Impact of processing on the in vitro protein quality, bioactive compounds, and antioxidant potential of 10 selected pulses

Pulses are consumed worldwide with different processing methods, which may impact their digestibility, protein quality, and composition. This study aims to analyze the effect of extrusion, baking, and cooking on protein nutritional parameters; bioactive compounds; and the impact on antioxidant capacity (AOX) of 10 selected pulses. Sodium dodecyl sulphate‐polyacrylamide gel electrophoresis (SDS‐PAGE) revealed that thermal processing causes modifications to the main storage proteins in pulses. Heating decreased saponin content from 12% to 44% in most heat‐processed samples; phytates were reduced 30%–84%, and polyphenol content decreased 28%–66%. In addition, the in vitro protein digestibility (IVPD) was enhanced 2.5%–9.5%, 3.5%–10.7%, and 2.2%–8.4% by extrusion, cooking, and baking, respectively. AOX showed an improvement in all processed samples (compared to raw flour) evaluated by the 2,2‐diphenyl‐1‐picrylhydrazyl (DPPH) method and by the oxygen radical absorbance capacity (ORAC). Fe2+ chelation showed that extruded and baked chickpea exhibited a decrease in IC50 by 40% and 70%, respectively. Extruded green and yellow split pea presented the highest Fe2+ chelation, improving by 11%–17% and 13–80%, respectively, when compared to the raw samples. Reducing power was enhanced by 26% in extruded chickpea, 18% and 29% in extruded and baked faba bean, respectively, and 50% in baked navy bean, when compared to the raw samples. Extrusion showed the highest β‐carotene AOX improvements (IC50 90%–96%). In this study, it was demonstrated that pulses AOX attributes can be enhanced by thermal processing; however, this will depend on the legume species and heating process applied. Furthermore, cooking seems to be the most effective thermal method to decrease saponins and phenolics, while extrusion reduced effectively phytic acid on bean samples, and cooking for the rest of pulses. All heating treatments affected positively IVPD, while the highest in vitro protein‐digestibility corrected amino acid score (IVPDCAAS) values were observed for baked pulses. Employing adequate processing methods represents an effective strategy to improve the digestibility of their proteins, as well as increasing the antioxidant potential of the resulting ingredients.

Samples of faba beans (FB, Vicia faba L.) were provided by SaskCan Pulse Trading (Regina, SK, Canada). Samples of chickpeas (CC, Cicer arietinum L.) were provided by Saskcan Pulse Trading, Thompsons Ltd., and Viterra. Prior to processing, samples of similar legume species from different suppliers were combined and thoroughly mixed.
Flours of pulses were obtained by milling using a Jacobson 120-B hammer mill (Minneapolis, MN) with screen hole size of 0.050 inch (0.127 cm), round. Milled pulses were subjected to three heat treatments as presented in Figure  Once the heat treatments were done, all samples were milled as described above.

| Amino acid profile
The amino acid analysis of samples was conducted in accordance with the Agilent method (Long, 2015). Briefly, 4 mg of protein of raw, extruded, cooked, and baked samples was hydrolyzed with 6 N HCl containing 0.1% phenol and Norvaline (as internal standard) for 24 h at 110 ± 2 C in glass tubes sealed under vacuum. The hydrolyzed samples were cooled to room temperature, and solutions were evaporated with nitrogen until dryness. Once dry, the amino acids were dissolved by the addition of 10 mM sodium borate buffer (pH 8.2, containing 0.1% [w/v] HCl) and then filtered with 0.22 μm polyvinylidene difluoride (PVDF) filters (low protein binding) (Sigma-Aldrich, USA). Analysis was performed using an Agilent Poroshell HPH-C18 reversed-phase column monitored with Agilent 1200 series HPLC system (Agilent Technologies Canada Inc., Mississauga, ON, Canada), utilizing an automatic post-column OPA and 9-fluorenylmethoxycarbonyl group (FMOC) derivatization and detection using absorbance at 338 nm. The separation was performed at a flow rate of 1.5 mL/min, employing a mobile phase of: (A) 10 mM Na 2 HPO4, 10 mM Na 3 B4O7, 5 mM NaN3, adjusted to pH 8.2 with HCl; and (B) ACN: MeOH:water (45:45:10,v/v/v). The elution program was as follows: 0 min 2% B, 20 min 59% B, and 25 min 2% B. All samples were performed in triplicate.
The content of tryptophan in the pulses samples was determined separately by alkali hydrolysis following the method of Yust et al. (2004), with slight modifications. Samples (15 mg of protein) were F I G U R E 1 Processing methods workflow dissolved in 3 mL of 4 N NaOH, sealed in hydrolysis tubes and incubated in an oven at 110 C for 24 h. Hydrolysates were cooled down, neutralized to pH 7 using 12 N HCl and diluted to 25 mL with 1 M sodium borate buffer (pH 9). Aliquots of these solutions were filtered through a 0.45 μm PVDF filters and then injected into a Nova-Pack C18 column (Waters, Mississauga, ON, Canada). An isocratic elution system consisting of 25 mM sodium acetate, 0.02% sodium azide (pH 9)/acetonitrile (91:9) delivered at 1 mL/min was used.

| Bioactive compounds
The quantification of saponins in the sample was performed following the method of Hiai et al. (1976), using diosgenin as a standard (0-0.125 mg/mL). Samples were mixed twice with methanol 80%, 200 μL of supernatants was pooled after centrifugation, and 250 μL of Vanillin and 2.5 mL of sulfuric acid were added. Absorbance was measured at 520 nm against distilled water as a blank. The content of saponins was reported as mg equivalents of diosgenin/g of sample.
The content of phytates was determined following the method of Vaintraub and Lapteva (1988), using phytic acid as a standard (0-0.160 mg/mL). Samples (0.5 g) were mixed with 10 mL of 5% HCl, stirred for 1 h and then centrifuged (5000 rpm × 10 min). Subsequently, 200 μL of extract was diluted with 2.8 mL of distilled water and mixed with Wade reagent (1.0 mL). The absorbance was measured at 500 nm against distilled water as a blank. The phytate content was expressed as mg equivalents of phytic acid/g of sample.
Finally, the content of total phenolics was determined in the raw and processed flours following the Folin-Ciocalteu method (Singleton et al., 1999), using gallic acid as a standard (0-100 mg/mL). Samples (0.5 g) were mixed with 10 mL of acidified methanol (1% HCl) and extracted for 18 h in dark conditions at room temperature.
Afterwards, samples were centrifuged at 5000 rpm for 10 min. Supernatants (20 μL) were vortexed with distilled water (1.58 mL), Folin-Ciocalteau reagent (100 μL), and Na 2 CO 3 10% (300 μL). Then, samples were kept away from light until reading; absorbance was measured at 760 nm against distilled water as a blank. Results were expressed as mg gallic acid equivalents per mg of sample (GAE mg/mg sample).

| In vitro protein digestibility-corrected amino acid score: Hydrolysates
Samples were digested following the method reported by Tinus et al. (2012), with few modifications. Briefly, the equivalent of 62.5 mg of protein was rehydrated in 10 mL of Milli-Q water, heated to 37 C and adjusted to pH 8.0. The samples were monitored for 10 min to record the stability of the pH, followed by the addition of a multienzyme cocktail containing trypsin (16 mg, 13,000-20,000 BAEE units/mg protein), chymotrypsin (31 mg, 40 units/mg protein) and protease (50-100 units/g solids). After the addition of the digestive cocktail, the subsequent pH drop was recorded for 10 min. Subsequently, the samples were transferred in a 95-99 C bath for 15 min and cooled down in an iced bath. Then the samples were centrifuged at 4 C, 6000 g for 30 min, and the supernatants were recovered. The in vitro protein digestibility (IVPD) was calculated as follows: IVPD % ð Þ= 65:66 + 18:10 × pH 0min -pH 10min ð Þ Meanwhile the in vitro protein-digestibility corrected amino acid score (IVPDCAAS) was calculated as a product of the amino acid score (AAS) and IVPD% (Nosworthy et al., 2018 where A 0 is the absorbance of the negative control (distilled water) and A 1 is the absorbance of the samples.

| β-carotene bleaching activity
The β-carotene bleaching was measured according to Marco (1968 where DR is the degradation rate among the absorbance at the beginning (ABS 0min ) and the end of the reaction (ABS 60min ) for control and samples. Using the DR, the antioxidant activity (AA) was calculated as % inhibition relative to the control as follows: where DR control is the degradation rate of β-carotene in the absence of sample. Results were expressed as percentage (%) AA.

| ORAC assay
The ORAC assay was performed as previously described (Huang et al., 2002 All the components were analyzed.

| SDS-PAGE
The electrophoretic pattern of raw and processed pulses displayed multiple bands ranging from 10 to 105 kDa ( Figure 2). For RL, bands at 70 kDa (associated to convicilin subunits) were still present in all the processed samples, but they were less intense in the baked sample than in the extruded and cooked samples. A set of bands at 40-50 kDa (associated to vicilin subunits) was of similar intensity for the extruded sample compared to the raw sample; they were slightly less intense in the cooked sample but dramatically less intense for the baked sample. Similar results were observed for a set of bands at 28-35 kDa (associated to α-subunits of legumin); bands at 12-18 kDa (associated to γ-vicilin or albumin subunits) were slightly intensified in cooked and baked samples, while the extruded sample showed the least intensity. In GL, similar band patterns were observed as for RL but with slight differences. The intensity of a set of bands at 40-50 kDa was more intense for the cooked and baked GL samples than for the corresponding RL samples, while being still less intense than for the raw and extruded GL samples. A similar pattern was observed for the bands detected at 28-35 kDa. For the set of bands at 12-18 kDa (associated to γ-vicilin), the intensity of the baked sample was more intense than for the corresponding RL sample.
GSP showed bands at 50 kDa (associated to α-subunits of legumin) and at 30 kDa (associated to β-legumin subunits) that were slightly less intense after processing; a set of 10-12 kDa (associated to albumin subunits) had a decreased intensity in cooked and baked flours. In YSP, bands at 75 kDa showed a slight decrease with processing, especially for the cooked and baked samples; a similar pattern was observed for the 50 kDa bands. Bands at 35-37 kDa (associated to lectins) had a high intensity in extruded and cooked samples but were slightly less intense in baked flour. For the 30 kDa bands, the extruded sample showed similar intensity as for the raw sample, while the cooked and baked flours had bands of lesser intensity. And lastly, 10-12 kDa bands showed less intensity for the processed samples than for the raw sample, especially for the cooked and the baked flours.
In CC, bands at 95 kDa (associated to β-subunits of conglycinin) were more intense in raw samples than for the processed samples.
For the bands at 72-75 kDa (associated to α 0 -or α-subunits of conglycinin), the bands were barely affected in the extruded sample, but bands were less intense in the cooked and baked samples. Bands at 45-50 kDa (associated to basic β-subunits of glycinin or β-subunits of conglycinin) were less intense after processing, especially for the cooked and baked flours. The set of bands at 30-35 kDa (associated to basic β-subunits of glycinin) and bands at 18-20 kDa (associated to acid α-subunits of glycinin) were barely affected by processing.
In FB, bands at 100-105 kDa (associated to legumin) showed a decrease in intensity for all the processed samples. Other sets of bands at 55-70 kDa (associated to phaseolin) were about the same intensity in cooked samples but decreased in extruded and baked flours. A band at 35 kDa (associated to α-type phaseolin) was not affected by processing, except for the extruded sample, which was less intense.
Another set of bands at 15-20 kDa (associated to β-type phaseolin or phytohemagglutinin) was barely affected, except for the extruded samples, which were less intense. RKB displayed bands at 109-115 kDa (associated to legumin), which were less intense in processed flours. A set of bands at 37-57 kDa (associated to α-type phaseolin) was less intense after processing, especially for the extruded sample. In BB, bands at 51-55 kDa (associated to phaseolin) were less intense in processed flours than in the raw sample, especially for the extruded and the baked samples; a similar pattern was observed for the bands at 6 of 18 S ANCHEZ-VEL AZQUEZ ET AL. 31 kDa (associated to α-type phaseolin). In PB, the main set of bands was observed at 53-69 kDa (associated to phaseolin), as well as bands of 30 kDa (associated to α-type phaseolin), whose intensity was decreased with processing, except for the baked sample, which was barely affected. In NB, a band at 51 kDa (associated to phaseolin) had a strong intensity, which decreased in processed flours. Bands at 31 kDa (associated to α-type phaseolin or erythroagglutinating phytohemagglutinin) were less intense after processing, especially after extrusion.
Bands at 12 kDa (associated to β-subunit of α-amylase inhibitor) were slightly less intense after processing.
Thermal processing is the most important treatment employed to 3.2 | Bioactive compounds of raw and processed pulses

| Saponins
Saponin content of raw and heat-treated pulses is shown in Figure 3a.

| Phytic acid
The phytic acid content of raw pulses reported in this study is within the range reported by Gupta et al. (2013) for some other legume seeds (2.2-23.2 mg phytic acid/g dw). El-Adawy et al. (2000) observed that the reduction in phytic acid content during the cooking F I G U R E 3 Bioactive compounds evaluated on selected raw and heat-processed pulses.
(a) Saponins content in raw and processed pulse flours (mg equivalent of saponin/100 g dw); (b) phytic acid content in raw and processed pulse flours (mg equivalent of phytic acid/100 g dw); (c) total phenolic compounds content in raw and processed pulse flours (mg equivalent of gallic acid/100 g dw). Same letters indicate nonstatistical differences (p < 0.05) between samples process occurred during the soaking step by leaching (depending on the nature of the phytates) and due to the increase of phytase activity induced by the soaking water. This is in agreement with our results.
Furthermore, Habiba (2002) attributed the decrease of phytates during traditional cooking (boiling water) to their binding with other nutrients such as proteins and minerals, limiting their bioavailability and detection by spectrophotometric methods. Patterson et al. (2016) and Rathod and Annapure (2016) reported a loss of phytic acid that was more or less significant depending on the extrusion parameters used (moisture, pressure and temperature) in different pulses. This reduction may be due to thermal degradation and/or a reactive chemical change and/or insoluble complex formation and could be enhanced by the increase of moisture in the extruder (Rathod & Annapure, 2016). Patterson et al. (2016) also observed a reduction in phytic acid after roasting, which could suggest that the baking process in our study might have the same effect. Champ (2002) explained that the typical process used to reduce phytic acid content in seeds is enzymatic degradation (phytase), germination, and fermentation, but in this study extrusion, soaking + cooking and baking processes showed interesting results for developing pulse food products with low phytic acid content.

| Polyphenols
The phenolics content is shown in Figure 3c. Phenolic compounds in raw samples showed a high variability, ranging between 0.87 and 6.93 mg of gallic acid equivalents/g dw (RL < GSP < NB < YSP < CC < PB < RKB < FB < BB < GL). Extrusion only resulted in a reduction (p < 0.05) of phenolic compounds in GSP, CC, and PB of 11%, 21%, and 9%, respectively. Meanwhile, cooking resulted in a decrease (p < 0.05) in lentils, GSP, FB, RKB, BB, and PB of between 28% and 66%. Baking showed a reduction (p < 0.05) of 12%, 19%, 17%, and 22% in GL, RKB, BB, and PB, respectively. As was observed for phytic acid, soaking/cooking resulted in the highest decrease of phenolic compounds for most of the pulses in this study.
Phenolics are mainly located in the seed coats of pulses, since they act as the first layer of protection. Therefore, phenolic concentration and diversity vary due to a wide variety of factors such as environmental conditions and varietal traits (Singh et al., 2007).
Food processing methods may modify the profile of phenolic compounds in pulses; as an example, germination and moisture treatments successfully enhanced their degradation (Khandelwal et al., 2010). This same study highlighted the varietal effect on tannin content, showing that colored seeds contain a higher tannin content, which is also observed with the phenolic content presented in this research. In addition, as observed for phytic acid, soaking seeds before cooking is known to decrease the polyphenol content, not only by leaching into the soaking water but also by activating polyphenol oxidase, which degrades the polyphenols, leading to their destruction (Khandelwal et al., 2010). Ordinary cooking treatment as described by Habiba (2002) also showed an effect on the loss of some phenolics. Ragaee et al. (2014) observed that an increase of phenolic compounds could be observed in some samples after thermal processing produced browning reactions, as was the case with extrusion (Brennan et al., 2011). This is due to the dissociation of conjugated phenolics and to the polymerization and/or oxidation reactions and the formation of new phenolics. In our work, the increase of phenolics was particularly significant in FB after extrusion, but for most of the other samples, a slight increase or decrease was observed. Additionally, Ragaee et al. (2014) noticed both effects-an increase or decrease of phenolics-after roasting, which is the process closest to baking. Additionally, phenolics from germinated or microwaveroasted black chickpea, as well as solid-state fermented lentils, showed a significant increase after these processing methods were applied (Dhull, Punia, Kidwai, et al., 2020;Kumar et al., 2020). Our results showed an increase of phenolics after baking for the samples containing a low amount of phenolics in the raw samples (RL, GSP, YSP, CC, and NB), whereas a decrease was observed in those with a high phenolic content in the raw samples (GL, RKB, BB, and PB).

| In vitro protein digestibility
IVPD of processed and raw pulses is presented in Table 1. Raw Phaseolus vulgaris beans showed between 75% and 78% protein digestibility, followed by CC, FB, lentils, and split peas with 79%, 82%, 83% to 84%, and 84% to 85%, respectively. After processing, all samples showed slight (p < 0.05) IVPD enhancement. Extrusion showed an increase of IVPD that ranged between 2.5% and 9.5%, while the increase observed for cooked samples was between 3.5% and 10.7% and between 2.2% and 8.4% for baked samples. Different studies observed an increase in protein digestibility after the application of various processing methods. Habiba (2002) and El-Adawy et al. (2000) observed an increase in digestibility after traditional cooking, microwave cooking, pressure cooking (autoclaving), and soaking alone. Protein digestibility was improved after cooking, not only due to the reduction of phytic acid and tannin content but also as discussed previously by Habiba (2002), due to the heating effect on the tertiary structure of the proteins, which allows better enzyme accessibility. El-Adawy et al. (2000) suggested that soaking activates proteases, which hydrolyze high molecular weight proteins into low molecular weight subunits, changing their conformation, which could enhance proteolysis during gastrointestinal digestion. Furthermore, Rathod and Annapure (2016) also found an increase of protein digestibility after extrusion, supporting the idea that the reduction in the content of some bioactive compounds, which can interfere with protein digestion, can have a positive effect on protein digestibility.
These authors also reported a higher protein digestibility when extrusion was used, in comparison with other processing methods, such as ordinary cooking and microwave cooking. This was also observed in this study but only for lentil, chickpea, and one bean variety (NB). (2007) obtained an increase of protein digestibility for kidney beans, as we did, after several processing techniques such as hydration, cooking, soaking + cooking, autoclaving, and germination.

Shimelis and Rakshit
T A B L E 1 Nutritional parameters of selected raw and heat-processed pulses' flours Abbreviations: AAS, amino acid score; B, baked; BV, biological value; C, cooked; E, extruded; EAA/TAA, essential amino acid/total amino acid; HT, heat treatment; IVPDCAAS, protein digestibility corrected amino acid score; PER, protein efficiency ratio; R, raw.

| Nutritional parameters
The calculated nutritional parameters are presented in  (Gilani & Sepehr, 2003). Moreover, during heat processing and/or alkaline extraction, racemization of L-amino acids and the formation of crosslinked peptide chains such as lysinoalanine are prone to be produced, resulting in a loss of lysine, cysteine, and threonine, thus reducing protein digestibility and having a negative impact on protein quality parameters (Gilani & Sepehr, 2003;Sarwar, 1997). Pastor-Cavada et al.

| Antioxidant potential of protein hydrolysates
The antioxidant capacity (AOX) evaluated in raw and processed pulse flours is summarized in Supplementary material 2. For practical purposes, the AOX is presented as the half-maximal inhibitory concentration (IC 50 ), which is the concentration required to result in a 50% AOX (Alam et al., 2013).

| DPPH
The 50% inhibitory concentration (IC 50 ) of the DPPH radical ( Figure 4a) was calculated from DPPH scavenging results. All processed pulses showed statistically lower (p < 0.05) IC 50 values when compared to those of raw samples. Raw CC presented an IC 50 = 47.77 mg/mL, making it the least effective DPPH scavenger raw sample. Nevertheless, after processing, the extruded, cooked, and baked CC showed lower IC 50 values of 6.43, 7.01, and 5.34 mg/mL, respectively. For RL, the highest decrease of IC 50 was calculated for cooked RL (>79%), while in GL, the best IC 50 was found in the extruded sample (>56%). Drastic changes for IC 50 values were observed in GSP and YSP, since the IC 50 in raw samples were 32.53 and 40.00 mg/mL, respectively. However, for processed samples IC 50 ranged between 5.91 and 2.77 mg/mL in GSP and between 5.69 and 4.49 mg/mL in YSP; thus, an increase of DPPH scavenging capacity of between 85.4% and 91.5% and between 85% and 88%, respectively, was observed. Similar results were obtained for NB, since the raw NB exhibited an IC 50 = 25.79 mg/mL and processed NB samples varied from IC 50 = 7.19-2.62 mg/mL; therefore, the scavenging capacity improved by between 72% and 90%. Other beans also showed enhanced DPPH scavenging activity after extrusion, cooking or baking processing, with values ranging between 53% and 83%.
It is known that processing methods can enhance the antioxidant properties and other health beneficial properties in pulses (Dhull, Punia, Kidwai, et al., 2020;Dhull, Punia, Sandhu, et al., 2020). Extruded powder of lentil-orange peel at 130 C and between 12% and 20% of feed moisture showed DPPH scavenging activity over 93.6% compared to non-extruded orange-lentil peel powder (Rathod & Annapure, 2016). Morales et al. (2015) and Amarowicz et al. (2009) found that the DPPH scavenging activity of pulse flours increased >90% with extrusion and the addition of some other food ingredients. Contrary to the results published by Lv et al. (2018), our results suggest that the extrusion (with slow screw speed) and cooking processing enhanced the DPPH scavenging activity in lentils.

| Metal chelating properties
Compared to DPPH scavenging capacity, the Cu 2+ chelating activity did not show a clear tendency among raw and processed samples ( Figure 4b). The IC 50 of raw RL (3.50 mg/mL) was statistically decreased (p < 0.05) after extrusion and cooking (3.12 mg/mL for both). In GL and GSP, the treatments enhanced (p < 0.05) the Cu 2+ activity, except in extruded samples, where the IC 50 showed a significant (p < 0.05) increase of >6% and >42%, respectively. A similar behavior was found in YSP and CC, since not only did extruded samples show less Cu 2+ chelating activity but also baking process showed no improvement. Also, extruded and baked YSP samples presented an IC 50 increase of approximately 32% and 11%, respectively, and 40% and 70% for extruded and baked CC, respectively. However, the Cu 2+ chelating activity was improved for almost all processed beans. Raw FB showed an IC 50 = 6.48 mg/mL; a decrease (p < 0.05) of 25% was observed for extruded FB (IC 50 = 4.82 mg/mL), and a decrease of 40% was observed for cooked and baked FB (IC 50 = 3.87 and 3.88 mg/mL, respectively). For RKB, cooking showed the highest Cu 2+ chelating improvement (IC 50 = 2.90 mg/mL, 20%). A similar value was observed in baked PB (IC 50 = 2.74 mg/mL), compared to the raw PB sample (IC 50 = 3.41 mg/mL). However, for BB, the extrusion process enhanced (p < 0.05) the Cu 2+ chelating activity up to 44%. Finally, NB only showed an enhancement (p < 0.05) for the baked sample, with an IC 50 = 2.11 mg/mL, which is 23% lower than for raw NB (IC 50 = 2.74 mg/mL).
The Fe 2+ chelating activity is shown in Figure 4c. The IC 50 (p < 0.05) was decreased by 11% in extruded RL (IC 50 = 5.12 mg/mL) and 26% for cooked RL (IC 50 = 6.19 mg/mL), in comparison to raw RL (IC 50 = 7.00 mg/mL). Meanwhile, extruded GL showed an IC 50 = 4.15 mg/mL and an IC 50 = 3.70 mg/mL for baked GL, which represents a decrease of 13% and 22%, respectively, when compared to the raw GL (IC 50 = 4.78 mg/mL). GSP and YSP showed a decrease in IC 50 (p < 0.05) by about 11% to 17% and 13% to 80%, respectively, after any heat processing applied. A decrease in IC 50 of around 9% and 24% for cooked and baked CC, respectively, was observed against the raw CC. In FB and RKB, the antioxidant capacity of extruded and cooked samples was improved by 12% to 16% in FB and 4% to 41% in RKB. The extruded BB and the baked PB showed a decrease in IC 50 by about 14% (IC 50 = 4.37 mg/mL) and 22% (IC 50 = 4.36 mg/mL), respectively, when compared to the corresponding raw samples. Statistically significant (p < 0.05) improvements of Fe 2+ chelating activity were detected for processed NB.
Chelation activity of transition metals is related to antioxidant effects, since metal ions promote oxidative damage in the human body, due to Fenton reactions (Saiga et al., 2003). Cu 2+ chelating activity seems to be somewhat higher than Fe 2+ . These results are in

| β-carotene bleaching activity
As with DPPH, β-carotene bleaching activity showed statistical improvements for the 10 processed pulses studied (Figure 4e). For all samples, extrusion showed the highest improvements of bleached β-carotene, with IC 50 decreases of between 90% and 96% compared to raw pulses. On the other hand, the IC 50 of cooked pulses showed an increased AOX from 30% to 75%, while baked pulses were enhanced by between 17% and 78% compared to raw samples.

| ORAC
As observed for DPPH and β-carotene assays, ORAC test showed statistical differences (p < 0.05) among all raw and processed pulses. Interestingly, the highest ORAC IC 50 decreases were found in extruded samples, ranging from 38% to 55%, followed by cooked RL, GL, GSP, and FB (24% to 42%) and baked YSP, CC, RKB, BB, PB, and NB (28% to 40%). Xu et al. (2007) reported ORAC values about 10 times higher in lentils than in green or yellow peas, as well as >20% lower than in beans. Processed FB, lentils, and peas did not show improvement in ORAC AOX compared to raw samples; the same effect was observed by Liu et al. (2020). The authors attributed this effect to phenolic releasing during any heat processing (boiling, pressure, microwave, and slow cooking). Moreover, in bioprocessed CC, Sánchez-Magaña et al. (2014) reported an enhancement of 64% in ORAC values, while in RKB and CC, ORAC improvements of 33% and 64% were found, respectively (Wu et al., 2012). Our results showed not so contrasting differences among raw pulses versus processed pulses.

| PCA
The PCA allowed us to detect similarities between samples and identify the main associations between variables responsible for the total variability of the data studied. The first two principal components ( Figure 5), PC1 and PC2, represented 40.27% and 17.12%, respectively, of the total variance of the data. Through the analysis of PC1, it is possible to observe ( Figure 5a): β-carotene, ORAC, and DPPH are three measurements that are similar and are especially related to factor 1 (strong negative correlation); reducing power is a measure rather independent of the previous ones and is mainly related to factor 2; Cu 2+ is a measure rather independent of the previous ones and is mainly related to factor 3; Fe 2+ is a measure rather independent of the previous ones and is mainly related to factor 4 (not illustrated); protein digestibility is mainly linked to factors 1 and 2; protein content is mainly related to factors 1, 3, and 5 (not shown); protein digestibility is inversely proportional to the IC 50 of β-carotene, ORAC, and DPPH values (negatively correlated). the Factor 1 to Factor 2 projection) ( Figure 5d); in this group, we find the "beans" in the upper part of factor 2, the "lentils" around neutrality, and the "split peas" in the negative portion; the three heat processes form a fairly well-mixed whole (no clear trend); the F1-F3 ( Figure 5e) projection confirms the grouping of grains without cooking on the negative side of the F1; this projection highlights the fact that "extruded" seems to be more to the right than the cooking and baking treatments. Therefore, the IC 50 measured by β-carotene, ORAC, and DPPH (separation according to factor 1) are markedly higher for all the varieties in raw samples than for the cooked varieties. Between the uncooked varieties, we can differentiate the families ("beans" vs "split peas" vs "lentils") according to its IC 50 measured by the reducing power method (factor 2). The F1-F3 projection indicates that the IC 50 measured by β-carotene, ORAC, and DPPH (factor 1) are slightly lower for all varieties (more to the right according to factor 1).

| CONCLUSION
This work demonstrated that heat treatments have a positive effect overall on protein quality parameters. Processing can alter protein digestibility; IVPD of all samples was higher than 75%; additionally, extrusion and cooking improved EAA/TAA, BV, and PER in most pulses; however, baked beans showed the best IVPDCAAS values.
Thus, the nutritional parameters of processed pulses indicated that these could be considered as good protein sources according to FAO/WHO (2013). Baking, cooking, and extrusion are recommended to reduce phytic acid and increase digestibility. Furthermore, the slight increase in protein digestibility could probably be explained by the low success obtained in eliminating and/or removing the bioactive compounds that are known to form complexes with proteins, leading to reduced accessibility for digestive enzymes action. The AOX was enhanced in most cases after heat treatments, especially after extrusion, which increased the β-carotene bleaching activity and ORAC values; reducing power was enhanced only in a couple of extruded and baked samples, while Cu 2+ and Fe 2+ presented a mixed behavior depending on pulse and varieties, unlike DPPH, where all treatments on the 10 pulses positively influenced the AOX. Saponins were more effectively reduced with cooking; phytic acid was decreased by extrusion and baking in beans, while for other pulses cooking and baking were more effective. Phenolics were decreased by cooking in most of cases. Finally, the IVPD was improved by all heating treatments, and IVPDCAAS was enhanced by baking in most of the pulses, but S ANCHEZ-VEL AZQUEZ ET AL.
extrusion was most effective for peas. Due to all mentioned above, choosing the best heat-processing treatment can raise the nutritional and antioxidant potentials in these selected pulses, which may be considered as excellent plant sources for developing functional and/or nutraceutical foods.

ACKNOWLEDGMENTS
The authors are grateful for the funding received from the Agriculture