Yellow pea flour and protein isolate as sources of antioxidant peptides after simulated gastrointestinal digestion

Although peas are widely consumed legumes throughout the world, the bioactivity of the peptides released by the gastrointestinal digestion has not been sufficiently studied so far. The objective of the present work was to evaluate the potential of flours and protein isolates obtained from two varieties of yellow peas as sources of antioxidant peptides. Flours and protein isolates were prepared and submitted to a simulated gastrointestinal digestion. Protein hydrolysis degree (TNBS method) and protein solubility (in phosphate buffer saline, pH = 7.4) values were independent on the starting material. Antioxidant activity measured by oxygen radical absorbance capacity (ORAC) and hydroxyl radical averting capacity (HORAC) showed no differences between varieties. A lower activity was registered for protein isolates with respect to flours in the case of HORAC, which could be associated with a loss of molecules with molecular masses lower than 43 kDa in the protein isolates. A significant increase in activities was evidenced by both methods after gastrointestinal digestion, except in the case of HORAC activity of flours. Digested from protein isolates presented a greater ratio of molecules smaller than 1.4 kDa and a lower ratio of those larger than 6.5 kDa with respect to digested flours, according to electrophoresis and gel filtration chromatography studies. Results suggested that the presence of other components or/and the initial state of proteins would affect proteolytic attack of digestive enzymes. Both, pea flours and protein isolates, present interesting potential as antioxidants food ingredients.


| INTRODUCTION
In recent years, the role of dietary proteins as physiologically active components has been exhaustively studied, demonstrating that they can be the source of biologically active peptides. These peptides are inactive within the sequence of parent protein and can be released during food processing or gastrointestinal digestion. Once peptides are released, they may cause different physiological actions such as antioxidant activity. Studies of structure-activity relationship have shown that physic-chemical and structural features, such as charge, amino acid sequence, molecular size, and hydrophobicity, may determine the bioactivity of peptides (Sarmadi & Ismail, 2010). Conditions in the gastrointestinal tract, such as activity of digestive enzymes and pH values might influence the structures and functions of the released peptides (Segura-Campos, Chel-Guerrero, Betancur-Ancona, & Hernandez-Escalante, 2011). The ability of peptides to resist the enzymatic attack is related to their amino acid composition due to the specificity of digestive enzymes. Gastrointestinal digestion can be in vitro simulated by different methodologies that try to mimic physiological conditions (temperature, agitation, pH, enzyme activities, and fluids composition) and the sequence of events in the gastrointestinal tract. Static (or biochemical) methods are the simplest ones and include two or three digestion steps (oral, gastric, intestinal) which products stay in one only reactor (Minekus et al., 2014).
Pea (Pisum sativum) seeds are an important source of nutrients and healthy compounds (20% to 26% w/w protein, 1% to 3% w/w lipids, 46% to 50% w/w carbohydrates, and 14% to 18% w/w fiber) providing approximately 317 kcal/100 g of grain (Zulet & Martínez, 2001). Peas are a source of ingredients such as flours, protein isolates, starches, and fiber, which are of increasing importance in the design of healthy foods and foods for special diets (Agboola, Mofolasayo, Watts, & Aluko, 2010). As a negative aspect, peas contain anti-nutritional factors, including α-galactosides, trypsin inhibitors, and phytates, which concentrations differ widely between varieties, and whose elimination is essential to improve the nutritional quality.
There are simple and economical processing techniques capable to effectively remove anti-nutritional factors, such as soaking, cooking, and germination (Vidal-Valverde, Frias, & Valverde, 1992).
Diets rich in dry peas have showed to be effective in decreasing the incidence of colon cancer, type-2 diabetes, LDL-cholesterol, and heart disease (Roy, Boye, & Simpson, 2010). As for other legumes, these beneficial effects have been related to micronutrients, phytochemicals, and bioactive peptides. Inhibitory activity of the angiotensin converting enzyme (ACE), antioxidant activity, antitumor activity, among others, have been described for peptides released from legume proteins by using diverse proteases and proteolysis conditions (López-Barrios, Gutiérrez-Uribe, & Serna-Saldívar, 2014). In this sense, it has been reported that the nonhydrolyzed pea protein showed no ACE inhibitory activity, but this activity was observed after in vitro gastrointestinal digestion (Barbana & Boye, 2010;Jakubczyk & Baraniak, 2014).
Both flours and pea protein isolates are ingredients used in food formulation. The presence of diverse components and the complexity of the matrix could have an effect on the gastrointestinal digestion processes. The aim of this work was to study the effect of the simulated gastrointestinal digestion on flours and protein isolates from two varieties of yellow peas, focusing on the release of peptides with antioxidant activity. In this way, the potentiality of both types of ingredients as sources of antioxidant molecules was evaluated.

| Pea protein isolate (I)
Protein isolates were obtained from flours of the two pea varieties by a protocol adapted from Makri, Papalamprou, and Doxastakis (2005) and Qayyum, Butt, Anjum, and Nawaz (2012). Flour was defatted by lipid extraction with hexane (overnight, room temperature). Defatted flour was dispersed (10% w/v) in MilliQ water, and pH was adjusted to 9.5, agitating for 40 min at room temperature. After centrifugation (11,000×g, 20 min, 4 C), soluble proteins were precipitated by adjusting the supernatant to the isoelectric pH (pH = 4.5) and centrifuging (10,000×g, 20 min, 4 C). The precipitated proteins were suspended in MilliQ water, neutralized (pH = 7), freeze-dried, and stored at 4 C.

| Antitrypsin activity evaluation
Seeds were soaked in tap water for 4 h. Soaked seeds were boiled (seed to water ratio: 1:5 w/v) for 30 min, and, finally, they were dried at 54 C overnight (Khattab & Arntfield, 2009). Flour was obtained from the thermally treated seeds (Ft) according to Section 2.2.1. Antitrypsin activity of samples of F, Ft, and I of the two varieties of peas was evaluated. Dispersions (10% w/v) of these samples in phosphate buffered saline (PBS) solution (pH = 7.4) were prepared, kept overnight at 4 C, and then centrifuged (21,380×g, 25 min). Antitryptic activity was determined in the supernatant according to Manassero, Vaudagna, Sancho, Añón, and Speroni (2016)

| Antioxidant activity
Antioxidant activity of soluble fractions (see Section 2.2.5) was evaluated by the oxygen radical absorbance capacity (ORAC) and the hydroxyl radical averting capacity (HORAC) assays using previously

| Flours (F) and protein isolates (I) characterization
Flours were prepared from raw seeds (FY and FN) as well as from boiled seeds (FYt and FNt) of the two yellow pea varieties (Yams and Navarro). Inhibition of the trypsin activity by dispersions (10 mgÁml −1 ) of all these samples was evaluated. Dispersions obtained from untreated flours showed values between 20% and 30% of inhibition of the trypsin activity (TI %). When values were normalized by the protein content of samples, FN presented a significantly lower (p < 0.05) specific inhibition than FY (0.8 ± 0.3 and 1.7 ± 0.1 TI %Ámg protein −1 , respectively). Vidal-Valverde et al. (2002) analyzed 18 varieties of Spanish peas, which presented a broad range of antitrypsin activity (measured by a different methodology than in the present work). Differences were attributed to the diversity of climatic and soil conditions during cultivation. The authors also informed that antitrypsin factor in peas was 10 times lower than in soy. Different processing treatments such as dehulling, soaking, cooking, fermentation, and germination have been used to reduce antinutritional factors in food legumes. Frequently, one method is not enough, and the combination of two or more process is required. Soaking is used to remove soluble antinutritional factors, whereas thermal treatment is useful to inactivate heat-sensible factors such as antitrypsin factors due to the thermal denaturation of these proteins. In our case, antitrypsin activity of flours was strongly reduced after soaking and boiling treatment of seeds (0.36 ± 0.08 and 0.33 ± 0.08 TI %Ámg protein −1 for FYt and FNt, respectively). Khattab and Arntfield (2009)  According to these results, the only slight difference registered between the two varieties of yellow peas was a higher antitrypsin factor activity both in the flour and in the protein isolate of Yams variety. However, it is important to remark that antitrypsin activity was low in both flours and even much lower in soaked and thermally treated flours and in protein isolates, suggesting that, for these peas, antitrypsin factor is not a major problem from a nutritional point of view. The protein contents of gastrointestinal digests are also shown in Table 2. Protein solubility (PBS, 10 mg sampleÁml −1 ) was analyzed before and after digestion. Proteins in flours presented a solubility of about 70% with respect to total protein, without significant differences between pea varieties (p > 0.05). Under the extraction conditions used, albumins and globulins should be solubilized. Protein isolates showed a protein solubility value similar (p > 0.05) than flours.

| Simulated gastrointestinal digestion
In agreement with this, Ladjal-Ettoumi, Boudries, Chibane, and Romero (2015) informed a protein solubility of 65% at pH = 7 and 70% at pH = 8 for Algerian pea protein isolates. After simulated gastrointestinal digestion, none of the samples showed significant changes (p > 0.05) in protein solubility (Table 2). These results suggest that the proteins attacked by digestive enzymes were mainly those that were soluble in PBS before proteolysis. However, other variables that could affect protein solubility cannot be ruled out, such as treatment at 85 C carried out to inactivate the digestive enzymes could cause aggregation and insolubilization of some-especially hydrophobic-polypeptides or peptides, and this effect could be dif-  (Figure 1a, lane 3 and lane 4). In the case of the T A B L E 1 Centesimal composition (g × 100 g −1 wet basis) of yellow pea ingredients  suggesting that these peptides were resistant to the simulated gastrointestinal process. New bands (10 to 13) with MW lower than 6.5 appeared after digestion of F and I (Figure 1b, lanes 5 to 8).
Chromatograms of soluble fractions from protein isolates showed a different relative proportion of molecules with respect to flours, with a greater ratio of peaks 1 + 2 and lower proportion of the area of the other peaks. These results showed again a partial loss of low molecular weight polypeptides (<43 kDa) during the protein isolate preparation (Figure 2c,d). After gastrointestinal digestion, molecules smaller than 11 kDa decreased and new peaks appeared: peak III (3.2 to 11.5 kDa), peak V (2.5 to 3.2 kDa), peak VI (1 to 1.7 kDa), and peaks IV and VII (MW < 1.7) (Figure 2c,d). A Superdex 30 column (optimal separation in the range of MW < 10 kDa) was used in order to analyze low MW peptides. Chromatograms obtained from flour soluble fractions presented a peak corresponding to MW greater than 10 kDa, molecules greater than 6.5 kDa (peak 3), and four small peaks (MW between 2 and 0.1 kDa) (Figure 3a,b). The simulated gastrointestinal digestion produced a diminution of the peaks 1, 2, and 3 (MW > 6.5 kDa) and the appearance of molecules in the range of 0.6-4.3 kDa (peak III), and between 0.1 and 0.5 kDa (peaks IV and V) corresponding to peptides of about four amino acids to free amino acids, which could also include other kinds of molecules of low molecular weight released by gastrointestinal digestion (Figure 3a,b). In the case of protein isolates, peptides lower than 6.5 kDa were much less abundant than in flours (Figure 3c,d). After gastrointestinal digestion, peaks corresponding to MW > 6.5 kDa diminished and diverse peaks associated to molecules in the range of 0.1-7.1 kDa appeared, the more relevant between 0.6 and 4.3 kDa (peak III). Also in this case, molecules with very low MW could be evidenced (peak VIII) (Figure 3c,d).

| Antioxidant activity
The antioxidant activity of soluble fractions of flours and protein isolates before and after simulated gastrointestinal digestion was evaluated. ORAC assay measures hydrophilic chain-breaking antioxidant (P) very low molecular weight standard. All lanes were stained with silver F I G U R E 3 Gel filtration chromatograms using a Superdex 30 column (optimal separation range < 10 kDa). Molecular weight markers are shown F I G U R E 2 Gel filtration chromatograms using a Superdex 75 column (separation range: 1 to 70 kDa). Molecular weight markers are shown capacity against ROO• radicals induced by AAPH, which proceed as a classic hydrogen atom transfer mechanism (Ou, Hampsch-Woodil, & Prior, 2001 (Table 3). This parameter did not present significant differences (p > 0.05) between flours and protein isolates. This fact indicated that flours did not contribute additional soluble components capable to modify the ROO• scavenging activity with respect to the protein isolates, suggesting that this activity is due mainly by polypeptides or other components associated with the protein fraction.
HORAC assay, in which the oxidative degradation of fluorescein occurs by hydroxyl radicals generated by the Fenton reaction (Ou et al., 2002), was also performed. Dose-response curves presented a linear fitting in this case. The yellow pea variety had no influence (p > 0.05) on the IC 50 values of flours and neither on those of protein isolates ( whereas HORAC evaluates mainly the capacity to chelate metals inhibiting the formation of hydroxyl radicals (Ou et al., 2002). A possible explanation to the observed behavior may be given by certain differences in the composition of both ingredients. As was previously described, they presented some differences in the molecular composition with a diminution of molecules smaller than 43 kDa in the case of the protein isolates. The present results suggested that these lost components would have relevant activity as metal chelator but not as radical scavengers.
Gastrointestinal digests exhibited significantly lower IC 50 values in ORAC assay (p < 0.05) than the starting materials in all cases, without significant differences among them. These results indicated than the digestion process produced an increment of about four times in the ORAC activity of flours and protein isolates by releasing antioxidant compounds, presumably peptides, although the presence of other antioxidant components could not be discarded. IC 50 values for HORAC assay did not present significant differences (p > 0.05) for flours before and after simulated gastrointestinal digestion (Table 3).
However, digestion from protein isolates showed a significant decrease (p < 0.05) in the IC 50 value, indicating an increment of HORAC activity by release of active molecules (  (Girgih et al., 2015). Also, we cannot discard the presence of other kinds of nonprotein components, which the content would be differential between flours and protein isolates and/or could become active after digestion, such as polyphenolic compounds.
The presence of this type of compound in free, esterified, and/or linked to protein fractions has been demonstrated in diverse legumes (Fratianni et al., 2014;Vaz Patto et al., 2015) and will be later studied in our pea varieties. Literature about antioxidant activity of peptides derived from pea proteins is scarce and, as far as we know, the effect of gastrointestinal digestion on flours and protein isolates have not been studied until now. In the present work, a first evaluation of yellow peas flours and protein isolates as antioxidant functional ingredients was performed. Two pea varieties were studied, and there were no substantial differences between them in molecular composition of flours and protein isolates, before and after simulated gastrointestinal digestion.
The proteolysis degree achieved during gastrointestinal digestion was independent on the variety of pea as well as on the pea ingredient

CONFLICT OF INTEREST
There are no conflicts to declare.

ETHICS STATEMENT
This research did not include any human subjects and animal experiments.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.