Extraction of Antioxidative and Antihypertensive Bioactive Peptides

Food Chemistry 141 (2013) 3435–3442

Contents lists available at SciVerse ScienceDirect

Food Chemistry

journal homepage: www.elsevier.com/locate/foodchem

Extraction of antioxidative and antihypertensive bioactive peptides from Parkia speciosa seeds ⇑ Hwee-Leng Siow a,b, Chee-Yuen Gan b, a Bioprocess Division, School of Industrial Technology, Universiti Sains Malaysia, 11800 USM, Penang, Malaysia b Centre for Advanced Analytical Toxicology Services, Universiti Sains Malaysia, 11800 USM, Penang, Malaysia article info abstract

Article history: Antioxidative and antihypertensive bioactive peptides were successfully derived from Parkia speciosa Received 29 March 2013 seed using alcalase. The effects of temperature (25 and 50 °C), substrate-to-enzyme ratio (S/E ratio, 20 Received in revised form 3 June 2013 and 50), and incubation time (0.5, 1, 2 and 5 h) were evaluated based on 2,2-diphenyl-1-picrylhydrazyl Accepted 6 June 2013 (DPPH), ferric reducing antioxidant power (FRAP) and angiotensin-converting enzyme (ACE) assays. Bio- Available online 15 June 2013 active peptide extracted at a hydrolysis condition of: temperature = 50 °C, S/E ratio = 50 and incubation time = 2 h, exhibited the highest DPPH radical scavenging activity (2.9 mg GAE/g), reducing power Keywords: (11.7 mM) and %ACE-inhibitory activity (80.2%). The sample was subsequently subjected to fractionation Antioxidative and the peptide fraction of <10 kDa showed the strongest bioactivities. A total of 29 peptide sequences Antihypertensive Bioactive peptide from peptide fraction of <10 kDa were identiﬁed as the most potent contributors to the bioactivities. Parkia speciosa These novel bioactive peptides were suggested to be beneﬁcial to nutraceutical and food industries. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction to health (Duranti, 2006). In view of this, plant-derived bioactive peptides have been attracting a great deal of interest in the Today, diet and a healthy lifestyle play a significant role in the scientific community. Among all plant sources, legumes (the sec- quality of our lives. The increasing awareness of the importance ond largest food crops after cereals for world agriculture) is a well of a healthy diet leads to the development of new, safe and healthy known source for the generation of bioactive role of peptides and foods. Natural food-derived peptides with specific bioactivity were are believed to provide nutritional benefits as they are rich in high therefore aroused a great deal of interest among researchers. quality of protein (Duranti, 2006). Food-derived bioactive peptides was defined as the specific protein Parkia speciosa, a Southeast Asian legume of the Mimosae fragments that show beneficial pharmacological properties in the subfamily, grows wild in the lowland tropical forests and is often human body beyond the normal and adequate nutrition cultivated in Malay villages. The tree bears long, flat, edible bean (Hartmann & Meisel, 2007). In recent years, it was acknowledged pods with bright green plump almond shaped seeds that have a that the dietary protein is a good source of bioactive peptides with unique Shitake mushroom-like flavour. The seeds of P. speciosa a broad spectrum of biological activities, including antioxidant, are flattened and elliptical in shape with a nutty and firm texture. antihypertensive, anti-inflammatory, opioids and immuno- Besides that, P. speciosa has earned its nickname ‘‘stink bean’’ due stimulants (Hartmann & Meisel, 2007). The precursor of bioactive to its strong and pungent odor, which could cause the body to ex- peptides can be categorised according to animal origin, marine crete a distinct smell through the skin, urine and faeces. P. speciosa origin as well as plant origin. Bioactive peptides derived from the seeds have always been a popular ingredient in cooking and animal source (milk, egg and meat muscle) and the marine source usually served with sambal, dried shrimp and chili pepper, as a (fish, salmon, oyster and seahorse) have been investigated in asso- popular local delicacy in Malaysia. Besides culinary uses, it is ciation with their pharmacological properties (Udenigwe & Aluko, reported to contain crucial chemical medicinal compounds which 2012). Nevertheless, plant proteins are emerging as an important exhibit potential biological activity such as anticancer (Ali, Hough- food ingredient to be used for the improvement of modern foods ton, & Soumyanath, 2006), antibacterial (Sakunpak & Panichayupa- in the aspect of nutrition, processing technology and contribution karanant, 2012), antioxidant (Aisha, Abu-Salah, Alrokayan, Ismail, & Abdul Majid, 2012), antiangiogenic (Aisha et al., 2012) as well as hemagglutinating activity (Chankhamjon, Petsom, Saw- ⇑ Corresponding author. Address: Bioprocess Division, School of Industrial asdipuksa, & Sangvanich, 2010). However, the potential roles of Technology, Universiti Sains Malaysia, 11800 USM, Penang, Malaysia. Tel.: +60 4 P. speciosa seed-derived bioactive peptides with a biofunctional 6595606x2461; fax: +60 4 656 9869. activity have not been investigated. E-mail address: [email protected] (C.-Y. Gan).

Peptides are inactive when it is encrypted in the parental solution containing 40% methanol and 10% acetic acid. The image protein but can be released with various biological activities was subsequently captured using Fujifilm luminescent image (Korhonen & Pihlanto, 2006). The production, characterisation analyzer LAS-3000 (Fujifilm, Tokyo, Japan). Multi Gauge Version and in vitro or in vivo evaluation in food has been extensively 3.0 software (Fujifilm, Tokyo, Japan) was used to analyse the elec- reviewed in recent years. Enzymatic hydrolysis is found to be the trophoresis pattern. Bio-Rad’s prestained SDS–PAGE standard with most common way to produce bioactive peptides using proteolytic broad range molecular weight (MW 6.0–202.4 kDa) was used as enzymes such as pepsin, trypsin, alcalase, and pancreatin (Korho- standard for comparison. nen & Pihlanto, 2006). The main objective of this project was to extract antioxidative and antihypertensive bioactive peptide from 2.5. 2,2-Diphenyl-1-picrylhydrazyl (DPPH) free radical scavenging P. speciosa seed using enzyme alcalase. Following, the identification activity of the extracted bioactive peptide using mass spectrometry approach was performed. The antioxidant activity of non-hydrolysed sample and protein hydrolysate (at a concentration of 0.5 mg/ml) extracted from P. 2. Materials and methods speciosa seed was determined using the DPPH free radical scavenging assay, as described by Lee, Kwon, Shin, and Yang (2009). Prior 2.1. Materials to analysis, the samples were pre-diluted at a factor of 5 and DPPH stock solution (0.1 mM) in ethanol was prepared. A 500 ll DPPH P. speciosa seeds were purchased from local markets (i.e. Air stock solution was added to 16.65 ll sample and incubated at Itam market, Jelutong market and Apollo market), Penang. The 30 °C for 30 min in the dark. After incubation, the sample was cen- seeds were lyophilised using Labconco Freeze-dryer (Fisher Scien- trifuged at 14,000g for 5 min. The absorbance of the sample was tific, USA) and ground into powder (average diameter of 30 mesh then measured at 517 using a spectrophotometer (Spectramax M5, screen size) prior to extraction of bioactive peptide. Alcalase with Molecular Devices, USA). A control sample (16.65 ll of methanol activity of 2.4 AU-A/g was purchased from Novoenzyme A/S, Den- and 500 ll DDPH solution) was prepared as mentioned above. mark. All other chemicals and reagents used in the experiment The antioxidant activity was expressed as percentage of DPPH free were of analytical grade purchased from Sigma–Aldrich (Malaysia) radical scavenging activity (%DPPHsc) and calculated using company or otherwise mentioned. formula:

%DPPHsc ¼ðAcontrol AsampleÞ=Acontrol 100% ð1Þ 2.2. Extraction of bioactive peptides from P. speciosa seeds where Asample is the absorbance of sample at t = 30 min and Acontrol is the absorbance of control sample. Garlic acid at a concentration P. speciosa seed ﬂour (1 g) was suspended in 10 ml of phosphate range of 0–60 lg/ml was used as standard. buffer at pH 8.0 and alcalase was added into the suspension with substrate-to-enzyme (S/E) ratio of 20 or 50. The hydrolysis was 2.6. Ferric reducing antioxidant power (FRAP) assay performed via incubation at 25 or 50 °C with constant shaking at 200 rpm for 0.5, 1, 2, and 5 h. The hydrolysis reaction was then ter- The reducing abilities of samples (at a concentration of 0.5 mg/ minated by heating the sample at 95 °C for 30 min in a water bath ml) towards ferric ions were determined according to the method (Memmerit, Germany), followed by centrifugation (10,000g)at of Benzie and Strain (1996) with some modiﬁcations. The working 4 °C for 30 min. The supernatant was collected and stored at FRAP reagent contained 10 mM 2,4,6-tri(20-pyridyl)-s-triazine 80 °C for further analysis. Non-hydrolysed sample was extracted (TPTZ) solution in 40 mM HCl, 20 mM FeCl36H20 and 0.3 M acetate according to the method aforementioned without the presence of buffer at pH 3 at a ratio of 1:1:10. The reagent was pre-warmed at enzyme. 37 °C and the samples were pre-diluted at a factor of 5 prior to analysis. The pre-diluted sample (2.7 ll) was then added to 2.3. Amino acid analysis 200 ll of FRAP reagent and mixed thoroughly. The sample was subsequently incubated at 37 °C for 1 h and the absorbance was A 0.1 g hydrolysed samples were heated at 110 °C for 24 h after measured at 593 nm using a spectrophotometer (Spectramax adding with 5 ml of 6 N HCl and purged with nitrogen. The samples M5, Molecular Devices, USA). Iron (II) sulphate heptahydrate were derivatised and then analysed using Waters-HPLC-System (FeSO47H2O) at a concentration range of 0–2 mM was prepared (USA) coupled with Waters 24675 Multi-k Fluorescence Detector as standard. The ability of ferric reducing antioxidant potential (Zarkadas et al., 2007). Methionine and cysteine were determined was expressed in mM FeSO4. separately according to the performic acid procedure of Moore (1963). Each analysis was performed in three replicates. 2.7. Angiotensin-converting-enzyme (ACE) inhibitory activities

2.4. Sodium dodecyl sulphate polyacrylamide gel electrophoresis The analysis of antihypertensive activities was determined as (SDS–PAGE) outlined by Cushman and Cheung (1971) with some modifications. Pre-diluted sample (50 ll, at a concentration of 0.5 mg/ml) was SDS–PAGE analysis of the non-hydrolysed sample and protein mixed with 50 ll ACE solution (50 mU/ml). The mixtures were hydrolysate extracted from P. speciosa seed was performed using pre-incubated at 37 °C for 10 min in a heating block (Grant, Fisher 15% resolving gel and 4% stacking gel. Samples (10 lg of protein) Scientific, UK). A 150 ll of 4.15 mM substrate (i.e. hippuryl-histi- were added with 10 ll of Laemmli sample buffer and 1 llof dyl-leucine in borate buffer containing 0.3 M NaCl, pH 8.3) was 2-mercaptoethanol. The samples were then heated at 95 °C for then added into the sample and incubated at 37 °C for 30 min. 5 min using heating block (Grant, Fisher Scientific, UK). Samples The reaction was subsequently stopped by adding 500 llof1M were loaded into the well and ran the electrophoresis at a constant HCl. Ethyl acetate (1.5 ml) was then added to extract the hippuric voltage of 80 V for 10 min, followed by 120 V for 120 min using acid. The resulting mixture was vortex for 1 min and stood for Mini Protean III Cell (Bio-Rad, USA). The gel was then stained for 5 min. An 800 ll of ethyl acetate layer was then transferred into 2 h using 0.1% Bio-Rad Coomassie Brilliant Blue R-250 in 40% 2 ml microcentrifuge tube and vacuum dried in a vacuum concen- (v/v) methanol and 10% acetic acid, followed by destaining with trator (Concentrator 5301, Eppendorf, Germany) at 45 °C for H.-L. Siow, C.-Y. Gan / Food Chemistry 141 (2013) 3435–3442 3437

30 min. After drying, 1 ml of deionised distilled water was added the eight most abundant ions from the parent mass list of pre- and vortex until the residual was fully dissolved. The concentration dicted peptides with rejection of singly or unassigned charge state. of hippuric acid was determined using spectrophotometer (Spec- The ITMS analysis was performed using the same resolving power tramax M5, Molecular Devices, US) at 228 nm. A positive control (60,000) and CID was conducted with isolation width of 2 Da, nor- and negative control sample, which contains 50 ll 1 M HCl and malised collision energy of 35, activation q of 0.25, activation time 50 ll distilled water respectively, as a replacement of sample of 50 ms and charge state of 2 or higher. was prepared. The calculation of the % ACE inhibition as follows: Data acquisition was performed using Xcalibur ver. 2.1 (Thermo Scientific, San Jose, CA, USA) with mass tolerance threshold of %ACE inhibition A A 100=A 2 ¼ð negative sampleÞ negative ð Þ 5 ppm. Data analysis was performed using PEAKS studio ver. 6.0. where Anegative is the absorbance of negative control and Asample is the absorbance of diluted sample. The absorbance of positive con- 2.9. Statistical analysis trol was used as the reading correction. Statistical analysis was performed using SPSS for Windows, Ver- 2.8. Fractionation and identification of the bioactive peptides extracted sion 12.0 (SPSS Institute, Inc., Cary, NC, USA). The means of results from P. speciosa seeds were compared using one-way analysis of variance (one way ANO- VA) and significant level was of P 6 0.05. All the measurements The most reactive extract was chosen for this part of the study. were performed in at least three replicates. The sample was fractionated according to the molecular mass and the biological activities (e.g. antioxidant and antihypertensive 3. Results and discussion activity) of each fraction were determined. The most active fraction was used for identification of bioactive peptide using mass spec- 3.1. Amino acid content of P. speciosa seeds protein hydrolysates trometry approach. Analysis of amino acid composition was carried out prior to the 2.8.1. Fractionation of bioactive peptides extraction in order to identify the potency of bioactive peptides The bioactive peptides were fractionated through a series of from P. speciosa seeds protein hydrolysates. Table 1 shows the ami- centrifugal ultrafiltration filters (Amicon, Milipore, Ireland) with no acid composition of P. speciosa seeds protein hydrolysates, molecular weight cut-off (MWCO) of 10, 30 and 50 kDa. The sam- which was expressed as per 1000 amino acid residuals. The results ple was first passed through a filter of 50 kDa via centrifugation at showed that Glu and Asp were found in the highest amount, 4,500g for 15 min. (Thermoscientific, Fisher Scientific, Germany). approximately 154 and 132 per 1000 amino acid residuals, respec- Deionised distilled water was added to the retentate and was tively. Previous studies revealed that these two amino acids pos- washed for 3 times using the same filter via centrifugation. The sess strong antioxidant activities. This is due to the presence of retentate was then collected as fraction >50 kDa. The resulting excessive electrons that can be donated when they interact with filtrate was then passed through a filter with 30 kDa MWCO mem- free radicals (Udenigwe & Aluko, 2011). Cys was the third abun- brane, via centrifugation (Centgrifuge 5417R, Eppendorf, Germany) dant amino acid residuals found in the hydrolysates, approxi- at 10,000g for 15 min. The retentate was washed with deionised mately 84.8 per 1000 amino acid residuals. According to distilled water for 3 times using the same filter via centrifugation. Patterson and Rhoades (1988), the SH group in Cys can also act The retentate was collected as fraction 30–50 kDa. Again, the as radical scavenger and it can protect tissue from oxidative stress. filtrate was added into a filter with 10 kDa MWCO and centrifuge Apart from that, the presence of Tyr, Met, His, and Lys were also at 4500g for 15 min and subsequently washed for 3 times. The suggested to be potent antioxidants (Udenigwe & Aluko, 2011). retentate was collected as fraction <10 kDa. The antioxidant and Table 1 also revealed that the protein hydrolysates were rich in antihypertensive properties of bioactive peptide fractions were hydrophobic, branched and aromatic amino acids, such as Ile, Val, then analysed using the DPPH (Section 2.5), FRAP (Section 2.6) Phe and Tyr, to give a comparative lower IC50 value of ACE-inhibi- and ACE-inhibitory (Section 2.7) assays as previously described. tory activities (He et al., 2007). The high content of Glu, Asp, Pro, Gly and Ala have also been found in many other ACE-inhibitory peptides (Zhao et al., 2007). This result served as a preliminary 2.8.2. Identification of bioactive peptides using mass spectrometry LC MS and MS/MS analyses on peptides fraction of <10 kDa were carried out using Thermo LTQ/Orbitrap Velos (Thermo Scien- Table 1 tific, San Jose, CA, USA) coupled with Easy-nLC II system (Thermo Amino acid composition of Parkia speciosa seeds protein hydrolysates (expressed as Scientific, San Jose, CA, USA). per 1000 amino acid residuals). Chromatographic separation of peptides was performed using Amino acid codes Amount (per 1000 amino acid residuals) Easy-Column C18-A2 (100 0.75 mm i.d., 3 lm; Thermo Scientific, Asp 132.1 ± 0.3 San Jose, CA, USA) coupled with pre-column (Easy-Column, Ser 44.2 ± 0.9 20 0.1 mm i.d., 5 lm; Thermo Scientific, San Jose, CA, USA) at Glu 154.0 ± 0.3 flow rate of 0.3 ll/min and injection volume of sample was 10 ll. Gly 45.2 ± 0.9 Pre-column was equilibrated for 15 ll at flow rate of 3 ll/min His 17.5 ± 0.2 Arg 68.5 ± 1.7 whereas analytical column was equilibrated for 4 ll at flow rate Thr 43.1 ± 1.1 of 0.3 ll/min. Running buffers used were: (A) deionised distilled Ala 45.8 ± 0.0 water with 0.1% formic acid, and (B) acetonitrile with 0.1% formic Pro 35.9 ± 0.8 acid. The pump gradient elution of nano-LC was as follows: 0– Cys 84.8 ± 8.6 Tyr 57.5 ± 1.5 70 min, 5–45% B; 70–85 min, 45–100% B; and 85–100 min, 100% B. Val 37.5 ± 15.9 The eluant was sprayed into mass spectrometer at 2.3 kV Met 28.5 ± 26.0 (source voltage) and capillary temperature of 200 °C was used. Lys 70.4 ± 1.5 Peptides were detected by full scan mass analysis from m/z 200– Ile 36.1 ± 0.5 2,000 at resolving power of 60,000 (at m/z 400, FWHM; 1-s acqui- Leu 61.5 ± 0.7 Phe 37.2 ± 0.3 sition) with data-dependent MS/MS analyses (ITMS) triggered by 3438 H.-L. Siow, C.-Y. Gan / Food Chemistry 141 (2013) 3435–3442 screening to access the potency of bioactivity from P. speciosa 38 kDa was not hydrolysed, suggesting that this subunit is in the seeds-derived peptides. A rich source of potential functional amino inner part of the protein that hindered alcalase from accessing it acid identified in the protein hydrolysates was therefore suggested (Larré, Chiarello, Dudek, Chenu, & Gueguen, 1993). At incubation to contribute to the antioxidant and ACE inhibition activities. temperature of 50 °C with the used of high amount of enzyme (S/E ratio of 20), a more extensive hydrolysis process occurred and produced a dominant hydrolysed protein with molecular 3.2. SDS–PAGE analysis of protein hydrolysates from P. speciosa seeds weight of 1.5–2.9 kDa (Fig. 1c). On the other hand, a partial hydrolysis of protein could be observed when a lower amount of enzyme Fig. 1 shows the SDS–PAGE profiles of hydrolysed and non- was used (S/E ratio of 50) (Fig. 1d). Hydrolysed protein with molec- hydrolysed samples of P. speciosa seeds protein under different ular weight of 1.5–6.2, 11.8 and 13.2 kDa could be obtained when process conditions. Result indicated that the protein bands of the hydrolysis took place up to 2 h. After the incubation period ex- non-hydrolysed samples have molecular weight ranging from 13 tended to 5 h, only hydrolysed protein with molecular weight of to 51 kDa regardless of incubation time and temperature. The high- 1.5–6.2 kDa could be observed. These differences in molecular est molecular weight of protein bands (48 and 51 kDa) are weight could result in different biological activities (Kim, Byun, likely to represent the subunit of vicillin as reported in the study Park, & Shahidi, 2001). of pea seed globulin by Tzitzikas, Vincken, De Groot, Gruppen, and Visser (2005). The protein bands with molecular weight of 30–45 kDa are considered as the acidic subunit of legumin (She- 3.3. 2,2-Diphenyl-1-picrylhydrazyl (DPPH) free radical scavenging wry, Jenkins, Beaudoin, & Clara Mills, 2004) and the two lower activity molecular weight bands (13–17 kDa) may correspond to the d- vicillin or polypeptides from the post-translational cleavage of Table 2a shows the DPPH scavenging activities (DPPHsc) of the the storage proteins (Shewry et al., 2004). This result is in line with non-hydrolysed samples ranging from 1.6 to 2.2 mg GAE/g seed. other research that revealed legume seeds are rich in globulin pro- After the protein was hydrolysed, a slightly higher value of DPPHsc teins, which can be classified into two major classes: legumin and was observed (2.1–2.9 mg GAE/g seed). This improvement suggests vicilins. the ability of these peptides to scavenge free radicals, which was With presence of alcalase, it could be observed that the enhanced through enzymatic hydrolysis. Furthermore, it is sug- enzymatic hydrolysis led to a gradual breakdown of the protein. gested that a partial hydrolysed protein chain has a higher expo- Higher temperature, longer incubation period and a lower S/E sure of amino acid residue which could act as a hydrogen donor ratio, enhanced the hydrolysis process. Fig. 1a and b shows that and thus enhance the scavenging activities of peptides (Rehman most of the proteins were hydrolysed by alcalase at a temperature & Shah, 2005). of 25 °C regardless of S/E ratio. The hydrolysed protein with molec- Temperature showed its importance in affecting the DPPHsc of ular weights of 1.5–5.0, 11.8 and 13.2 kDa were produced. It is protein hydrolysates, with values increasing from approximately interesting to note that the protein with molecular weight of 2.1–2.2 to 2.1–2.9 mg GAE/g seeds at S/E ratio of 50 when

Fig. 1. SDS–PAGE proﬁle for samples produced at following conditions: (a) S/E ratio of 20, temperature of 25 °C; (b) S/E ratio of 50, temperature of 25 °C; (c) S/E ratio of 20, temperature of 50 °C; and (d) S/E ratio of 50, temperature of 50 °C. Lane 1: standard marker. Lanes 2, 4, 6 and 8 are non-digested samples incubated for 0.5, 1, 2 and 5 h, respectively. Lanes 3, 5, 7 and 9 are alcalase-digested samples which incubated for 0.5, 1, 2 and 5 h, respectively. H.-L. Siow, C.-Y. Gan / Food Chemistry 141 (2013) 3435–3442 3439

Table 2

(a) DPPH free radical scavenging activity (expressed as mg GAE/g of seeds); (b) ferric reducing antioxidant potential (expressed as millimole of iron (II) sulphate, FeSO4) and (c) percentage of ACE-inhibitory activity (expressed as% ACE-inhibitory activity) of non-hydrolysed and hydrolysed samples from Parkia speciosa seeds.

Sample T* (°C) S/E ratio Time (h) 0.5 1 2 5

(a) DPPHSC Non-hydrolysed 25 1.9 ± 0.1ab 2.2 ± 0.2cd 2.2 ± 0.2cd 1.9 ± 0.4ab 50 1.8 ± 0.2ab 1.6 ± 0.1a 1.8 ± 0.3ab 1.7 ± 0.1a Hydrolysed 25 20 2.1 ± 0.1c 2.1 ± 0.1c 2.1 ± 0.1c 2.1 ± 0.1c 50 2.1 ± 0.1c 2.1 ± 0.2c 2.2 ± 0.2cd 2.2 ± 0.1cd 50 20 2.1 ± 0.1c 2.2 ± 0.1cd 2.3 ± 0.1cd 2.3 ± 0.2cd 50 2.3 ± 0.2cd 2.6 ± 0.2e 2.9 ± 0.1f 2.1 ± 0.1c (b) FRAP Non-hydrolysed 25 3.9 ± 0.3a 4.4 ± 0.9a 4.6 ± 0.7ab 4.9 ± 0.3ab 50 4.4 ± 0.2a 4.8 ± 0.6ab 5.1 ± 0.6ab 5.3 ± 0.2ab Hydrolysed 25 20 6.1 ± 0.7b 6.3 ± 0.0b 7.4 ± 0.6c 7.9 ± 0.8cd 50 6.6 ± 0.3bc 7.5 ± 0.7c 8.1 ± 1.0d 8.9.±0.5de 50 20 6.5 ± 0.7b 7.1 ± 0.2c 8.4 ± 0.2d 7.8 ± 0.5cd 50 7.3 ± 0.6c 8.8 ± 0.7de 11.7 ± 0.8f 9.1 ± 0.8e (c) %ACE-inhibitory Non-hydrolysed 25 ND** ND ND ND 50 ND ND ND ND Hydrolysed 25 20 51.7 ± 4.9a 57.5 ± 1.5ab 55.5 ± 2.7a 55.7 ± 1.6a 50 59.0 ± 3.6ab 53.9 ± 9.4a 56.5 ± 2.3a 58.7 ± 0.0ab 50 20 50.6 ± 3.9a 64.8 ± 0.7c 65.6 ± 3.5c 54.1 ± 0.2a 50 65.2 ± 5.0c 74.5 ± 2.2d 80.2 ± 2.8e 73.9 ± 4.1d

Values are expressed as means ± sd. a–fMeans the column by different letters are signiﬁcantly different (P < 0.05). * T means incubation temperature. ** ND means non-detected.

incubation temperature increased from 25 to 50 °C. However, there time regardless of incubation temperature and substrate-to- was no significant (P > 0.05) changes in samples hydrolysed at S/E enzyme ratio. Again, it is suggested that a certain size of peptide ratio of 20. On the contrary, the non-hydrolysed sample has the (partially hydrolysed) is required to posses the biological activity. reverse effect. As the temperature increased, the DPPHsc values de- If the incubation periods are extended, an extensive hydrolysis creased, particularly at incubation time of 1–2 h. This is due to the would occur and result in shorter hydrophilic peptides that are modification of protein structure caused by thermal treatment. The inaccessible to DPPH free radicals (Bamdad, Wu, & Chen, 2011). study of Rehman and Shah (2005) described that the accessibility of protein to enzyme attack is increased due to the exposure of 3.4. Ferric reducing antioxidant power protein active site after structure modification. In addition, the pro- ton donating residues exposed at high temperature (around 50 °C) For the non-hydrolysed samples, the FRAP values ranged from due to unfolding of protein molecules, could contribute to the anti- 3.9 to 6.3 mM as shown in Table 2b, and therefore it is suggested oxidant activity. Temperature is not the only factor in improving that the P. speciosa seeds protein could be a strong antioxidant. It the DPPHsc of the hydrolysates with results showing that some is interesting to note that the FRAP was significantly (P < 0.05) im- hydrolysates produced at different process conditions were not af- proved after hydrolysed by alcalase. The FRAP values of hydrolysed fected by increasing of temperature. This suggests temperature samples were in the range of 6.1–11.7 mM. Lin, Tian, Li, Cao, and could only improve the DPPHsc of hydrolysates if the process vari- Jiang (2012) and Udenigwe and Aluko (2011) reported that the ables such as the S/E ratio and incubation time are appropriate. use of this enzyme could increase the surface exposure of amino Without hydrolysis, the temperature had a negative effect towards acid residues (e.g. tyrosine, histidine, lysine and methionine) that the DPPHsc of samples which could be due to the instability of the contribute to the antioxidant activity. This is supported by the protein (i.e. conformation change of protein) at high temperature result of amino acid composition (Table 1) where the protein that reduces its ability to quench the DPPH radicals. . hydrolysates contained a high amount of these amino acid that is The role of S/E ratio is also vital in affecting the DPPHsc of the highly exposed after hydrolysis. produced bioactive peptides. By using a lower concentration of en- Table 2b shows that temperature has a significant effect in pro- zyme, resulted in higher DPPHsc values. At incubation temperature ducing bioactive peptides with high reducing power. For example, of 50 °C, the hydrolysed sample with S/E ratio of 50 was signifi- the FRAP values of protein digests produced at S/E ratio of 50 were cantly (P < 0.05) higher after hydrolysed for 1–2 h. This occurrence approximately 6.6–8.9 mM at 25 °C. The reducing power was sig- suggests that the higher concentration of enzyme could cause nificantly (P < 0.05) increased to approximately 7.3–11.7 mM extensive hydrolysis of proteins, whereas lower concentration of when temperature increased to 50 °C. The strongest reducing enzyme caused only partial hydrolysis of the proteins. power of 8.8 and 11.7 mM was observed at particular incubation Results showed that the DPPHsc values were time-dependent. time of 1–2 h. A similar trend was also observed at S/E ratio of The DPPHsc of hydrolysed sample produced at 50 °C with a S/E 20, where significantly (P < 0.05) higher FRAP values of 7.1 and ratio of 50, was significantly (P < 0.05) increased along with the 8.4 mM was observed for the same incubation period. incubation time. A maximum DPPHsc (2.9 mg GAE/g seed) was As mentioned in Section 3.3, high temperature may increase the obtained at 2 h, however a lower value of 2.1 mg GAE/g seed was exposure of active structures for producing effective electron produced after 5 h of hydrolysis. There were no significant differ- donors which could improve the reducing power of hydrolySed ence (P < 0.05) observed in other samples along the incubation samples. However, the effect of temperature was not significant 3440 H.-L. Siow, C.-Y. Gan / Food Chemistry 141 (2013) 3435–3442

(P < 0.05) for the non-hydrolysed samples. The FRAP values of trends were observed at higher S/E ratio, with the highest inhibi- non-hydrolysed samples were almost similar in the range of 3.9– tory activity (80.2%) observed in a sample that had been hydroly- 5.3 mM regardless of the change in temperature. This could be sed for 2 h. Inouye, Nakano, Asaoka, and Yasukawa (2009) due to its inactive parental protein chains that are not affected reported that thermal treatment increases enzyme-protein inter- by the application of external energy. actions due to the thermal-induced unfolding of the proteins that The effect of S/E ratio on the reducing power of hydrolysates eventually enhance the enzymatic hydrolysis in releasing potent was investigated and shown in Table 2b. Results showed that bioactive peptides. Apart from that, it was suggested that alcalase hydrolysed sample with S/E ratio of 50 showed better reducing was stable at temperature of 50 °C, which could further facilitate power than the sample with S/E ratio of 20. With a sample of a the hydrolysis process. S/E ratio of 50, incubation temperature of 25 °C and incubation It was also found that S/E ratio used in production of bioactive time of 1 and 2 h, resulted in FRAP values of 7.5 and 8.1 mM, peptide had an effect on the %ACE-inhibitory activities. S/E ratio of respectively. When the S/E ratio of 20 was used, the FRAP values 50 produced a %ACE-inhibitory activity of protein hydrolysate of decreased to 6.3 and 7.4 mM, respectively. Result demonstrated approximately 65.2–80.2% that was higher than that of S/E ratio that the use of S/E ratio of 50 provide a desirable composition of of 20. The S/E ratio of 50 is believed to give a sufficient amount enzyme and substrate in a hydrolysis process, which is able to pro- of enzyme to produce desirable peptides with various bioactivities. duce peptides with high reducing power. High concentration of en- However, the role of S/E ratio was not significant (P < 0.05) when zyme did not improve the reducing power of hydrolysates even the process was carried out at room temperature and there was though it could facilitate a complete hydrolysis process in a short no apparent change in the %ACE-inhibitory activities (approxi- time frame. Partial hydrolysis process would be more favourable mately 51.7–59.0%) of the protein hydrolysates regardless of incu- to generate potent bioactive peptides, which is easily controlled bation time. by using a lower S/E ratio. The %ACE-inhibitory activities of hydrolysed samples were in- Result also showed that incubation time critically affects the creased with incubation time. For example, this trend could be ob- production of bioactive peptides. At incubation temperature of served in the sample with S/E ratio of 50 and incubated at 50 °C. 25 °C, all hydrolysed sample possessed a significant increase in The %ACE-inhibitory activity increased from 65.2% to 80.2% for reducing power from incubation time of 2 to 5 h. On the other the sample produced up to 2 h of incubation time. Subsequently, hand, samples hydrolysed at a temperature of 50 °C and a S/E ratio the activity decreased to 73.9% when the sample was hydrolysed of 50, only increased the reducing power after 1 h of hydrolysis, for 5 h. However, the incubation time did not significantly and achieved a maximum values of 11.7 mM at a incubation time (P < 0.05) affect the %ACE-inhibitory activity (approximately of 2 h. After 5 h of hydrolysis, there was a decrease in reducing 51.7–59.0%) of hydrolysates at an incubation temperature of power to 9.1 mM and it has been identified that the size of peptide 25 °C. Hydrolysis process within a 2 h timeframe, seems to be is a key factor in generating potent antioxidant peptides. Extensive the most desired period to produce peptides with a high potency hydrolysis may result in the formation of shorter peptides that of ACE-inhibitory activity. This is in parallel with the results shown could hinder its ability as an electron donor, as supported by the in DPPH and FRAP (Sections 3.3 and 3.4). study of the hemp (Cannabis sativa L.) hydrolysates which obtained its highest reducing power at 2 h using the same protease (Tang, Wang, & Yang, 2009). A 0.5–1 h of incubation period may not be 3.6. Fractionation of the bioactive peptides enough for a hydrolysis process to generate the potent antioxidant peptides, resulting in the samples with lower FRAP values. It has The protein hydrolysates which possess the highest DPPHsc, been reported that extensive hydrolysis process would lead to FRAP and %ACE-inhibitory activities were further subjected to frac- the formation of low molecular weight fractions which might de- tionation. Based on the previous results, samples with a S/E ratio of crease the reducing power of the sample and loss the property as 50 incubated at 50 °C for 2 h was selected for fractionation with a an antioxidant (Bamdad et al., 2011). This could explain the de- total of four peptide fractions obtained: <10, 10–30, 30–50 and crease in reducing power observed in the sample hydrolysed at a >50 kDa peptide fraction. incubation temperature of 50 °C for 5 h with a S/E ratio of 50. This Table 3 shows that the fractions had an decreasing activity in a result is supported by the SDS–PAGE profile (Section 3.2), where a similar order of <10, 30–50, >50 and 10–30 kDa. The peptide frac- 5 h of hydrolysis was performed resulting in the most peptide tion of <10 kDa had significantly highest (P < 0.05) %DPPHsc activ- bands been hydrolysed (Fig 1c and d). ity (67.4%), reducing power (1.6 mM) and %ACE-inhibitory activity (35.9%) in comparison with other peptide fractions. Other research 3.5. Angiotensin-I-converting enzyme (ACE) inhibitory activity reported that smaller and low-molecular weight peptides usually exhibit higher antioxidant and ACE-inhibitory activities than that Table 2c reveals that the protein hydrolysates have the ability to of high-molecular weight peptides (Zhao et al., 2007). The result inhibit the ACE in the range of 50.6–80.2% (after 5 dilution), also indicated that the peptide fraction of 30–50 kDa exhibited rel- whereas the non-hydrolysed samples did not show any ACE-inhib- atively high bio-functional activities. This suggests that the peptide itory activities. This result suggested that biological active form of peptides with specific amino acid sequence were produced via enzymatic hydrolysis (Korhonen & Pihlanto, 2006). Whereas, the Table 3 parent chain protein could not contribute to ACE inhibition activity Antioxidant and ACE-inhibitory activity of fractionated P. speciosa seeds-derived due to the long and bulky structure. peptides by ultrafiltration membrane filters. Temperature also played an important role in improving the Peptide DPPHsc activity FRAP ACE-inhibitory activity ACE-inhibitory activity. Results showed that sample produced at fraction (%) (mM) (%) an incubation temperature of 25 °C with a S/E ratio of 20, gave a <10 kDa 67.4 ± 8.2c 1.6 ± 0.0c 35.9 ± 0.7c lower inhibitory activity ranging from 51.7% to 57.5%. Bioactive 10–30 kDa 6.4 ± 1.0a 0.1 ± 0.0a 11.1 ± 0.7a b b c peptide produced at a higher temperature (50 °C), resulted in an 30–50 kDa 34.9 ± 3.3 0.5 ± 0.0 32.6 ± 0.4 >50 kDa 8.7 ± 0.6a 0.1 ± 0.0a 17.8 ± 0.6a increase of inhibitory activities, particularly with an incubation period of 1–2 h, where significantly (P < 0.05) higher inhibitory Values are expressed as means ± sd. a–c activities were observed of 64.8% and 65.6%, respectively. Similar Means the column by different letters are significantly different (P < 0.05). H.-L. Siow, C.-Y. Gan / Food Chemistry 141 (2013) 3435–3442 3441

Fig. 2. Example of MSMS spectra from peptide fraction of <10 kDa. Panel (a) shows MSMS spectra whereas panel (b) shows the ion tables, standard error and MSMS fragments for the peptide sequence.

fraction of 30–50 kDa may consist of potent bioactive peptides se- Table 4 quences even though it is larger in size. Identification of peptide sequences from peptide fraction of <10 kDa. Peptide sequence (<10 kDa) Molecular mass (Da) 3.7. Identification of bioactive peptides Ala-Ser-Pro-Ala-Pro-Ala-Gly-Leu-Ser-Tyr 467.2619 Cys-Val-Pro-Glu-Val-Asp-Pro-Leu-Leu-Leu-Arg-Ala-Glu 829.4482 Amino acid sequence of the fractionated peptides (<10 kDa) Asp-Lys-Leu-Asp-Leu-Ser-Asp-Leu-Ala-His-Glu-Tyr 716.3675 that possess the highest bioactivities, were further identified using Glu-Ala-Ala-Val-Tyr-Pro-Leu 381.7134 an advanced mass spectrometry approach. Glu-Tyr-Leu-Ala-Arg-Lys-Gly-Gly-Val-Glu-Val-Asn 667.8621 Fig. 2a displays the representative MSMS spectra from peptide Glu-Tyr-Asn-Asn-His-Gly-Ser-Phe-Arg-Glu-Ala-His-Pro- 843.3735 fraction of <10 kDa with a high resolution of 60,000 that is Gly-Ala Phe-Asp-Gly-Leu-Pro-Ala-Lys-Pro-Ala-Lys-Lys-Glu-Glu- 695.8508 essential for the identification and confirmation of the molecular Thr-Arg formula of the peptide sequences and thus determines the mass Phe-Leu-Asp-Ala-Leu-Leu-Lys-Thr-Val-Asp-Cys-Asn 757.8835 accuracy of the molecular ions. Ion peaks with doubly or higher Gly-Gly-Asn-Tyr-Val-Gly-Val-Pro 381.7137 charges were then selected to be fragmented. Fig. 2b shows the rep- Gly-Thr-Val-Phe-Glu-Lys-Lys-Ala-Leu-Pro-Gly-Pro-Met 695.8483 resentative MSMS fragments and the ion tables. It was observed that His-Asp-His-Ala-Glu-Arg-Phe-Ser-Gly-His-Asp-Asn-Glu- 833.3802 Asp the sequence of VLNSNAAPLPN could be derived from the fragment Leu-Ala-Arg-Asn-Ala-Thr-Pro-Pro-Arg 498.3045 spectra, which has high signal-to-noise ratio and show complete or Leu-Leu-Arg-Ser-Val-Gly-Val-Pro-Leu 477.2616 near-complete backbone fragmentation as well as including a low Met-Leu-Glu-Leu-Asp-Pro-Leu-Asn-Asn-Leu-Pro-Arg- 778.4091 error (<0.8 Da) in the data. All other MSMS spectra were of similar Met Met-Asn-Gly-Glu-Lys-Lys-Lys-Ala-Pro-Leu-Gly-Asp-Glu 708.8801 high quality and subjected to peptide sequence identification. Met-Ser-Glu-Gly-Ser-Gly-Ala-Ala-Asp-Val-Glu-Gly-Pro- 647.3242 Table 4 shows that there were several peptide sequences Ser identified from peptide fraction of <10 kDa as the most potent con- Met-Tyr-Val-Val-Glu-Leu-Met-Met-Ser-Phe 633.324 tributors for antioxidant and antihypertensive properties. A total of Asn-Gly-Pro-Leu-Leu-Pro-Ser-Asp-Glu-Met-Asp-Arg-Ala 707.8572 29 amino acid sequences from the peptide fraction were observed. Asn-Met-Ala-Val-Pro-Asp-Ala-Leu-Ala-Gly-Pro-Leu 584.8326 Asn-Met-Gly-Pro-Leu-Leu-Pro-Gln 435.2648 Results showed that the sequences contained 7–16 amino acid Asn-Val-Asn-Tyr-Val-Glu-Leu-Leu-Val-Ala-Pro-Asn-Val- 800.4105 residuals per peptide. It is further supported by the research that Asp-Leu bioactive peptides are short peptides, which usually consist of 2– Pro-Thr-Gly-Pro-Asn-Asn-Gly-Leu-Ser-Tyr 467.2621 20 amino acid residues per peptide (Korhonen & Pihlanto, 2006). Gln-Thr-Asp-Ser-His-Gly-Asn-Asp-Arg-Leu-Leu-Leu-Thr 735.398 Ser-Met-Leu-Gly-Asp-Gly-Thr-Leu-Pro-Pro-Asp-Leu-Ala 643.8059 Most of the peptide sequences from peptides fraction of Val-Cys-Asp-Ala-Asp-Pro-Ser-Ser-Pro-Val-Pro-Glu-Ala 643.8093 <10 kDa contained a high amount of hydrophobic amino acids Val-Leu-Glu-Glu-Ser-Lys-Gly-Ala-Pro-Lys-Pro-Asn 555.2952 including Gly, Val, Ala, Pro and Leu. The presence of these hydro- Val-Leu-Asn-Ser-Asn-Ala-Ala-Pro-Leu-Pro-Asn 555.2945 phobic amino acids in the peptide sequence is suggested to Trp-Gly-Leu-Glu-Thr-Ala-Gly-Ser-Val-Val-Leu-Leu-Asp- 829.4472 improve the solubility of peptides in lipid. This could help to Leu-Gln Trp-Leu-Leu-Tyr-Leu-Asp-Gly-Pro-Pro-Met-Gly-Leu-Thr 738.4321 enhance lipid inhibitory activity by facilitating the interaction 3442 H.-L. Siow, C.-Y. Gan / Food Chemistry 141 (2013) 3435–3442 between peptides and radical species. The hydrophobic amino acid Importance of the COOH-terminal dipeptide sequence. Journal of Biological could also help to exhibit higher antihypertensive potential Chemistry, 255, 401–407. Cushman, D. W., & Cheung, H. S. (1971). Spectrophotometric assay and properties of (Cheung, Wang, Ondetti, Sabo, & Cushman, 1980). the angiotensin-converting enzyme of rabbit lung. Biochemical Pharmacology, The peptide sequences were also found to possess a high 20, 1637–1648. amount of repeating amino acid peptides. For example, repeating Duranti, M. (2006). Grain legume proteins and nutraceutical properties. Fitoterapia, 77, 67–82. dipeptides that constitute the same amino acid residuals such as Hartmann, R., & Meisel, H. (2007). Food-derived peptides with biological activity: Ala-Ala; Leu-Leu and Val-Val are frequently found as the compo- From research to food applications. Current Opinion in Biotechnology, 18, nent in the peptide sequence. According to Kawashima, Itoh, Miyo- 163–169. He, H. L., Chen, X. L., Wu, H., Sun, C. Y., Zhang, Y. Z., & Zhou, B. C. (2007). High shi, and Chibata (1979), several di- and tri-peptides have exhibited throughput and rapid screening of marine protein hydrolysates enriched in better biological activity if compared to their constituent amino peptides with angiotensin-I-converting enzyme inhibitory activity by capillary acids. It was also reported that amino acid residues in di- and electrophoresis. Bioresource Technology, 98, 3499–3505. Inouye, K., Nakano, K., Asaoka, K., & Yasukawa, K. (2009). Effects of thermal tri-peptides, could be absorbed more rapidly than free amino acid treatment on the coagulation of soy proteins induced by subtilisin Carlsberg. (Silk et al., 1980). It is therefore possible that the antioxidant and Journal of Agricultural and Food Chemistry, 57, 717–723. antihypertensive activity of the protein hydrolysates may be re- Kawashima, K., Itoh, H., Miyoshi, M., & Chibata, I. (1979). Antioxidant properties of lated to the abundance of these repeating peptides. branched-chain amino acid derivatives. Chemical and Pharmaceutical Bulletin, 27, 1912–1916. The positioning of amino acid residue at the N- or C-terminal of Kim, S. K., Byun, H. G., Park, P. J., & Shahidi, F. (2001). Angiotensin I converting the peptide sequences is also a crucial factor in determining its bio- enzyme inhibitory peptides purified from bovine skin gelatin hydrolysate. activity. Cheung et al. (1980) had reported that the presence of Journal of Agricultural and Food Chemistry, 49, 2992–2997. Korhonen, H., & Pihlanto, A. (2006). Bioactive peptides: Production and amino acid Pro, Tyr and Phe at the C-terminal peptide sequence functionality. International Dairy Journal, 16, 945–960. is highly favoured for ACE inhibition. Also, Val and Ile were shown Larré, C., Chiarello, M., Dudek, S., Chenu, M., & Gueguen, J. (1993). Action of as the most effective for enhancing the binding between peptides transglutaminase on the constitutive polypeptides of pea legumin. Journal of Agriculture and Food Chemistry, 41, 1816–1820. and ACE. Hence, peptides that have these structural features are Lee, J. R., Kwon, D. Y., Shin, H. K., & Yang, C. B. (1999). Purification and identification suggested to contribute to the bioactivity of P. speciosa seeds. of angiotensin-I converting enzyme inhibitory peptide form kidney bean protein hydrolysate. Food Science and Biotechnology, 8, 172–178. Lin, S., Tian, W., Li, H., Cao, J., & Jiang, W. (2012). Improving antioxidant activities of 4. Conclusion whey protein hydrolysates obtained by thermal preheat treatment of pepsin, trypsin, alcalase and flavourzyme. International Journal of Food Science & P. speciosa seed has shown to be an excellent source of antiox- Technology, 47, 2045–2051. Moore, S. (1963). On the determination of cystine as cysteic acid. The Journal of idant and antihypertensive peptides which could be developed as Biological Chemistry, 238, 235–237. a natural functional food ingredient or novel nutraceutical. The Patterson, C. E., & Rhoades, R. A. (1988). Protective role of sulfhydryl reagents in parameters (i.e. temperature, S/E ratio and incubation time) stud- oxidant lung injury. Experimental Lung Research, 14, 1005–1019. Rehman, Z. U., & Shah, W. H. (2005). Thermal heat processing effects on ied showed significant effect onto the production of the bioactive antinutrients, protein and starch digestibility of food legumes. Food Chemistry, peptide. Results showed that at a temperature of 50 °C, a S/E ratio 91, 327–331. of 50 and an incubation time of 2 h, gave the highest biological Sakunpak, A., & Panichayupakaranant, P. (2012). Antibacterial activity of Thai edible plants against gastrointestinal pathogenic bacteria and isolation of a new broad activities. The peptide sequences of the bioactive peptide were also spectrum antibacterial polyisoprenylated benzophenone, chamuangone. Food successfully predicted. Chemistry, 130, 826–831. Shewry, P. R., Jenkins, J. A., Beaudoin, F., & Clara Mills, E. N. (2004). The classification, functions and evolutionary relationships of plant proteins in relation to food Acknowledgement allergies. In E. N. Clare Mills & P. R. Shewry (Eds.), Plant food allergens (pp. 24–41). Oxford, UK: Blackwell Science. This work is financially supported by short-term grant of Uni- Silk, D. B., Fairclough, P. D., Clark, M. L., Hegarty, J. E., Marrs, T. C., Addison, J. M., et al. versiti Sains Malaysia (304/PDOPING/6312045). (1980). Use of a peptide rather than free amino acid nitrogen source in chemically defined ‘‘elemental’’ diets. Journal of Parenteral and Enteral Nutrition, 4, 548–553. References Tang, C. H., Wang, X. S., & Yang, X. Q. (2009). Enzymatic hydrolysis of hemp (Cannabis sativa L.) protein isolate by various proteases and antioxidant properties of the resulting hydrolysates. Food Chemistry, 114, 1484–1490. Aisha, A. F., Abu-Salah, K. M., Alrokayan, S. A., Ismail, Z., & Abdul Majid, A. M. (2012). Tzitzikas, E. N., Vincken, J. P., De Groot, J., Gruppen, H., & Visser, R. G. F. (2005). Evaluation of antiangiogenic and antioxidant properties of Parkia speciosa Hassk Genetic variation in pea seed globulin composition. Journal of Agricultural and extracts. Pakistan Journal of Pharmaceutical Sciences, 25, 7–14. Food Chemistry, 54, 425–433. Ali, H., Houghton, P. J., & Soumyanath, A. (2006). a-Amylase inhibitory activity of Udenigwe, C. C., & Aluko, R. E. (2011). Chemometric analysis of the amino acid some Malaysian plants used to treat diabetes; with particular reference to requirements of antioxidant food protein hydrolysates. International Journal of Phyllanthus amarus. Journal of Ethnopharmacology, 107, 449–455. Molecular Sciences, 12, 3148–3161. Bamdad, F., Wu, J., & Chen, L. (2011). Effects of enzymatic hydrolysis on molecular Udenigwe, C. C., & Aluko, R. E. (2012). Food protein-derived bioactive peptides: structure and antioxidant activity of barley hordein. Journal of Cereal Science, 54, Production, procesing, and potential health benefits. Journal of Food Science, 77, 20–28. R11–R24. Benzie, I. F. F., & Strain, J. J. (1996). The ferric reducing ability of plasma (FRAP) as a Zarkadas, C. G., Gagnon, C., Gleddie, S., Khanizadeh, S., Cober, E. R., & Guillemette, R. measure of ‘‘antioxidant power’’: The FRAP Assay. Analytical Biochemistry, 239, J. D. (2007). Assessment of the protein quality of fourteen soybean [Glycine max 70–76. (L.) Merr.] cultivars using amino acid analysis and two-dimensional Chankhamjon, K., Petsom, A., Sawasdipuksa, N., & Sangvanich, P. (2010). electrophoresis. Food Research International, 40, 129–146. Hemagglutinating activity of proteins from Parkia speciosa seeds. Zhao, Y., Li, B., Liu, Z., Dong, S., Zhao, X., & Zeng, M. (2007). Antihypertensive effect Pharmaceutical Biology, 48, 81–88. and purification of an ACE inhibitory peptide from sea cucumber gelatin Cheung, H. S., Wang, F. L., Ondetti, M. A., Sabo, E. F., & Cushman, D. W. (1980). hydrolysate. Process Biochemistry, 42, 1586–1591. Binding of peptide substrates and inhibitors of angiotensin-converting enzyme.