. This agreed 110 x B with the significant interaction observed for cultivar 6- I I S I r temperature x soil moisture (Table 6). For each cultivar, -J I I C, I lunasin concentration can vary from 35 to 75% when the most E 2 favorable temperature and moisture conditions are compared 0- _Intermediate r1 Low LiHigh with the least favorable. Highest percentage difference was Temperature observed in Imari (75%), and decreased in the order of Queen (52%). Jack (45%), Dwight (38%), and Loda (35%). Recently, Jeong et al. (28) found that lunasin concentration Figure 5. Lunasin concentration (mg/g flour) of soy of 8 Korean soybean cultivars ranged from 4.40 ± 0.32 to flour as affected by (A) cultivar, and (B) temperature. 70.49 ± 1.38 mg lunasinlg protein, and the amounts correlated Average temperatures: intermediate (23CC), low (18C), and high (28CC). with the extent of inhibition of core historic acetylation. These 944 WANG FT AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 91, NC). 4, 2008

Table 4. Effect of cultivar and temperature on lunasin Bowman Birk inhibitor (BBI) was considerably higher than concentration (mglg) that of KTI. BBI values ranged from 0.6 to 6.3% of total Cuttivar Temperaturea Mean ± SDb extractable protein, whereas that of KTI ranged from 4.3 to 6.9% of total extractable protein. Also, the variety with the highest KTI (Krajina) also had the highest BBI, and the variety Dwight I 7.98 ± 0.64 Ell with the lowest Ku (Vojvodjanka) also had the lowest BRI. L 9.54 ± 1.10 BDCE Genotypes with a high percentage of trypsin inhibitor; 7.41 ± 0761GH H especially BBI, could have a significant role from the Imari I 8.38 ± 1.4 CDEll nutraceutica] point of view and might be used in cancer L 8.11 ±O.gE1° prevention and therapy. Deshimaru ci al. (31) were able to H 6.10±1.14H isolate 9 protcinasc inhibitors from wild soy seeds. Two inhibitors were classified as a soybean KI family and the others Jack 1 962216BDC were stable to heat and extreme pH, suggesting that these L 978114ABC belonged to the BBI family. Vollmann et al. (32) studied 8.98 ± 079BDCEF H environmental influences and the effects of nitrogen and sulfur Loda I supply on TIA of soybean. TIA was affected significantly by L 9.7O±l.27ABC environment (geographical location), fertilization type, and H 1029099AB genotype. They used 3 macroenvironmcnts (of different elevations in 3 years), and the environmental means of TIA was Queen I 8.57 ± 123CDEFG 69.5-104.8 mg tiypsin inhibitcdlg defatted soy flour. Their L 7.01 ± 097GH results showed that significant genetic variation in TIA was 7.61 ± 0,26FGH H found within the genotype and suggested that TIA of soybean may be modified considerably by genetic improvement and Average temperature of growth: I = intermediate temperature appropriate agronomic management. Friedman et al. (33) (23°C); L = low temperature (18°C); and H = high temperature (28°C). studied Williams 82 standard soybean cultivar, an isoline Values, mean ± standard deviation (SD), followed by the same (L81-4590) lacking the KTI and 13 cultivars from the USDA letter(s) are not significantly different from each other (P> 0.05); soybean gerrnplasrn collection. L81-4590 had only 54% TIA n = 4. and 791N, chymotmypsin inhibitory activity. All 13 cultivars exhibited values ranging from 4.7 to 13.2 mg chymotmypsin inhibited/- flour and 37.2 to 61.4 11mg TIAIg. Becker-Ritt et lunasin concentration on commercially available soy protein al. (34) found significant differences in the proteinase inhibitor and isofiavone-cnriclied products. They showed that lunasin content of 6 soybean cultivars. Differences in the inhibitory concentration of commercially available soy protein ranged activity towards tlypsin might be related to different levels of from 13 ± 2 mg lunasinlg flour (soy flour and soy flakes) to the BBI since the levels of KTI appeared to be similar, as 44 ± 6 nig lunasinlg flour. Lunasin concentration in lunasin-enriched flour was 27.3 rng/g flour. However, the isoflavone-enriched products contained from zero to a low lunasin concentration, possibly due to the poor solubility of Table 5. Effect of cultivar and soil moisture on lunasin lunasin in ethanol, which was used in the extraction of concentration (mglg) soybean isoflavones. Jcong et al. (29) found that defatted soy Level of cultivar Moisturea Mean ± SDb flour had the lowest lunasin concentration (5.48 ± 0.17 mg/g protein) when compared to soy concentrate (8.72 ± 0.19 to Dwight L 8.01 ± 1.25 BCD 16.52 ± 0.23 mg lunasin/g protein) and soy isolate (6.92 ± H 8.61 ± 0.14 mg lunasin/g protein). They also showed that water-washed soy concentrate had a higher concentration of man L 6.86 ± lunasin (16.52 ± 0.23 mg lunasin/g protein) than H 820151BC alcohol-washed concentrate (8.72 + 0.19 mg lunasing Jack L 1044123A protein). It has been demonstrated that lunasin concentration H 8.48 ± in soy products is affected by processing conditions. Loda L 10. 18 ± 1.29A H 1062104A Other Biologically Active Proteins and Peptides in Soybean Queen L 7.22 ± 0.82 H 82411OBI

Pesic et al. (30) reported the influence of different genotypes ° Soil moisture conditions: L = low soil moisture; H = high soil on the level of trypsin inhibitory activity (TIA) in soybean. moisture. Significant differences were found among 12 soybean Values, mean ± SD, followed by the same letter(s) are not significantly different from each other genotypes used in the study. Also, the extent of variation in (P> 0.05); n = 6. LII-

WANG El AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 91, No. 4.2008 945

Table 6. Effect of cultivar, temperature of growth, and BAPNA as substrate, Voliman et al. (32) used azocasein as one soil moisture on lunasin concentration (mglg) of the substrates and detennined TIA by measuring the residual activity of bovine trypsin used as standard. Cultivar Temperaturea Moisturea Mean ±8Db In summary, other biologically active peptides and proteins present in soybean are also affected by cultivar and growing Dwight I L 7.71 ± 079JKL1 8.24 ± 057FGHIJ}< conditions, the same observations found regarding lunasin I H concentration in soybean. L L 9.03 ± 1.52 11F1H1 10.05 ± 051B0CE L H Conclusions H L 7.28±1.27°- 7.53 ± 023JKL1 H H This paper discusses techniques recently used for the 7.40 ± 083JKL marl I L analysis of biologically active peptides in . It is apparent that these techniques are available for general use I H 9.36±1.16c0EF and require relatively expensive equipment, laboratory L L 7.82 ± 001HIJKL 8.41 ± 1•45FGHIJK supplies, and highly trained personnel. Choosing a method L H should be based on chemical composition and stability of the 5.36 ± 040M H L material to be analyzed. As shown in this review, the best H H 6.85±1.25KML method for the analysis of different bioactive peptides in food Jack I L 11.40±1.17AB includes a combination of different LC techniques because 7.85 ± 008GHIJKL one method increases the efficacy and efficiency of other H methods. Also, this review reports the effect of cultivar, L 1053127ABCD L environmental factors and processing on lunasin L H 9.03 ± concentration of soybean and soybean products. Soybean 9.40 ± 072CDEFGH H L cultivars and environmental conditions where soybean was 8.57 ± 081EFGHIJ H H grown also affected the concentration of lunasin in soy flour. 1Q77085ABC Loda I L The interactions of cultivars, soil moisture conditions, and 1168034A growing temperature significantly affected lunasin I H + 046EFGHIJ concentration, suggesting that its concentration in soybean L L 8.64 ABC can be manipulated. Other bioactive peptides and proteins L H 10.76 ± present in soybean were also affected by cultivar and 1115013AB H L environmental conditions. The significant effect of processing 9.43 ± 000CF0 H H was shown by the differences in the concentration of lunasin 7.63 ± 054JKL1 Queen I L in defatted soy flour, soy isolate, alcohol- and water-washed soy concentrates, and isoflavone products. H L L 6.28 ± 053LM 7.75 ± 061JKL1 Acknowledgments L H IKLI H L 7.76 ± 7.45 ± 006JKL1 We acknowledge the financial support of the Illinois H H Soybean Association and the USDA Functional Foods Initiative. L = Low; I = intermediate; H = high temperature. Values, mean ± SD, followed by the same letter(s) are not significantly different from each other (P> 0.05); n = 2. References

(1) Wang. W, & (IC Mcjia, E.G. (2005) Comp. Rev Food Sd. detected by immun ob lotting. Vasconcelos etal. (35) studied the Food Sate/v 4, 63-78 antinutritional and toxic proteins present in 2 Brazilian soybean (2) de Lumen. B.O. (2005) Nun: Rev 63, 16-21 cultivars. Rio Balsas and Bays. Their results showed that the (3) Nielsen. N.C. (1996) Soybean: Genetics, MolecularBiolog Bays cultivar presented significantly higher, about 2-fold, TIA and Biotechnology. D.P.S. Verna & R.C. Shoemaker (Eds), than the Rio Balsas cultivar. Giami (36) reported the chemical CAB International, Wallingford, UK, pp 127-163 composition and antinutritional attributes of 3 promising (4) Panthce. DR., Kwanyucn, P., Sams, CE., West, DR., advanced breeding lines of soybean in Nigeria. The raw seeds Saxton, AM.. & Pantalone, V.R. (2004) .J. Am. Oil Chem. contained 8.6-18.2 rng trypsin inhibitor/g defatted flour. The Soc. 81, 1005-1012 assay used in the analysis of TIA used (5) Ji, M.P., Cai, T.D., & Chang, K.C. (1999) J. Food Sci. 64, a-N-benzoyl-DL-argininc-p-nitroafllhde hydrochloride (BAPNA) 763-767 as substrate, and the activity was determined by measuring the (6) Liu. K.S. (1997) Soybeans: Chemistry, Technology and change in absorbance at 40 5-420 nm. In addition to the use of UiiIi:atzon. Chapman and Hall, New York. NY. 532 pp 946 WANG ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 91, No. 4, 2008

(7) Maruyama, N., Katsube, T., Wada, Y., Oh, M.H., Barba De (23) de Mejia. E.G. Bradford, T.. & Hasler. C. (2003) Nut,: Rei La Rosa, A.P., Okuda, E., Nakagawa, S., & Utsumi, S. (1998) 61,239-246 Eu,: 1. Bioc/zern. 258,854-862 (24) de Mejia, E.G., Vasconez, M.. de Lumen. B.O.. & Nelson, R. (8) Natarajan. S., Xu. C., Bae, H., & Bailey, B. (2007)1. Plant (2004)1 Agiic. Food Chem. 52,5882-5887 Phvsio/. 164, 756--763 (25) Vasconcz-Costa, M. (2004) Wasters Thesis, University of (9) Lo, W.M.Y.. & Li-Chan, E.C. (2005) J. Agric. Food Ghe,n. Illinois. Urbana-Champaign. IL 53,3369-3376 (26) Lozovaya, V.V., Lygin, A.V., Ulanov, A.V., Nelson. R.L., (10) Zhang, J.H., Tatsumi, E., Deng. C.. & Li, L. (2006) Food Dayde. J., & Widholm. iD. (2005) Crop Sc!. 45. 1934-1940 Chern. 48, 551-557 (27) Vasconcz, M., Nelson. R.. & dc Mcjia, E.G (2005) lust. (11) Gibbs, B.F., Zougman. A., Masse, R., & Mulligan, C. (2004) Food Tee/no!. 85, 8 Food Res. Int. 37. 123-131 (28) Jeong, H.J., Jeong, J.B., Kim. D.S., & de Lumen, B.O. (12) Kuba, M., Tanaka, K., Tawata, S., Takeda, Y., & Yasuda. M. (2007)1. Agric. Fooc/ C/tern. 55, 632--637 (2003) Biosci. Biotechnol. Biochern. 67, 1278-1283 (29) Jeong, H.J., Park, J.H., Lam, Y.. & de Lumen, B.O. (2003)1. (13) Mallikaijun Gouda, KG., Gowda, L.R., Appu Rao, A.G., & Aguk. Food Chem. 51. 7901-7906 Prakash, V. (2006)1 Agric. Food Chain. 54, 4568-4573 (30) Pesic, MB., Vucelic-Radovic, B y., Barrac, M.B., (14) Kodera, T., & Nio, N. (2006)1 Food Sc!. 71, C164—C173 Stanojevic. S.P.. & Ncdovic, V.A. (2007) Sensors 7. 67-74 (15) Kuba, M., Tana, C., Tawata, S., & Yasuda, M. (2005) (31) Deshimaru, M.. 1-lanamoto. R., Kusano. C., Yoshimi. S.. & Process Biochern. 40,2191 —21196 Terada, S. (2002) Biotech. Biochern. 66. 1897-1903 (16) Zhong, F., Liu, J., Ma, J., & Shoemaker, C.F. (2007) Food (32) Vollmann, I., Grausguber. H., Wagentristl, H., Wohlescr. H., Res. Int. 40, 661-667 & Michele. P. (2003)1. Sci. Food Agric. 83, 1581-1586 (17) Zhong, F., Zhang, X., Ma, J., & Shoemaker, C.F. (2007) (33) Friedman, M., Brandon. DL., Bates, A.H., & Hymowitz, T. Food Res. Int. 40, 756-762 (1991)1 Agile. Food C/tern. 39. 327-335 (18) Gianazza, E., Eberini, 1., Arnoldi, A., Wait, R., & Sirton, C. (34) Becker-Ritt, A.B., Mulinari, F., Vasconcelos. TM.. & Carlini. (2003)1. Note 133, 9-14 C . R. (2004) 1 Sc!. FoodAgric. 84, 263-270 (19) Lee, S.H., Yang, J.1., Hong, S.M., Hamh, D.H., Lee, S.Y., (35) Vasconcelos, I.M., Maia, A.A.B., Siebra, E.A., Oliveira, Kim, 1.1-I., & Choi, S.Y. (2005) Biofàctocs 23, 121-128 J.T.A., Carvalho, A.F.F.U., Melo, V.M.M., Carlini, CR.. & (20) Rho, S.J., Park, S., Ahn, C.W., Shin, J.K., & Lee, H.G. (2007) Castelar, L.I.M. (200 1).J. NuIr. Biochem. 12, 55-62 1 Sci. Food Agric. 87, 908-913 (36) Giami, S.Y. (2002)1 Sd. FoodAgric. 82, 1735-1739 (21) Saz. J.M.. & Marina. M.L. (2007)1 Sep. Sci. 30, 431-451 (37) Tsuruki, T., Kishi, K., Takahashi, M., Tanaka, M., (22) Galvez, A., Chen, N., Macasieb, J.. & de Lumen, B.O. Matsukawa, T., & Yoshikawa, M. (2003) FEBS Leil. 540, (2001) Cancer Res. 61, 7473-7478 206-210