Lima Bean ( Phaseolus lunatus ) Protein Hydrolysates with ACEI Inhibitory Activity

Several protein sources can be used to produce bioactive peptides with angiotensin I-converting enzyme (ACE) inhibittory activity. Protein concentrates from ungerminated and germinated lima bean Phaseolus lunatus seed flours were hydrolyzed with Alcalase 2.4 L or pepsin-pancreatin sequential hydrolysis, and ACE inhibitory activity measured in the different hydrolysis treatments. Protein hydrolysate production was analyzed with a 2 factorial design with four replicates of the central treatment. Evaluated factors were protein concentrate source (ungerminated seeds, PC1; germinated seeds, PC2), enzyme/substrate ratio E/S (1/50 or 1/10) and hydrolysis time (0.5 or 2.0 h for Alcalase; 1 or 3 h for pepsin-pancreatin). Degree of hydrolysis (DH) was high for the Alcalase hydrolysates (24.12% 58.94%), but the pepsin-pancreatin hydrolysates exhibited the highest ACE inhibitory activity (IC50 = 0.250 0.692 mg/mL). Under the tested conditions, the hydrolysates with the highest ACE inhibitory activity were produced with sequential pepsin-pancreatin using either PC1 at 1 h hydrolysis time and a 1/10 E/S ratio or PC2 at 1 h hydrolysis time and a 1/50 E/S ratio. Lima bean protein hydrolysates prepared with Alcalase or pepsin-pancreatin are a potential ingredient in the production of physiologically functional foods with antihypertensive activity.


Introduction
Hypertension is a major risk factor for stroke and other cardiovascular diseases.Among other strategies, angiotensin I-converting enzyme inhibitors (ACE-I) are used in the prevention of hypertension and congestive heart failure [1].Studies have also been done of ACE-I in the treatment of chronic heart failure and myocardial infarction [2], as well as cancer [3,4].ACE (peptidyldipeptide hydrolase, EC 3.4.15.1) catalyzes various reactions, two of which play key physiological roles in regulating blood pressure: conversion of inactive decapeptide angiotensin I into angiotensin II, a powerful vasoconstrictor and saltretaining octapeptide; and inactivation of the vasodilator nonapeptide bradykinin, which is conducive to lowering blood pressure [5].ACE-I have been shown to exhibit antihypertensive activity in spontaneously hypertensive rats and hypertensive patients [6].
Recent research has focused on ACE-I hydrolysates and peptides from animal and vegetal sources.The primary animal sources are proteins from casein, whey pro-tein, fish protein, pig and chicken muscle, hemoglobin, blood plasma protein, gelatin [7] and egg yolks [8,9].The primary plant sources are proteins from legume seeds such as mung bean [10], soybean [11,12], pea [13], and chickpea [14]; oilseeds such as rapeseed [15,16]; and cereal crop derivatives such as maize gluten, wheat germ and buckwheat [7].These studies involve hydrolysate and peptide extraction with commercial enzymes (Alcalase, Flavourzyme, pepsin, pancreatin, trypsin, chymotrypsin, neutrace, etc.) at different hydrolysis times and enzyme/substrate ratios.An important variable in these studies is the concentration (mg protein/mL) of hydrolysate or peptide required to inhibit 50% of ACE-I activity under assayed conditions (IC 50 ); most enzymatic hydrolysates and peptides have values between 0.20 and 246.7 mg/mL [10].Protein hydrolysis can be done chemically using acids or alkalis, although this oxidizes, destroys or modifies some amino acids, thus reducing protein quality [17].Enzymatic hydrolysis is therefore preferred since it can improve the physicochemical, functional and sensory properties of native proteins without affecting their nutritional value [18].In addition to modifying technological functionality, enzymatic hydrolysates have also been used to reduce allergenicity, and formulate pharmaceutical products and substances for clinical applications [19,20].
Legume seeds have higher nutritional protein content (200 -400 g/kg) than other seeds, such as cereals (70 -140 g/kg) [21].The high protein content of legume protein isolates or concentrates are making them increasingly important in the food industry, where they represent a promising alternative ingredient for preparing and developing new foods [22].Germinated legumes currently account for an increasingly larger proportion of total worldwide food legume consumption because germination produces beneficial modifications in their chemical components; among other applications, they are used to produce high nutritional value flours [23].Lima bean (Phaseolus lunatus) is grown in tropical regions such as Mexico.Its seeds have high protein (210 -260 g/kg) and high carbohydrate contents (550 -640 g/kg), low fat (10 -23 g/kg) and fiber levels (32 -68 g/kg), high levels of minerals such as K, Zn, Ca and Fe, and low levels of Na and P [24].Use of plant products as functional ingredients in foods is rising rapidly, due both to their cost-effectiveness and the improvements they provide in product nutritional quality [25].Phaseolus lunatus is a promising source of protein hydrolysates with antihypertensive properties due to aminoacidic composition.The present study aim was to obtain hydrolysates from protein concentrates produced from ungerminated and germinated Phaseolus lunatus using Alcalase or pepsin-pancreatin sequential hydrolysis, and evaluate ACE-I inhibitory activity.

Seeds and Chemicals
Phaseolus lunatus L. seeds were purchased at a local market in Merida, Yucatan, Mexico.Any impurities were manually removed from the seeds.Flours from ungerminated and germinated seeds were prepared by cleaning the seeds, milling them in a Mycros impact mill, and screening through a 20 mesh screen (0.85 mm pore opening).ACE, Hippuryl-L-histidyl-L-leucine, 2,4,6-trichloro-s-triazine, o-phthaldialdehyde, serine standard, gastric porcine pepsin, and pancreatin were purchased from Sigma.Alcalase 2.4 L was obtained from Novo Nordisk, Bagsvaerd, Denmark.All other chemicals were analytical grade.

Seed Germination
Seeds were washed with water, soaked in distilled water and immersed for 1 min in 20 mL sodium hypochloride solution (0.2 w/v), and washed again three times in 25 mL distilled water.They were then germinated for 48 h in a germination room in darkness at 16˚C -18˚C, with moisture levels maintained by spraying with distilled water every 8 h [26].

Protein Concentrates from Ungerminated and Germinated Seeds
Protein concentrates (PC) were extracted from the ungerminated and germinated seeds according to Chel-Guerrero et al. [21].Three, 1 Kg batches of flour were processed by preparing flour: water (1:6 w/v) suspension, and adjusting pH to 11 with 1N NaOH.After soaking for 1 h, the suspensions were milled in a disk mill and passed through 80 and 100 mesh screens to separate the fiber-containing solid fraction from the liquid fraction, which contains the protein and starch.Residual solids were washed 5 times with distilled water at a 1:3 v/v ratio and then passed through 150 mesh screen to eliminate the finest fiber.Wash water was mixed with the supernatants from the initial suspension and this solution allowed to sediment for 30 min to recover the starch fraction and separate the solubilized protein.The pH of the solubilized proteins was adjusted to the isoelectric point (4.5) with 1N HCl.The suspension was then centrifuged at 1317 × g for 12 min (Mistral 3000i, Curtin Matheson Sci.), the supernatants discarded and the precipitates freeze-dried at −47˚C and 13 × 10 −3 mbar.

Amino Acid Analysis of Protein Concentrates
Samples (2 mg) were hydrolyzed with 4 mL of 6N HCl, the solutions sealed in tubes under nitrogen and incubated in an oven at 110˚C for 24 h.Amino acids were determined after derivatization with diethyl ethoxymethylenemalonate by high-performance liquid chromatography (HPLC) [28], using D,L-g-aminobutyric acid as an internal standard.The HPLC system consisted of a Model 600E multi-system with a 484 UV-vis detector (Waters, Milford, MA) equipped with a 300 × 3.9 mm i. Eluted amino acids were detected at 280 nm and the column temperature kept at 18˚C.

Alcalase Protein Hydrolysis
Following the method of Li et al., [10], protein isolate suspensions (4 g protein in 100 ml water) were prepared from the ungerminated and germinated seed flours.Hydrolysis of these suspensions was done using Alcalase 2.4 L, a commercial protease preparation with endopeptidase activity produced from Bacillus licheniformis [14].Its specific activity is 2.4 Anson Units (AU) per gram, one AU being the amount of enzyme which, under standard conditions, digests hemoglobin at an initial rate that generates an amount of trichloroacetic acid-soluble product that produces the same color with Folin reagent as one milliequivalent of tyrosine released per min.This enzyme is a microbial protease preparation used in the food industry to improve the functional and nutritional properties of protein preparations.For the present experiment, the specific activity of the alcalase was adjusted to 0.3 AU/g by dilution with distilled water.A 2 3  factorial design was used to analyze DH and ACE-I activity, with two replicates for both DH and ACE-I activity and four replicates of the central trial.Factors and their corresponding levels were: Factor A, protein concentrate source (ungerminated, PC 1 ; germinated, PC 2 ); Factor B, enzyme/substrate ratio (E/S) (1/50 or 1/10); and Factor C: hydrolysis time (0.5 or 2.0 h).Hydrolysis was done in a pH stat, 50 mL vessel, by adding 0.3 AU/g alcalase.The reaction was run at 50˚C, pH adjusted to 7.0 with 0.1 N NaOH and the suspension kept under constant agitation.Hydrolysis was finalized by inactivating the protease through heating to 85˚C for 10 min.The hydrolysates were centrifuged at 13211 × g for 20 min at 4˚C (Beckman) and the supernatants freeze-dried (Labconco) for later analysis.

Pepsin-Pancreatin Protein Hydrolysis
Protein hydrolysis was done with a sequential pepsinpancreatin system involving digestion with pepsin for half the hydrolysis time followed by incubation with pancreatin for the other half [29].The gastric porcine pepsin (P7000, Sigma) used here has an enzymatic activity of 800 units/mg protein (one unit will produce a ΔA280 of 0.001 per min at pH 2.0 and 37˚C, measured as TCA-soluble products using hemoglobin as substrate), and the porcine pancreatin (P1500, Sigma) can hydrolyze casein more than 25 times its weight in 60 min at pH 7.5 and 40˚C.A 2 3 factorial design was used again to analyze DH and ACE-I activity, with two replicates for both DH and ACE-I activity, and four replicates of the central trial.
Factors and their corresponding levels were: Factor A, PC source (PC 1 , PC 2 ); Factor B, E/S ratio (1/50 or 1/10); and Factor C, hydrolysis time (1.0 or 3.0 h).Hydrolysis was done in a 50 mL reaction vessel equipped with a stirrer, thermometer and pH electrode.Hydrolysis was run at 37˚C using a 4% (w/v) concentration for both enzyme solutions, and pH adjusted to 2 with 0.1 N HCl for pepsin and to 7.5 with 0.1 N NaOH for pancreatin.The hydrolysis reaction was stopped by heating to 80˚C for 20 min.The hydrolysates were centrifuged at 13211 × g for 20 min at 4˚C (Beckman) and the supernatants freezedried (Labconco) for later analysis.

Degree of Hydrolysis
Degree of hydrolysis (DH) was calculated by determining free amino groups through reaction with OPA reagent [30] using a serine standard.Total number of amino groups was determined in a 100% hydrolyzed sample by treatment with 6 N HCl at 110˚C for 24 h in a vacuum oven.This analysis was repeated twice.

ACE-I Activity
ACE-I activity was determined according to Hayakari et al. [31], with some modifications.ACE hydrolyzes HHL, yielding hippuric acid and His-Leu.This method relies on the colorimetric reaction of hippuric acid with TT in a 0.

Statistical Analysis
Statistical treatment of DH and ACE-I activity results was done with an analysis of variance (ANOVA), a Duncan comparison of means and a 2 3 factorial design regression analysis.A Student t test (5% significance level) was applied to compare differences between means for the flour and PC chemical compositions (ungerminated and germinated seeds) [32].All statistical analyses were done with the Statgraphics Plus version 5.0 statistical software.

Proximate Composition
Germination for 48 h significantly (p < 0.05) effected flour and PC composition (Table 1).(11).At this pH, fat saponifies with the alkali and these (mainly polar) fats precipitate together with the protein at the isoelectric point (pH 4.5), which has been reported in protein concentrates from Cicer arietinum L [25] and Canavalia ensiformis [21].The extent of this saponification probably depends on legume seed fatty acid composition.Germination resulted in higher protein content, probably due to synthesis of enzyme proteins for seed dicotyledon development in the mother plant and degradation of reserve nutrients (lipids and carbohydrates), both vital to the purpose of providing the energy required for protein synthesis in plant growth [35].Protein content in both flours and PCs was similar to previously reported levels in P. lunatus [21,[37][38][39].The protein recovery percentages (i.e.process efficiency) for PC production were 34.30% for ungerminated seeds and 30.84% for germinated seeds, which are lower than mentioned for other seeds: 44% in faba beans [40]; 60% in lupine seeds [41]; and 46% in safflower seeds [42].

Amino Acids
The PCs exhibited similar amino acid profiles (  ).The PC amino acid profiles observed here are very similar to amino acid profiles reported for lima bean [39] and mung bean protein isolates [10].
The mathematical model explaining the behavior of DH as a function of PC source (A); E/S ratio (B); hydrolysis time (C); PC source-E/S ratio (AB); PC sourcehydrolysis time (AC); E/S ratio-hydrolysis time (BC); and PC source-E/S ratio-hydrolysis time (ABC) had a determination coefficient (R 2 ) of 0.998 (Figure 1, Equation ( 1)).This model showed E/S ratio (B) and hydrolysis time (C) to have the most positive effects on DH (p < 0.05), with coefficients of 4.11 and 9.94, respectively.

Hydrolysis with Alcalase
ACE-I activity (IC 50 ) ranged from 0.61 -2.40 mg/mL (Table 3).The variance analysis showed it to be unaffected by the principal factors and two-factor interactions, but that it was affected (p < 0.05) by the three-factor interaction.The regression analysis showed it to fit a first-order polynomial model with an R 2 = 0.931.This explains the behavior of ACE-I activity as a function of PC source-E/S ratio-hydrolysis time (ABC) (Equation ( 2)).
Hydrolysis with the endoprotease Alcalase increased DH with longer hydrolysis time in both PC 1 and PC 2 , the highest DH being in PC 2 (DH = 58.94%)(Table 3).This greater Alcalase activity in PC 2 may have resulted from partial elimination of lipids and carbohydrates during germination (Table 2).Another possibility is that denaturation during the concentration process produced a better molecular orientation in PC 2 than in PC 1 , which would have enhanced the enzyme-substrate interaction, or that germination increased Ala, Met and Phe residues, upon which alcalase has broad specificity [19].
Given that the three-factor interactions had a significant (p < 0.05) effect on ACE-I activity when using the three evaluated factors, the first-order, 2 3 statistical model could be optimized using higher-order models, increasing the number of factors evaluated, and evaluating the response variable as a function of the substrate and enzyme employed.From a practical standpoint, Treatment 1 (IC 50 = 0.94 mg/mL) would be optimum because it involves lower costs (i.e.ungerminated seeds, The variance analysis indicated that all factors and interactions influenced (p < 0.05) DH (Equation ( 1)), and the regression analysis for this response variable indicated that the results fit (p < 0.05) a first-order polynomial model.% 39.52 2.08A 4.11B 9.94 C 0.93AB 01.40AC 1.68BC 1.28ABC 0.998  minimal enzyme and a shorter hydrolysis time are required), although it did not exhibit a major ACE-I effect (0.56 mg/mL).At 0.5 h hydrolysis time and the 1/50 E/S (Table 3, Figures 1(a1), (b1)), alcalase activity was highest in PC 2 (DH = 29.91%),whereas ACE-I inhibitory activity was highest in PC 1 (IC 50 = 0.94 mg/mL).This trend inverted when the E/S ratio increased to 1/10 (Table 3, Figures 1(a2), (b2)), in that alcalase activity was highest in PC 1 (33.40%) and inhibitory activity was highest in PC 2 (IC 50 = 0.61 mg/mL).At 2 h hydrolysis time, PC 2 exhibited the highest enzymatic activity at both E/S levels (1/50: DH = 46.67%,1/10: DH = 58.94%)(Table 3, Figures 1(b1),  2(b2)).In contrast, inhibitory activity was independent of PC source and E/S ratio, with the highest IC 50 values in PC 1 at 1/10 (0.56 mg/mL) and PC 2 at 1/50 (0.87 mg/mL) (Table 3, Figures 1(a1), (b1)).The inhibitory activity mathematical model and the fact that only the three-factor interaction (ABC) was significant, indicate that treatment 7 (PC 1 , 1/10 E/S, 2 h hydrolysis time) produced optimum ACE-I inhibitory activity (IC 50 = 0.56 mg/mL).
Inhibitory activity in treatments 2 (PC 2 , 1/50 E/S) and 3 (PC 1 , 1/10 E/S) improved with greater hydrolysis time (2.40 -0.87 and 2.05 -0.56 mg/mL, respectively), but declined in treatments 1 (PC 1 , 1/50 E/S) and 4 (PC 2 , 1/10 E/S) (0.93 -2.11 and 0.61 -1.16 mg/mL, respectively).This suggests that a peptide (or peptides) obtained at 0.5 h with good ACE-I activity was hydrolyzed at 2 h hydrolysis time, consequently decreasing its antihypertensive effect.In other words, there is an optimal hydrolysis point or DH above which a larger amount of ACE-I peptides are degraded than formed, lowering overall ACE-I activity.Similar behavior has been reported for protein hydrolysates from sunflower [29], chickpea [14] and fish [43].The central trial treatment had an average DH of 39.81% and an average IC 50 of 1.01 mg/mL, meaning that PC 1 -PC 2 mixtures are not recommended in terms of their inhibitory activity.Overall, the optimum ACE-I activity range among the Alcalase hydrolysates was observed in PC 1 (IC 50 between 0.56 -2.11 mg/mL), probable because this enzyme can produce peptides with molecular structures which enhanced substrate ACE-I activity.However, Alcalase reacted better with PC 2, because the Alcalase cleaved more peptidic bonds which in turn generated high DH values (Table 3).Both the DH and inhibitory activity values for the P. lunatus Alcalase hydrolysates are similar to reported results for Alcalase hydrolysates from other proteins: Alaska pollack skin (IC 50 = 0.63 mg/mL, 6 h hydrolysis time) [43]; chickpea protein isolate (IC 50 = 0.19 mg/mL, DH = 18%, 0.5 h hydrolysis time) [44]; chickpea legumin (IC 50 = 0.18 mg/mL, DH = 27%, 0.5 h hydrolysis time) [14]; and mung bean (IC 50 = 0.64 mg/mL, DH = 17%, 2 h hydrolysis time) [10].

Hydrolysis with Pepsin-Pancreatin
Pepsin-pancreatin sequential hydrolysis did not hydro-lyze the lima bean PCs as effectively as Alcalase, producing comparatively lower DH values (15.35% -37.07%), although ACE-I activity was generally higher in these hydrolysates (Table 4, Figure 2).These differences in hydrolytic behavior may have originated in the hydrolysis method (e.g.enzyme type and concentration) [45], or  the low aromatic and cationic amino acids contents (5.04% and 10.95% in PC 1 , respectively; 4.53% and 10.96% in PC 2 , respectively), upon which pepsin and pancreatin have broad specificity (Table 2) [47].Enzymatic activity for the three evaluated enzymes differed in both experiments (0.3 AU/mL for alcalase and 4% w/v for pepsin and pancreatin) even though hydrolysis conditions (pH and temperature) were optimum for each enzyme in all treatments.
The IC 50 values for the pepsin-pancreatin hydrolysates ranged from 0.25 to 0.69 mg/mL, with the overall higher values in PC 2 (0.28 -0.53 mg/mL).As occurred in the alcalase hydrolysates, DH improved with increasing hydrolysis time in both PC 1 and PC 2 , and inhibitory activity was independent of hydrolysis time.
The variance analysis showed that the E/S ratio, hydrolysis time, the PC source-E/S ratio interaction and the three-factor interaction (PC source-E/S ratio-hydrolysis time) influenced (p < 0.05) DH (Equation ( 3)) and the three-factor interaction (PC source-E/S ratio-hydrolysis time) influenced (p < 0.05) DH (Equation ( 3)).
The regression analysis indicated that the DH results fit (p < 0.05) a first-order polynomial model.This mathematical model also explained DH behavior as a function of E/S ratio (B), hydrolysis time (C), the PC source-E/S ratio (AB) interaction and the PC source-E/S ratio-hydrolysis time (ABC) interaction, with a determination coefficient of R 2 = 0.976.The model indicated that E/S ratio (B) and hydrolysis time (C) had most significant effect (p < 0.05) on DH (coefficients = 3.88 and 3.86, respectively).
Optimum ACE-I activity with this hydrolysis method (0.25 mg/mL) was observed in Treatment 12.The increase from 1 to 3 h hydrolysis time increased ACE-I activity only in PC 1 at the 1/50 E/S (0.69 -0.34 mg/mL), suggesting that this treatment generated peptides with good antihypertensive properties at 3 h hydrolysis time (Figure 2).Aside from a minor improvement in IC 50 in PC 2 at 1/50 E/S with the increase from 1 to 3 h (0.29 -0.28 mg/mL; p > 0.05), increased hydrolysis time generally compromised ACE-I activity, probably due to hydrolysis of some antihypertensive peptides with increased hydrolysis time.This coincides with Megías et al [29], who, in a study of sunflower protein hydrolysates produced with pepsin-pancreatin sequential hydrolysis, report that the highest ACE-I values were observed at DH 10% -22%, rather than at higher DHs of 30% -35%.
Structure-activity correlations between different peptide ACE inhibitors indicate that ACE binding is strongly influenced by the substrate's C-terminal tripeptide sequence.The C-terminal tripeptide can interact with the three sub-sites at an active ACE site.This enzyme apparently prefers substrates or competitive inhibitors containing hydrophobic (aromatic or branched-side chain) amino acid residues at each of the three C-terminal positions.Many studies have shown that tripeptides or dipeptides with potent inhibitory activity have Trp, Tyr or Pro at their C-terminal, and branched aliphatic amino acids at the N-terminal [43].A quantitative QSAR study of the structure-activity relationship of ACE-I peptides has also confirmed that tripeptides composed of amino acids with hydrophobicity at their C-and N-terminals have potent ACE-I activity [10].Given that both the tested P. lunatus PCs contained high proportions (33.34% -34.10%) of hydrophobic amino acids (Table 2), it is probable that this action mechanism contributed to their ACE-I activity.

Conclusion
Protein concentrates from ungerminated and germinated Phaseolus lunatus seeds are good sources of ACE inhibitory peptides when hydrolyzed with the proteases Alcalase or pepsin-pancreatin.The highest ACE inhibitory activity was observed using sequential hydrolysis with pepsin-pancreatin, a potentially important fact since pepsin is the main proteolytic enzyme generated in the stomach during food digestion and pancreatin includes proteases such as trypsin, chymotrypsin and elastase, all released by the pancreas into the small intestine.In other words, the pepsin-pancreatin protein hydrolysates produced in the present study represent a set of peptides resembling those generated during digestion of P. lunatus proteins in the human organism.Because they are resistant to pepsin and pancreatin, these peptides could be absorbed by digestive epithelial cells in the small intestine and possibly exert an ACE inhibitory effect within the organism.Lima bean protein hydrolysates produced with pepsin-pancreatin are a potential ingredient in the production of physiologically functional foods with ACE inhibitory activity.
Protein concentration levels were 56.40% in PC 1 and 69.14% in PC 2 .Fat content increased in both PC 1 (3.12%) and PC 2 (4.16%), probably caused by concentration of fats at the alkaline pH

Table 1 . Proximate composition of protein concentrates from ungerminated and germinated P. lunatus seed flours.
a-b,A-B Different letter superscripts in the same row indicate significant difference (P < 0.05); F 1 : Flour from ungerminated seeds; F 2 : Flour from germinated seeds; PC 1 : Protein concentrate from ungerminated seeds; PC 2 : Protein concentrate from germinated seeds; NFE: Nitrogen-free extract.