Vigna Unguiculata as Source of Angiotensin-I Converting Enzyme Inhibitory and Antioxidant Peptides

© 2013 Betancur-Ancona et al., licensee InTech. This is an open access chapter distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Vigna Unguiculata as Source of Angiotensin-I Converting Enzyme Inhibitory and Antioxidant Peptides


Introduction
The frequency of lifestyle-related diseases is steadily increasing, particularly of hypertension, a risk factor for cardiovascular diseases such as coronary heart disease, peripheral arterial disease and stroke. Indeed, cardiovascular diseases are the primary cause of morbidity and mortality in Western countries, with hypertension affecting about 20% of the world's adult population [1]. Blood pressure is controlled by various regulatory factors in the body, including angiotensin I-converting enzyme (ACE-I). ACE-I (peptidyldipeptidaseA, kininase II, EC 3.4.15.1) is a zinc dipeptidylcarboxypeptidase. This membrane-bound exopeptidase is found on the plasma membranes of various cell types, including vascular endothelial cells, microvillar brush border epithelial cells and neuroepithelial cells. It is thought to be physiologically important. The primary activity of ACE-I is to cleave broad specificity free carboxyl group oligopeptides. Substrates containing Pro at the P1' position and Asp or Glu at P2' are resistant to ACE-I. However, on certain substrates ACE-I can also function as an endopeptidase or a tripeptidylcarboxypeptidase. With ACE-I, endopeptidase activity is observed on substrates having amidated carboxyl groups where the enzyme can cleave a C-terminal dipeptide amide and/or a C-terminal tripeptide amide [2]. ACE-I is responsible for converting angiotensin I (Ang I) to the powerful vasoconstrictor angiotensin II (Ang II) and inactivating the vasodilator peptide bradykinin (BK) by removal of C-terminal dipeptides [3]. In a functional sense, therefore, the enzymatic actions of ACE-I potentially cause increased vasoconstriction and decreased vasodilation. ACE-I has attracted interest for development of orally-active ACE-I inhibitors to treat hypertension due to its central role in vasoactive peptide metabolism. Inhibition of ACE-I prevents conversion of Ang I into Ang II, making it becomes one of the most effective methods for suppressing increases in blood pressure [4]. Since discovery of ACE-I inhibitors in snake venom, extensive research has been done on synthesizing ACE-I inhibitors such as captopril, enalapril, alacepril and lisinopril, all currently in use for treatment of hypertension and heart failure in humans. However, these synthetic drugs occasionally produce side effects such as cough, taste alterations and skin rashes. Interest has consequently increased in natural ACE-I inhibitors as safer and lower cost alternatives to synthetic ones [5].
Antioxidant deficiency also has been implicated in the occurrence of hypertension and other degenerative diseases. Reactive oxygen species (ROS) such as the superoxide anion radical (O -2), hydrogen peroxide and hydroxyl radicals (•OH) are physiological metabolites formed as result of respiration in aerobic organisms. ROS are very unstable, and react rapidly with other substances including DNA, membrane lipids and proteins. Oxidative stress is produced by an imbalance between oxidizing species and natural antioxidants in the body, and has been associated with aging, cell apoptosis and severe diseases such as cancer, Parkinson, Alzheimer, and cardiovascular disorders [6]. Epidemiological studies have demonstrated an inverse association among intake of antioxidants from fruits and vegetables, and morbidity and mortality due to coronary heart diseases and cancer. In response, researchers are searching for natural antioxidants in food that may protect the body from free radicals and retard the evolution of many chronic diseases [7].
In recent years, food proteins have gained increasing value due to the rapidly expanding knowledge about physiologically active peptides. Peptides from various dietary sources have been shown to have clearly positive effects on health by functioning as antihypertensives, antioxidants, anticarcinogens, antimicrobials and anticariogenics, among others. These properties have led to their labeling as functional or biologically-active (i.e. bioactive) peptides. Bioactive peptides may be encrypted within the amino acid sequence of a larger protein. These peptides usually consist of 3-20 amino acids and are released from the original protein after degradation [8]. The most common way to produce bioactive peptides is through enzymatic hydrolysis of whole protein molecules. After enzymatic processing, amino acid sequences that were inactive in the core of the source protein are released and can exercise special properties. Many of the known bioactive peptides have been produced using gastrointestinal enzymes, usually pepsin and trypsin. Other digestive enzymes and different enzyme combinations of proteinases-including Alcalase ® , chymotrypsin, pancreatin, pepsin and thermolysin have also been utilized to generate bioactive peptides from various proteins [9].
Continued population growth worldwide, consequent food resource shortages in developing countries, and the health risks associated with excessive animal protein (and saturated fats) intake has led researchers to search for new sources of proteins from nonconventional raw materials. Legumes are cultivated worldwide and constitute an excellent protein source (protein content =20-30%). Cowpea (Vigna unguiculata) is a major legume crop worldwide, particularly in tropical and subtropical areas such as southeast Mexico. It serves as a major dietary protein source in both human and animal diets, and its protein content makes it a good raw material for preparation of protein extracts and hydrolysates. Protein extract has been used as substrates for production of hydrolysates with functional and/or nutritional properties better than the original extract [10]. Bioactive peptides with ACE-I inhibitory activity has been isolated from protein hydrolysates from a number of animal and vegetable sources. Vegetable origin proteins are of particular interest, and legumes are especially promising due to their high protein content and diverse physiological activities in the human organism. Extensive hydrolysis of V. unguiculata protein concentrates with commercial and digestives enzymes could therefore produce a number of peptides with a myriad of potential applications; for example, as natural-source therapeutic agents in medical treatments and/or as an ingredient in functional foods. Taking this into account, the aim of the present study was to modify enzymatically protein concentrates of V. unguiculata, evaluate the ACE-I inhibitory and antioxidant potential of the hydrolysates and relate the biological activity to their amino acid compositions.

Protein isolates
A single extraction was performed with 6 kg of cowpea seeds. Impurities and damaged seeds were removed, and sound seeds milled in a Mykros impact mill (Industrial Machinery, Monterrey, Mexico) until passing through a 20-mesh screen (0.85 mm) followed by milling in a Cyclotec 1093 (Tecator, Sweden) mill until passing through a 60-mesh screen (0.24 mm). The resulting flour was processed using the wet fractionation method of Betancur-Ancona [11]. Briefly, whole flour was suspended in distilled water at a 1:6 (w/v) ratio, pH was adjusted to 11.0 with 1 M NaOH, and the dispersion was stirred for 1 h at 0.178 x g with a mechanical agitator (Caframo Rz-1, Heidolph Schwabach, Germany). This suspension was wet-milled with a Kitchen-Aid ® food processor, and the fiber solids were separated from the starch and protein mix by straining through 80-and 150-mesh sieves and washing the residue five times with distilled water. The protein-starch suspension was allowed to sediment for 30 min at room temperature to recover the starch and protein fractions. The pH of the separated solubilized proteins was adjusted to the isoelectric point (4.5) with 1 N HCl. The suspension was then centrifuged at 1317 x g for 12 min (Mistral 3000i, Curtin Matheson Sci.), the supernatants were discarded, and the precipitates were freeze-dried until use.

Enzymatic hydrolysis
Hydrolysis of the protein extract was done using a totally randomized design with the treatments being the enzymatic system applied: Alcalase ® 2.4L FG; Flavourzyme ® 500MG; or a sequential system using pepsin from porcine gastric mucosa (Sigma, P7000-100G) and pancreatin from porcine pancreas (Sigma, P3292-100G). The response variable was degree of hydrolysis (DH).
Hydrolysis was done under controlled conditions (temperature, pH and stirring) in a 1000 mL reaction vessel equipped with a stirrer, thermometer and a pH electrode. Hydrolysis with Alcalase ® and Flavourzyme ® was done according to Pedroche et al. [12]. Protein extracts were suspended in distilled water to produce a 4% (w/v) protein solution. This solution was equilibrated at optimum temperature and pH for each protease before adding the respective enzyme. Protease was then added to the solution at a ratio of 0.3 UA/g for Alcalase ® and 50 UAPL/g for Flavourzyme ® . Hydrolysis conditions were 90 min at 50°C for both enzymes, and pH 8.0 for Alcalase ® and pH 7.0 for Flavourzyme ® . The pH was kept constant by adding 1.0 M NaOH during hydrolysis. Hydrolysis with the sequential pepsinpancreatin system was done with a pH-stat method for 90 min: pre-digestion with pepsin for 45 min followed by incubation with pancreatin for 45 min. Hydrolysis parameters were substrate concentration 4%; enzyme/substrate ratio 1:10; pH 2 for pepsin; pH 7.5 for pancreatin; and 37°C [13,14]. In all three treatments the reaction was stopped by heating to 80°C for 20 min, followed by centrifugation at 9,880 x g for 20 min to remove the insoluble portion.

Degree of hydrolysis
DH was calculated by calculating free amino groups with o-phthaldialdehyde [15]: DH= h/htot *100, where htot is the total number of peptide bonds per protein equivalent, and h is the number of hydrolyzed bonds. The htot factor is dependent on raw material amino acid composition.

Hydrolysate fractionation by ultrafiltration
Following Cho et al. [16], the hydrolysate was fractionated by ultrafiltration using a high performance ultrafiltration cell (Model 2000, Millipore, Inc., Marlborough, MA, USA). Five fractions were prepared using four molecular weight cut-off membranes: 1 kDa, 3 kDa, 5 kDa and 10 kDa. Soluble fractions prepared by centrifugation were passed through the membranes stating with the largest Molecular Weight Cut off (MWCO) membrane cartridge (10 kDa). The retentate and permeate were collected separately, and the retentate recirculated into the feed until maximum permeate yield was reached, as indicated by a decreased permeate flow rate. The permeate from the 10 kDa membrane was then filtered through the 5 kDa membrane with recirculation until maximum permeate yield was reached. The 5 kDa permeate was then processed with the 3 kDa membrane and the 3 kDa permeate with the 1 kDa membrane. This process minimized contamination of the larger molecular weight fractions with smaller molecular weight fractions while producing enough retentates and permeates for the following analyses. The five ultrafiltered peptide fractions were designated as >10 kDa (10 kDa retentate); 5-10 kDa (10 kDa permeate-5 kDa retentate); 3-5 kDa (5 kDa permeate-3 kDa retentate); 1-3 kDa (3 kDa permeate-1 kDa retentate); and <1 kDa (1 kDa permeate).
Vigna Unguiculata as Source of Angiotensin-I Converting Enzyme Inhibitory and Antioxidant Peptides 187

ACE-I inhibitory activity
ACE-I inhibitory activity in the hydrolysate and its purified peptide fractions was analyzed following Hayakari et al. [17]. ACE-I hydrolyzes hippuryl-L-histidyl-L-leucine (HHL) to yield hippuric acid and His-Leu. This method relies on the colorimetric reaction of hippuric acid with 2,4,6-trichloro-s-triazine (TT) in a 0.5 mL incubation mixture containing 40 μmol potassium phosphate buffer (pH 8.3), 300 μmol sodium chloride, 40 μmol 3% HHL in potassium phosphate buffer (pH 8.3) and 100 mU/mL ACE-I. The mixture was incubated at 37ºC for 45 min and then reaction terminated by addition of TT (3% v/v) in dioxane and 3 ml of 0.2 M potassium phosphate buffer (pH 8.3). After centrifuging the reaction mixture at 10,000 x g for 10 min, enzymatic activity was determined in the supernatant by measuring absorbance at 382 nm. All runs were performed in triplicate. ACE-I inhibitory activity was quantified by a regression analysis of ACE-I inhibitory activity (%) versus peptide concentration and defined as an IC50 value, that is, the peptide concentration in (μg protein/mL) required to produce 50% ACE-I inhibition under the described conditions.

Antioxidant activity by ABTS assay
2,2'azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radical cation (ABTS •+ ) was produced by reacting ABTS with potassium persulfate following Pukalskas et al. [18]. To prepare the stock solution, ABTS was dissolved at a 2 mM concentration in 50 mL phosphatebuffered saline (PBS) prepared from 4.0908 g NaCl, 0.1347 g KH2PO4, 0.7098 g Na2HPO4, and 0.0749 g KCl dissolved in 500 mL ultrapure water. If pH was lower than 7.4, it was adjusted with NaOH. A 70 mM K2S4O8 solution in ultrapure water was prepared. ABTS •+ radical was produced by reacting 10 mL of ABTS stock solution with 40 L K2S4O8 solution and allowing the mixture to stand in darkness at room temperature for 16-17 h before use. The radical was stable in this form for more than 2 days when stored in darkness at room temperature.
Antioxidant compound content in the hydrolysates and their UF peptide fractions was analyzed by diluting the ABTS •+ solution with PBS to an absorbance of 0.800 ± 0.030 AU at 734 nm. After adding 990 L of diluted ABTS •+ solution (A734 nm= 0.800 ± 0.030) to 10 L antioxidant compound or Trolox standard (final concentration 0.5-3.5 mM) in PBS, absorbance was read at ambient temperature exactly 6 min after initial mixing. All analyses were run in triplicate. The percentage decrease in absorbance at 734 nm was calculated and plotted as a function of the concentration of Trolox for the standard reference data. The radical scavenging activity of the tasted samples, expressed as inhibition percentage, was calculated by the following formula: Where AB was the absorbance of the blank sample (t=0), and AA was the absorbance of sample with antioxidant after 6 min.
Trolox equivalent antioxidant coefficient (TEAC) was quantified by a regression analysis of % Inhibition versus Trolox concentration using the following formula: Where b was the intersection and m was the slope.

G-50 gel filtration chromatography
After filtration through 10, 5, 3 and 1 kDa membranes in a high performance ultrafiltration cell, 10 mL of the fraction with highest ACE-I inhibitory activity was injected into a Sephadex G-50 gel filtration column (3 cm x 79 cm) at a flow rate of 25 mL/h of 50 mM ammonium bicarbonate (pH 9.1). The resulting fractions were collected for to assay ACE-I inhibitory activity. Peptide molecular masses were determined by reference to a calibration curve created by running molecular mass markers on the Sephadex G-50 under conditions identical to those used for the test samples. Molecular mass standards were thyroglobulin (670 kDa), bovine gamma globulin (158 kDa), equine myoglobin (17 kDa), vitamin B12 (1.35 kDa) and Thr-Gln (0.25 kDa). Fractions selected for further purification of peptides were pooled and lyophilized before RP-HPLC.

HPLC C18 chromatography
The fractions isolated with the Sephadex G-50 column were redissolved in deionized water and injected into a preparative HPLC (Agilent, Model 1110, Agilent Technologies, Inc. Santa Clara, CA, USA) reverse-phasecolumn (C18 Hi-Pore RP-318, 250 mm x10 mm, Bio-Rad). The injection volume was 100 L, and the sample concentration was 20 mg/mL. Elution was achieved by a linear gradient of acetonitrile in water (0-30% in 50 min) containing 0.1% trifluoroacetic acid at a flow rate of 4 mL/min and 30°C [13]. Elution was monitored at 215 nm, and the resulting fractions were collected for assay of ACE-I inhibitory activity as described above.

Amino acid composition
Protein amino acid composition was determined for the hydrolysate, and the peptides were purified by ultrafiltration, gel filtration chromatography and HPLC [19]. Samples (2-4 mg protein) were treated with 4 mL of 6 mol equi/L HCl, placed in hydrolysis tubes and gassed with nitrogen at 110°C for 24 h. They were then dried in a rotavapor and suspended in 1 mol/L sodium borate buffer at pH 9.0. Amino acid derivatization was performed at 50°C using diethyl ethoxymethylenemalonate. Amino acids were separated using HPLC with a reversed-phase column (300 x 3.9 mm, Nova Pack C18, 4 mm; Waters), and a binary gradient system with 25 mmol/L sodium acetate containing (A) 0.02 g/L sodium azide at pH 6.0, and (B) acetonitrile as solvent. The flow rate was 0.9 mL/min, and the elution gradient

Statistical analysis
All results were analyzed in triplicate using descriptive statistics with a central tendency and dispersion measures. One-way ANOVAs were performed to evaluate protein isolate hydrolysis data and in vitro ACE-I inhibitory activity. A Duncan's multiple range test was used to determine differences between treatments. All analyses were performed according to Montgomery [20] and processed using the Statgraphics Plus version 5.1 software.
The AH had a 53.0% DH, which is lower than reported by Vioque et al. [21] for rapeseed protein hydrolysates (60% DH) produced with a mixture of Alcalase ® and Flavourzyme ® during 180 min. However, this DH was higher than that reported for mung bean protein hydrolysates (20.0% DH) produced with Alcalase ® for 10 h [22] and for V. unguiculata hydrolysates (32.3% DH) produced with Alcalase ® for 60 min [10].The variation in DH observed here is probably the result of protease specificity since Alcalase ® is an industrial alkaline protease produced from Bacillus licheniformis, the main enzyme component of which (serine endopeptidase subtilisin Carlsberg) presents broad specificity and hydrolyzes most peptide bonds, with a preference for those containing aromatic amino acid residues [23]. ACE-I prefers substrates or competitive inhibitors containing hydrophobic amino acid (aromatic or branched lateral chain) residues (Hong et al., 2005). Alcalase ® is therefore very suitable for production of bioactive peptides, such as those with ACE-I inhibitory activity. According to Pedroche et al. [24] the controlled liberation of biologically active peptides from protein by enzymatic hydrolysis is one of the most promising trends concerning medical applications of the protein hydrolysates with DH higher than 10% while hydrolysates with a low DH (lower than 10%) are used for the improvement of functional properties of flours or protein isolates. Therefore, the results suggest (DH=53.0%) that V. unguiculata protein is an appropriate substrate for producing these bioactive peptides when hydrolyzed with Alcalase ® .
Hydrolysis with Flavourzyme ® produced a V. unguiculata hydrolysate with 58.8% DH, somewhat higher than obtained with the Alcalase ® system. A similar discrepancy has been reported for chickpea protein hydrolysates (27.0% DH) produced with Flavourzyme ® for 180 min [25]. DH was higher with Flavourzyme ® since it is a protease complex produced by Aspergillus orizae, which contains endoproteinases and exopeptidases. The fungal protease complex Flavourzyme ® has a broader specificity; thus, when combined with its exopeptidase activity high DH values can be achieved, perhaps as much as 50% giving mostly dipeptides in the hydrolysate.
Sequential hydrolysis with pepsin-pancreatin produced cowpea protein hydrolysates with the lowest DH (35.74%) of the three studied enzymatic systems. This DH was similar to 37.0% at 360 min reported for sunflower protein hydrolysates obtained with the same system [13], but higher than the reported values for soy protein hydrolysates produced with pancreatin for 60 (11.0%) and 180 min (17.0%) [26]. The V. unguiculata PPH represents a pool of peptides resembling those generated during digestion of V. unguiculata proteins in the organism. This coincides with the behavior of extensively hydrolyzed sunflower protein reported by Megías et al. [13].Pepsin is the main proteolytic enzyme generated in the stomach during food digestion, while pancreatin includes proteases such as trypsin, chymotrypsin and elastase, which are released by the pancreas in the small intestine. The resulting peptides are therefore resistant to pepsin and pancreatin, suggesting that they might be absorbed by digestive epithelial cells in the small intestine, probably might be bioavailable and exercise their biological activity.

ACE-I inhibitory activity
ACE-I inhibitory activity of AH, FH and PPH was measured and calculated as IC50. Biological activity was highest in the PPH, as indicated by a lower IC50 value (1397.9 g/ml) compared to the AH (2564.7 g/ml) and FH (2634.4 g/ml). In other words, more ACE-I inhibitory active peptides were produced using the pepsin-pancreatin treatment.
ACE-I inhibitory activity has been reported for enzymatic hydrolysates from different protein sources with IC50 values ranging from 0.2 to 246.7 g/mL [22]. Many of these hydrolysates have been shown to have antihypertensive activities in spontaneously hypertensive rats (SHR). In the present study, the IC50 value for the hydrolysate prepared with pepsin-pancreatin at 90 min incubation is within the concentration range likely to mediate an antihypertensive effect. Therefore, it is to be expected that V. unguiculata protein derived ACE-I inhibitory peptides would have antihypertensive activity. However, further investigations are necessary to examine whether the peptide mixture may exert antihypertensive activity in vivo because the inhibitory potencies of the peptides on ACE-I activity do not always correlate with their antihypertensive activities found in SHR [22].
Ultrafiltration of AH, FH and PPH produced peptide fractions with increased biological activity (Fig. 1). ACE-I inhibitory activity of peptide fractions ranged from 24.3-123 g/ml in the AH, from 0.04 to 170.6 g/ml in the FH and from 44.7 to 112 g/ml in the PPH. ACE-I inhibitory activity was significantly (P<0.05) dependent on peptide fraction molecular weight, with the lowest activity being in the >10kDa fractions and the highest in the <1kDa fractions for all hydrolysates. Similar ACE-I inhibitory activity behavior was reported by Je et al. [5]for five peptide fractions from pepsin-hydrolyzed Alaska pollack frame protein run through an ultrafiltration membrane-bioreactor system with MWCOs of 30, 10, 5, 3 and 1 kDa. Higher ACE-I inhibitory activity (%) in lower molecular weight fractions was also reported by Xue-Ying et al. [27] for yak casein hydrolysate fractions separated using 6 and 10 kDa MWCOs: >10kDa (23.1%), 6-10kDa (29.2%) and <6 kDa (85.4%).
Of the three hydrolysates tested in the present study, FH, which had the highest DH, also exhibited the highest ACE-I inhibitory activity in the <1kDa fraction. The biological activity of this fraction provides it potential commercial applications as a 'health-enhancing ingredient' in functional food production.
Vigna Unguiculata as Source of Angiotensin-I Converting Enzyme Inhibitory and Antioxidant Peptides 191 In terms of ACE-I inhibitory activity, the pepsin-pancreatin system produced hydrolysate with the highest activity, while the Flavourzyme ® system produced peptide fractions with the highest activity. This also confirms that V. unguiculata is a good protein source for bioactive peptide extraction by gastrointestinal or commercial proteases.

Antioxidant activity by ABTS assay
Antioxidant activity of the protein hydrolysates and their corresponding UF peptide fractions was quantified and calculated as TEAC values (mM/mg protein). Antioxidant activity was not significantly different (P>0.05) between the three hydrolysis systems (14.7 for AH, 14.5 for FH and 14.3 mM/mg protein for PPH). Ultrafiltration improved antioxidant activity, which was dependent (P<0.05) on fraction molecular weight (Fig. 2).TEAC values were 303.2-1457 mM/mg protein for the AH, 357.4-10211 mM/mg protein for the FH, and 267.1-2830.4 mM/mg protein for the PPH. These values, and consequently the fraction's antioxidant activities, were higher (P<0.05) as fraction molecular weight decreased. According to Dávalos et al. [28] this behavior among the V. unguiculata peptide fractions may reflect the enhanced accessibility of small peptides to the redox reaction system, for the prescence of critical amino acid residues.
Antioxidant activity in the V. unguiculata protein hydrolysates and their UF peptide fractions was measured with an ABTS assay, which quantifies an antioxidant's (i.e. hydrogen or electron donor) suppression of the radical cation ABTS •+ based on singleelectron reduction of the relatively stable radical cation ABTS •+ formed previously by an oxidation reaction. When added to PBS medium (pH 7.2) containing ABTS •+ , the proteins in the hydrolysates and peptide fractions very probably acted as electron donors, transforming this radical cation (maximum absorbance at 734nm) into the non-radical ABTS. The higher antioxidant activity of the UF peptide fractions versus their source hydrolysates is related to unique properties provided by their amino acid composition. The fractions' increased ability to decrease free radical reactivity is linked to the greater exposure of their amino acids, which leads to increased peptide/free radical reactions. Overall, the <1 kDa peptide fraction from the FH had the highest TEAC values and was shown to undergo single-electron transfer reactions in the ABTS •+ reduction assay, demonstrating its antioxidant capacity. This is an extremely attractive property since oxidants are known to be involved in many human diseases and aging processes. Oxidants are associated with the chronic damage of ageing, and destructive oxidants and oxygen-free radicals can be extremely toxic to tissues by promoting tissue necrosis and cell damage. Some authors claim that proteins possess antioxidant properties; for instance, [29] reported that protein insufficiency aggravates enhanced lipid peroxidation and reduces antioxidative enzyme activities in rats, while Larson et al. [30] observed that proteins affect lipid metabolism in laboratory animals. The biological effect exhibited by the FH<1kDa fraction apparently reinforces the claim that proteins possess antioxidant properties. This makes the FH<1kDa fraction a potential "antioxidant" ingredient in functional food production.

Gel filtration chromatography
Because it exhibited the highest ACE-I inhibitory (IC50 value of 0.04 g/mL) and antioxidant activity (TEAC value of 10211 mM/mg protein), the <1 kDa fraction from FH was selected for further fractionation. Gel filtration chromatography was used to generate a molecular weight profile of this fraction (Fig. 3).  The profile was typical of a protein hydrolysate formed by a pool of peptides, with gradually decreasing molecular masses. Elution volumes between 406 and 518 mL included free amino acids and peptides with molecular masses ranging from 3.6 to 0.4 kDa. This range was fractionated into eleven fractions (1 to 11) and ACE-I inhibitory activity determined for each. Fractions with elution volumes smaller than 406 mL and greater than 518 mL were not analyzed because they largely included peptides with high molecular weights, as well as free amino acids. ACE-I inhibitory activity (%) in the eleven fractions ranged from 5.29 to 47.43% and differed (P<0.05) between fractions ( Table 1). The highest ACE-I inhibitory activity was observed in fractions F4 (47.43%; 437.5-444.5 mL elution volume) and F5 (45.14%, 448-455 mL elution volume), which were not statistically different (P<0.05). Their molecular masses were approximately 1.8 kDa (indicative of 7 amino acid residues) and 1.5 kDa (indicative of 10 amino acid residues), respectively. The IC50 value for F4 (14.19 g/mL) was similar than those of Fagopirum esculentum peptide fractions purified by Sephadex LH-20 gel filtration (15.1 g/mL) [31], but lower than those of gel filtration (Sephadex G-25) peptide fractions from tuna broth hydrolysate (210 to 25,260 g/mL) [32], from Fagopyrum esculentum Moench (Sephadex C-25 = 25,715.1 g/mL; Sephadex G-10 = 21,315.1 g/mL), and from the peptic hydrolysate of Acetes chinensis (Sephadex C-15 = 770 -1590 g/mL) [33]. A similar behavior pattern was observed with Konjac peptides purified by series connection of Sephadex G-25 and Sephadex G-15 columns that resulted in purified peptides with molecular weights of 1500 and 1000 Da and IC50 values of 120 g/mL and 88 g/mL, respectively [34]. Ji-Eun et al. [35] reported similar results to separate peptide fractions below 3 kDa through size exclusion chromatography (IC50= 500 g/mL) from textured and fermented vegetable protein (IC50 = 2190 g/mL). They purified peptide fraction with a molecular weight range of 500-999 Da with IC50 values of 94 g/mL and that represented peptides of approximately 7 amino acids residues.

Reverse-phase HPLC chromatography of pooled fractions
Fractions F4 and F5 from the gel filtration chromatography treatment were pooled and analyzed using RP-HPLC to produce a chromatographic profile from mass-transfer between stationary and mobile phases. Mixture components were separated by dissolving fractions F4 and F5 in acetonitrile and forcing them through a chromatographic column under high pressure. The mixture resolved into its components in the column, separating F4 and F5 based on differences in hydrophobicity. In this process, the components of both fractions passed over stationary-phase particles containing pores large enough for them to enter, and in which interactions with the hydrophobic surface removed them from the flowing mobilephase stream. The strength and nature of the interaction between the sample particles and stationary phase depended on hydrophobic and polar interactions. As the eluent organic solvent concentration increased, it reached a critical value for each analyte and desorbed it from the hydrophobic stationary-phase surface to allowing it to elute from the column into the flowing mobile phase. Because this elution depended on the precise distribution of hydrophobic residues in each specie, each analyte eluted from the column at a characteristic time and the resulting peaks were used to qualitatively analyze both fractions' components. Within each gel filtration fraction, the eluates were divided into four major fractions: F4-1, F4-2, F4-3, F4-4; F5-1, F5-2, F5-3, F5-4. The peptides were relatively pure, although a small shoulder still appeared behind the peaks in the chromatogram.
Enough material from each fraction was collected in successive analyses to determine ACE-I inhibitory activity (Fig. 4)
Compared to the <1 kDa fraction amino acid profile, the G-50 gel filtration chromatography fractions had higher Asx, Glx and Arg concentrations. Hydrophobic amino acid content decreased by 20.27% in F4 (34.56g/100g) and 21.19% in F5 (34.16g/100g) compared to the <1 kDa fraction (43.35 g/100g), while hydrophilic residues increased by 19.25% (48 g/100g) in F4 and 22.48% (50 g/100g) in F5. The ACE-I inhibitory activity observed in F4 (47.43%) and F5 (45.14%) could therefore be the result of their higher Arg, Asx or Asp concentrations. These residues are known to play an important role in the antihypertensive activity of peptides from white and red wines and from French flor-sherry wine [38]. The F4 and F5 fractions may also have the added benefit of low bitterness. Many of the ACE-I-inhibitory peptides isolated from food sources are composed of multiple food components and hydrophobic and/or aromatic amino acid residues. However, practical use of these food protein hydrolysates is complicated by formation of peptides, which impart a bitter taste, the result of the formation of low molecular weight peptides containing mostly hydrophobic amino acids. To address this problem, Kim et al. [40] recommended use of Flavourzyme ® , a fungal endoprotease and exo-protease complex that produces hydrolysates or peptides with ACE-I inhibitory activity and low bitterness. Fractions F4 and F5 are promising prospects for use in new product development because they had clear ACE-I inhibitory activity and low hydrophobic amino acid content, which may ensure that they have low bitterness.  Comparison of amino acid composition and properties between the cowpea V. unguiculata hydrolysate and its peptides isolated by ultrafiltration and G-50 gel chromatography showed the most active fraction to be the <1 kDa fraction, which had abundant aromatic (e.g., Phe) and cyclic amino acids (e.g., Pro). Amino acids such as Phe, with large bulky chains and hydrophobic side chains, are preferred in both positions of a dipeptide for their high steric properties and low lipophilicity, while amino acids such as Pro are preferred in the carboxyl terminus of active tripeptides for their low lipophilicity, and high steric and electronic properties [41]. Aromatic and basic amino acids are important to the ACE-I inhibitory activity of peptides. For instance, similar observations have been made for an orientase hydrolysate with notable ACE-I inhibition attributed to its high basic and aromatic amino acids contents [33]. Cheung et al. [43] reported a peptide with strong, competitive ACE-I inhibition in which aromatic amino acid residues at its C-terminal and basic or hydrophobic ones at its N-terminal played an essential role. Based on the above, the higher ACE-I inhibitory potential of the <1 kDa fraction from the cowpea hydrolysate can be attributed to the steric properties of its aromatic amino acids and the lipophilicity and electronic properties of its cyclic amino acids.
F4-2 was also unique in containing Met (0.5 g/100g) and Trp (0.2 g/100g), as well as higher concentrations of Ile (3.2 g/100g) and Leu (10.2 g/100g) than the other fractions eluted from F4, all of which could have significantly increased its relative ACE-I inhibition activity. This would coincide with previous reports of inhibitory activity in peptides with these amino acids. The residues Val, Leu and Ile are preferred in the amino terminal position in active tripeptides for their low lipophilicity and steric properties or side chain bulk/ molecular size [40]. Clearly, the structure of ACE-I inhibitory peptides influences their activity, as shown by Cheung et al. [42], who reported that peptides with Tyr at the C-terminus and Ile at the N-terminus exhibit highly potent inhibitory activity. In a randomized, double-blind, placebo-controlled human study, a significant depressor effect was observed in mild essential hypertensive volunteers and Val-Tyr was shown to be one of the predominant ACE-I inhibitory peptides involved in this effect [44]. In addition to its high Tyr content, F4-2 may also be absorbed intact into the human circulatory system and induce a reduction in blood pressure. For instance, intravenous and oral administration of Val-Tyr in SHR have shown that this di-peptide caused a long-lasting depressor effect. Val-Tyr is known to be absorbed intact into the human circulatory system, and studies using cross-mated transgenic mice carrying the human renin gene and the human angiotensinogen gene have shown that, as a natural ACE-I inhibitory dipeptide, Val-Tyr regulates the enhanced human reninangiotensin system and induces a prolonged reduction in blood pressure [45].
Of the fractions purified from F5 by preparative RP-HPLC chromatography (Table 3), F5-2, F5-3 and F5-4 exhibited ACE-I inhibition that was not different among them (P<0.05), but was greater than that of F5-1. The amino acids Arg, Tyr, Met, Ile, Leu, Phe and Lys very probably played a key role in providing greater activity to the first three fractions. As mentioned above, residues such as Leu and Ile are preferred at the amino terminus of tripeptides with ACE-I inhibitory activity due to their low lipophilicity and steric properties or side chain bulk/molecular size values. In addition, Lys and Arg are expected in positions adjacent to the amino terminus due to their low electronic properties and high lipophilicity and steric property values, while residues such as Phe are preferred at the carboxyl terminus due to their low lipophilicity and high steric and electronic property values [41].
Although the precise substrate specificity is not fully understood, ACE-I appears to prefer substrates containing hydrophobic amino acid residues at the three C-terminal positions, suggesting that the higher hydrophobic amino acid content in F5-2 (9.1 g/100g), F5-3 (89.8 g/100g) and F5-4 (34.2 g/100g) versus F5-1 (1.1 g/100g) probably made a substantial contribution to their inhibitory potency. This agrees with Wu et al. [40], who state that aromatic, positively-charged and hydrophobic amino acids are preferred in active tripeptides. Due to substrate specificity differences between the two ACE-I catalytic sites, ACE-I inhibitors may inhibit only one site. Moskowitz [46] proposed a model explaining the clinical superiority of hydrophobic ACE-I inhibitory drugs relative to hydrophilic ones: all ACE-I inhibitors bind to the C-terminal catalytic site, but only hydrophobic ones bind to the occluded N-terminal catalytic site and are therefore better at blocking Ang II production. This would also explain why hydrophobic ACE-I inhibitors have specific local benefits such as organ damage prevention, in addition to reducing blood pressure [46]. The high hydrophobic amino acid (particularly aromatic side-chains) content in the F5 may therefore make a substantial contribution to its fractions' ACE inhibitory activity by blocking Ang II production.
Overall, the highest in vitro ACE-I inhibitory activity (IC50) among the cowpea hydrolysate and its derivative fractions was present in the ultrafiltered <1 kDa fraction (0.04 g/mL), followed by the RP-HPLC F4-2 fraction (0.4704 g/mL), the gel filtration chromatography fraction F4 (14.195 g/mL) and finally the hydrolysate (2634.4 g/mL). Although the IC50 values for the hydrolysate, F4 and F4-2 fractions were substantially higher than that of the F<1 kDa fraction, they are still in the same order of magnitude as values reported for many other natural ACE-I inhibitory peptides. Nevertheless, the IC50 values for all the studied V. unguiculata derivatives are far higher than that of the synthetic ACE-I inhibitor Captopril ® (0.0013 g/mL) [32]. The biological potential in the peptides purified from V. unguiculata, and the high ACE-I inhibitory activity in the F<1 kDa fraction, reinforce the need for ACE-I inhibitory peptides to be rich in hydrophobic amino acids (aromatic or branched chains) and peptides rich in Pro. Pro is well-documented as the most favorable amino acid for ACE-I binding (most commercial inhibitors include this residue), but it was not present in the peptides with potential biological activity that had been purified by G-50 gel filtration chromatography and RP-HPLC.
The studied cowpea V. unguiculata protein hydrolysate has potential applications in the development of physiologically functional foods aimed at preventing and/or treating hypertension. An added benefit is the balanced amino acid profile of the protein hydrolysate and its peptide fractions, which makes them an appropriate protein source in human nutrition. It should be considered, however, that the in vitro ACE-I inhibitory potencies of peptides do not always correlate with their in vivo antihypertensive activities as quantified in SHR. This is because they must be absorbed and transported intact from the intestine to the blood stream (in the case of oral administration) and resist plasma peptidase degradation (in the case of oral and intravenous administration) to reach their target sites and exert an antihypertensive effect in vivo. Therefore, in vivo research is needed to determine to what extent any of the studied ACE-I inhibitory peptides can exercise their antihypertensive activity in vivo. 2.4 ± 0.073 Trp 0 ± 0 0.2 ± 0.040 0 ± 0 0 ± 0 0 ± 0 0 ± 0 0 ± 0 0 ± 0 Table 3. Amino acid contents of fractions from the F4 and F5 gel filtration chromatography fraction purified by preparative RP-HPLC C-18 chromatography.

Conclusions
After modification by Alcalase ® , Flavourzyme ® and pepsin-pancreatin cowpea V. unguiculata proteins proved to be a source of bioactive peptides with ACE-I inhibitory and antioxidant activity. Fractionation of V. unguiculata enzymatic hydrolysates by ultrafiltration enhanced their ACE-I inhibitory and antioxidant activity in all the resulting peptides, although the <1 kDa fraction of the Flavourzyme hydrolysate had the highest overall biological activity. Further purification of this fraction by gel filtration chromatography and RP-HPLC produced fractions with different activities, all of which were much higher than the source hydrolysate. Separation of protein hydrolysates by molecular weight and hydrophobicity clearly enhanced peptide ACE-I inhibitory activity, particularly purification by ultrafiltration and chromatography. The highest biological potential among the purified peptides was observed in the ultrafiltered <1 kDa fraction, which supports the importance of high hydrophobic amino acid and proline content in ACE-I inhibitory peptides. These results highlight the promise of controlled protein hydrolysis with a fungal protease complex to isolate bioactive peptides from cowpea V. unguiculata proteins, which can then be further purified and/or used as an ingredient in functional foods designed for specific diets.

Author details
Maira R.