Separation and identification of bioactive peptides from stem of Tinospora cordifolia (Willd.) Miers

Enzyme hydrolysates (trypsin, papain, pepsin, α-chymotrypsin, and pepsin-pancreatin) of Tinospora cordifolia stem proteins were analyzed for antioxidant efficacy by measuring (1) 1,1-diphenyl-2-picrylhydrazyl (DPPH•) radical scavenging activity, (2) 2,20-azinobis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS+) radical scavenging capacity, and (3) Fe2+ chelation. Trypsin hydrolysate showed the strongest DPPH• scavenging, while α-chymotrypsin hydrolysate exhibited the highest ABTS+ scavenging and Fe2+ chelation. Undigested protein strongly inhibited the gastrointestinal enzymes, trypsin (50% inhibition at enzyme/substrate ratio = 1:6.9) and α-chymotrypsin (50% inhibition at enzyme/substrate ratio = 1:1.82), indicating the prolonged antioxidant effect after ingestion. Furthermore, gel filtration purified peptide fractions of papain hydrolysates exhibited a significantly higher ABTS+ and superoxide radical scavenging as compared to non-purified digests. Active fraction 9 showing the highest radical scavenging ability was further purified and confirmed by MALDI-TOF MS followed by MS/MS with probable dominant peptide sequences identified are VLYSTPVKMWEPGR, VITVVATAGSETMR, and HIGININSR. The obtained results revealed that free radical scavenging capacity of papain hydrolysates might be related to its consistently low molecular weight hydrophobic peptides.


Identification of plant
The experimental plant has been systematically characterized as Tinospora cordifolia by local flora with the voucher specimen No.: B0223 at School of Bioengineering, SRM University, Chennai, India. The molecular characterization of the plant has also been done with genomic DNA isolation using matK gene. The obtained sequence was then subjected to a BLAST search of the GenBank database (NCBI, Bethesda, MD, USA) to identify the experiment plant. Alignment of the sequence plant with the similar sequences in the GenBank database revealed a high similarity (99%) to Tinospora cordifolia. Based on these results, the experimental plant was identified as a member of Tinospora cordifolia, and the sequence has been deposited in GenBank (NCBI) with Accession No: MF459677.

Extraction of T. cordifolia stem proteins
T. cordifolia plant stems were collected from SRM University campus, Kattankulathur, India. Extraction of proteins was performed according to the method described by Aranha et al. [22] with minor changes. The stems were washed with distilled water, rinsed with deionized Millipore water (18.2 mΩ), chopped into small pieces and air-dried at room temperature for 15 days to remove the residual moisture. The dried stems were then ground into fine dry powder using a mixer-grinder while maintaining at a temperature of 4˚C and then immediately stored at -20˚C for further processing. The powder was then re-suspended in deionized water at powder:liquid ratio of 1:5 (wt/vol). After stirring at 4˚C (for 2 h), the suspension was centrifuged (5000 g) for 20 min. The supernatant was subjected to salt precipitation by slowly dissolving ammonium sulfate by stirring to 80% saturation. The resulting suspension was allowed to stand overnight at 4˚C then centrifuged at 5000 g for 20 minutes. The precipitated proteins were re-suspended in a minimum amount of water and dialyzed using 3.5 kDa cut-off dialysis membrane (Thermo Scientific Snake Skin dialysis tubing, 3.5K MWCO, 16 mm ID) for 8 hours against Millipore (deionized) water. The dialyzed proteins were lyophilized (Lyodel Freeze Dryer Model: DPRG-1GH) to get dry powder which was aliquoted and stored at -20˚C. The lyophilized powder was used for further experiments and assays. Proteins quantification was done by Bradford method [23].

Preparation of in vitro T. cordifolia stem protein digests
A variety of fluids containing single or multiple enzymes were prepared to simulate the digestive activities of different regions of the gastrointestinal system in vitro. Both plant and animal digestive enzymes were used to determine the antioxidative efficacy of different enzyme digests (S1 Table).
To simulate in vitro gastric digestion, papain and pepsin enzymes were individually used for hydrolysis of the proteins. For papain digestion, 5 mg lyophilized powder of T. cordifolia stem extract was dissolved in 1ml of phosphate buffer (50 mM; pH 6.8). Papain activation solution contained 4 mg papain enzyme, 20 mM cysteine, and 50 mM EDTA in 20 ml pH 6.8 phosphate buffer [24,25]. 250 μl of papain solution (0.05 mg enzyme) was treated with a protein solution to give an enzyme/substrate ratio of 1:100, and 1500 U papain per mg proteins in the final reaction mixture. The reaction control consisted of proteins in buffer without the enzyme (1 ml, 5 mg/ml). The digestion mixture was incubated for digestion times of 30, 60, 80, 100, and 120 min in an orbital shaker (120 rpm, 37˚C) and 250 μl aliquots were withdrawn. The reaction was stopped by boiling the reaction mixture for six minutes to irreversibly denature papain enzyme. For pepsin digestion, suspensions of the T. cordifolia lyophilized powder were prepared in simulated gastric fluid (SGF)-10 mM HCl, 30 mM NaCl, and pH 2.2, at 10mg/ml. After completion of incubation (at 37˚C) for 10 min, pepsin (2000 U/mg, gastric buffer pH 2.2, 0.5 mg/ml) was added (6 ml) to yield 6000 U of pepsin/ mg of protein as a final digestion mix containing 10 mg of protein and 3.33 mg of pepsin (pepsin/ protein ratio of 1:3.33, w:w). The reaction control is comprised of proteins in SGF (0.5 ml, 10 mg/ml) without the enzyme. Samples were incubated in a shaking incubator (120 rpm, 37˚C) and aliquots (500 μl) collected after 60 and 120 min. Digestions were terminated by increasing the pH to 8 by adding 160mM of sodium carbonate (45 μl) to irreversibly inactivate pepsin [26,27].
For the simulation of in vitro intestinal digestion, trypsin and α-chymotrypsin enzymes were separately used for hydrolysis of the proteins. The suspension of lyophilized T. cordifolia stem protein extract was made in Tris-HCl buffer having pH 7.8 and 8.0, respectively, adjusted with 1 M HCl to yield a final concentration, 5 mg/ml. Solutions of enzymes were added to the protein solutions in buffer to give the following: 1) For α-chymotrypsin digestion-Tris-HCl buffer pH 7.8, 8 U bovine α -chymotrypsin per mg of protein extract; 2) For trypsin digestion -Tris-HCl buffer pH 8.0, 2000 U of porcine trypsin per mg of test protein extract [28,29]. This yielded an enzyme/substrate ratio of 1:5 (w/w) for both trypsin and chymotrypsin (i.e. 5 mg of protein incubated with 1 mg of trypsin and chymotrypsin respectively). 1 ml of protein solution (5 mg/ml) was incubated as reaction control without enzyme addition. Digestions were carried out in a shaking incubator (120 rpm, 37˚C) and aliquots (100 μl) taken out at 30, 60, 120, and 180 min for further analysis. Reactions were stopped by boiling at 100˚C for 6 min to irreversibly denature the α-chymotrypsin and trypsin in the digestion mixes.
Apart from single enzyme digestion, the digestibility of T. cordifolia proteins was checked in a simulated gastrointestinal fluid comprising of all major gastric and pancreatic enzymes. The lyophilized protein powder was exposed to the first step of digestion by incubation in SGF-10 mM HCl and 30 mM NaCl, pH 2.2, and pepsin (0.33% w/v, E/S ratio of 1:3.33). Incubation in SGF was performed with the following specifications: 666.67 U pepsin per mg of protein and was maintained in a shaking incubator (120 rpm, 37˚C). Two hours of digestion in SGF was quenched by an addition of 40 μl of 0.9 M NaHCO 3 per ml of digestion mixture to raise the pH to 7.5 followed by an addition of pancreatin (E/S ratio of 1:20). The reaction was continued for 4 hours with the same condition as for pepsin followed by boiling for 6 min to inactivate all enzymes.

SDS-PAGE analysis
Tricine-SDS-PAGE with the protein digests were performed by the Schagger (2006) method [30,31]. Samples were taken at different time-points during in vitro gastric and duodenal digestion was kept at -20˚C until SDS-PAGE analysis. Briefly, 20 μg (different sample volumes for different digests) of the digested protein aliquots (for all enzymes), undigested protein, and enzyme (reaction mixture equivalent, i.e. 0.2 μg of papain, 6.67 μg of pepsin and 4 μg of trypsin and chymotrypsin respectively) was incubated with equal volume of 1X Tricine-SDS-PAGE reducing loading dye for 15 min at 45˚C before being loaded onto the gel along with lower range Genei protein markers (6 μl) for comparison. Protein bands were visualized by silver staining method. BIORAD ChemiDoc XRS+ imaging system was used for taking the images of the gels.

Effect on intestinal enzyme activity
The effect of the T. cordifolia proteins on the hydrolyzing capacity of the intestinal enzymes, trypsin, and α-chymotrypsin was analyzed both qualitatively and quantitatively using disc diffusion and spectrophotometric assays, respectively. Gelatin-agar plates were made by adding 2.3 g of agar, 0.4% (w/v) gelatin to 100 mL of Tris-HCl buffer (10 mM; pH 8.0 for trypsin and pH 7.8 for α-chymotrypsin), and the solution was then autoclaved. The mixture was transferred into sterile Petri dishes under the sterile condition and solidified. A mixture of protein solutions at different concentrations and both trypsin and α-chymotrypsin enzymes (5 μl, 1 μg/μl; i.e. 50 U of trypsin and 0.2 U of α-chymotrypsin) in Tris-HCl buffer (pH 8.0) were pre-incubated for 1 hour and then pipetted into the respective wells. Both enzyme suspensions in buffer only were used as positive controls and Tris-HCl buffer alone was used as a negative control. For comparison of the efficacy of trypsin enzyme inhibitory action, a standard soybean trypsin inhibitor was also loaded into a new experimental gelatin plate along with all the other reaction mixtures as mentioned above for trypsin. The plates were incubated for 20 hours at room temperature. The gelatin plates were then stained using Coomassie Brilliant Blue R-250 (0.5%) staining reagent for 12 hours followed by destaining with water where the observations were recorded.
For the quantitative analysis of the effect of T. cordifolia stem proteins on both trypsin and α-chymotrypsin enzymes, spectroscopic analysis at 640 nm by Sigma's non-specific protease assay using casein as a substrate and Folin-ciocalteu's phenol reagent (FCP) (2N, Sigma-Aldrich) was adopted [32]. 500 μg trypsin enzyme solution (1 μg/μl, i.e. 5000 U; 10 mM sodium acetate buffer with 5 mM Calcium acetate; pH 8.0) was incubated with different amounts of lyophilized stem extract (50 μl sample volume) for 60 min. After pre-incubation, 5 ml of 0.65% casein in 50 mM Potassium phosphate buffer (pH 7.5) was added as a substrate and incubated at 37˚C for 60 min. the reaction was terminated by adding 110 mM cold Trichloroacetic acid (TCA; 5 ml) followed by 30 minutes incubation. The reaction mixture was centrifuged and the absorbance of the filtrate was observed at 640 nm after adding 5 ml Na 2 CO 3 and 1 ml of FCP. 50 μl of sodium acetate buffer was considered as the assay blank while 500 μg trypsin enzyme solutions were considered as a positive control. The conditions mentioned above were also maintained for α-chymotrypsin (1 μg/μl, i.e. 20 U) with a slight variation of pH of the sodium acetate buffer maintained at 7.8 to ensure optimum enzyme activity. The percentage of inhibition of the enzyme was estimated using the formula:
For DPPH radical scavenging assay, we followed the Brand-Williams et al. procedure [33] was modified and performed in 96-well plates while 0.625 μM DPPH was dissolved in HPLCgrade methanol to generate DPPH • free radicals. 20 μl aliquots of aqueous protein extracts at various concentrations (from 3 mg/ml to 0.5 mg/ml), and protein digests (2 mg/ml; digestion aliquots of 30, 60, 120, and 180 minutes) were added with 190 μl of DPPH solution diluted in methanol (160 μM, absorbance = 1 ± 0.05) and loaded onto a 96-well microplate. The mixture was shaken for a few seconds and incubated in the dark for 16 hours at room temperature. The absorbance was read at 517 nm in a microplate reader (Multiscan GO, Thermo Scientific). The radical scavenging activity of the samples was calculated as the percentage reduction of DPPH • according to the following formula: DPPH • scavenging activity of the samples was compared with the standard antioxidant that is ascorbic acid (3 mg/ml to 0.5 mg/ml) [10].
The ABTS •+ scavenging ability was estimated by following the decolorization assay by using the method of Re et al. [34] with additional modifications. An aqueous solution of 7 mM of ABTS and 2.45 mM of potassium persulphate was made and incubated in the dark (for 12-16 h) to generate ABTS •+ . The resulting dense-green colored solution was diluted with water prior to the assay to give an absorbance of 0.7 ± 0.02 at 734 nm. The RSA was then measured by using 10 μl samples (different concentrations of protein;2 mg/ml, protein digestion aliquots of 0, 30, 60, 90, 120, and 180 minutes, and 1mg/ml purified fractions) and 210 μl of the diluted ABTS •+ solution. The obtained results were shown as millimolar trolox equivalent antioxidant capacity (TEAC)/mg protein. For calculating the TEAC values a calibration curve was made using different amounts of trolox (50 to 1000 μM). Absorbance was noted at 1, 2, 5, and 10 min during the reaction. The percentage reduction of the radical cation ABTS •+ was calculated using the following formula: The TEAC values were estimated by dividing the slope value of trolox standard curve from the individual percentage reduction values of samples [35,36].
The superoxide radical (O 2 •-) scavenging activity of the FPLC fractions of the papain hydrolysates was analyzed by the pyrogallol autoxidation method of Marklund and Marklund [10,37] and was adapted for use in the 96-well microplates. The initial reaction mixture was prepared by adding 100 μl of the sample (in 50mM phosphate buffer, pH 6.9) to 180 μl of Tris-HCl buffer (50 mM; pH 8.2). The mixture was kept at 25˚C (10 min). Superoxide radicals were generated by adding 10 μl Pyrogallol (10 mM) prepared in 10 mM HCl). The absorbance of the final reaction mixture was read at 320 nm up to 4 minutes Phosphate buffer (50 mM, pH 6.9) was used as a blank. The O 2 •scavenging activity of ascorbic acid was also determined for comparison. The percentage of scavenged superoxide anion was estimated from the following formula:

Determining ferrous chelating activity
The ability of T. cordifolia aqueous extract and digests to chelate the transition metal ion Fe 2+ was calculated according to Wu et al. [38]. Samples (each 25 μL) were treated with 125 μL of 80 μM FeCl 2 (final concentration in the assay solution, 40 μM). After 3 min, the reaction was started by adding 100 μL of 0.5 mM ferrozine (Final concentration of 0.2 mM). The mixtures were incubated for 15 minutes and absorbance measured (at 562 nm) [35]. The chelation ability was determined using the following equation:

Liquid chromatography of papain digest
The purification of proteins/peptides present in the papain digests was done by utilizing the Fast Protein Liquid Chromatography [Ä KTA purifier Frac-950] using a size exclusion chromatography column (GE XK16, 16×100 mm) and superdex G-30 matrix with a separation range of MW 0.5-10 kDa. The matrix was packed into the column and equilibrated with phosphate buffer (0.05 M; pH 6.9) over 5 column volumes. After equilibration, 30 milligrams of double filtered (0.22 μm syringe filter) papain hydrolysate was loaded into the column. The different fractions were eluted out of the column at a specific flow rate (0.5 ml/min). Fractions were pooled together based on the protein and peptide maxima detected at 280 nm and 215 nm respectively and directly lyophilized and stored at -20˚C.

MALDI-TOF-MS and MS/MS
Lyophilized protein (fraction 9) was resuspended in 20 μl solution of 100 mM ammonium bicarbonate and 5% acetonitrile. The pH was adjusted to pH >7.5 by using 1M Tris-HCl (pH 8.0). 2 μl of 50 mM DTT was treated and the reaction mixture was kept at 65˚C (for 5 min). 2 μl of 100 mM iodoacetamide was included and kept for 30 minutes in dark at 30˚C. 2 μl of 100 ng/μl of stock trypsin (modified sequencing grade) was included and incubated overnight at 37˚C. The reaction was terminated by adjusting pH less than pH 6.0 by using 0.5% acetic acid. MALDI-TOF analysis was performed on the in-solution tryptic digest of the FPLC fraction. MS/MS of proteins/peptides of major peaks obtained in MS spectra of fraction 9 were performed. MS and MS/MS spectra of proteins and peptides were obtained by the positive ion mode on the MALDI-TOF/TOF Mass Spectrometer (Applied Biosystems 4700 Proteomics Analyzer, Farmingham, USA). Close external mass calibration for MS was carried out using 4700 Cal Mix (Applied Biosystems). The baseline was corrected and referred for the raw data.

Peptide sequencing
The MALDI-TOF-MS/MS data for each peptide was then processed by 'GPS explorer™ software-DeNovo explorer'. The following modifications were considered: oxidation [M] and carbamidomethyl [C].

Statistical analysis
Experiments were done with 2 or 3 independent replications. Antioxidant assays for all samples were carried out in triplicates. The obtained results were expressed in terms of mean ± standard deviation and subjected to analysis by analysis of variance (ANOVA) using Microsoft-Excel 2013 and IBM SPSS Statistics (Version 19 for Windows 8.1, SPSS Inc.) software. The significant differences (P<0.05) among individual means were calculated.

SDS-PAGE analysis of T. cordifolia digests
The extracted T. cordifolia stem proteins were digested using both plant and animal-derived digestive enzymes. Different protein profiles in the SDS-PAGE gels demonstrated the differential cleavage sites of the digestive enzymes on the T. cordifolia proteins. The results of SDS-PAGE indicate hydrolysis of 27 kDa protein band and accumulation of 11.3 kDa band along with the appearance of 4.9 kDa band demonstrating partial digestion of T. cordifolia stem proteins by papain enzyme (Fig 1(A)), as observed in the weak proteolytic effect of papain on commercial whey proteins [39]. The proteolytic effect of papain on the T. cordifolia proteins was more distinct than that of the other proteolytic enzymes used at higher than physiological ratios. While in-vitro simulation of proteins with trypsin enzyme (E:S = 1:5) showed the appearance of a new band at 23.8kDa molecular range in SDS-PAGE (Fig 1(B)), no profile change was observed for simulated digestion using α-chymotrypsin (E:S = 1:5, 8U/ mg protein) as indicated in Fig 1(C). These results are in contrast withμ-and m-calpain digestion with trypsin and chymotrypsin indicating the enzyme inhibitory effect of T. cordifolia stem proteins [29].
The SDS-PAGE analysis of T. cordifolia proteins treated with pepsin (at E/S = 1: 3.33 and 1:6.67) indicated limited hydrolysis (Fig 1(D)) even though the enzyme/substrate ratio was higher than that resembling the physiological ratio (1:20, wt/wt, 172 U/mg). This indicates that the protein resistance to pepsin digestion is comparable with flaxseed and egg white ovalbumin digestion mimics the physiological conditions [40]. The SDS-PAGE profile of in vitro pepsinpancreatin digests showed a decrease in intensity of 19kDa molecular weight proteins and appearance of new bands that indicate limited hydrolysis of these proteins as shown in Fig  1(E). Thus, the papain digests were applied for further purification and antioxidant activity analysis owing to their unique cleavage of the 27 kDa proteins among all the proteases used for in-vitro digestive simulation. This could be due to the combination of both endo-and exoprotease activity shown by the cysteine protease papain.

Effect of T. cordifolia proteins on intestinal enzyme activity
The zones of enzyme activity are formed on the gelatin plates due to the proteolytic activity of proteases on the gelatin protein. The decrease in zone diameter after adding protein indicates inhibition of enzyme activity. Gelatin disc diffusion plate assay was performed for the stem proteins which were incubated with trypsin and α-chymotrypsin enzymes in suitable buffers. The respective enzymes alone were used as positive control while the zone of activity was compared for analysis of the enzyme inhibitory activity of the proteins. Results indicate that with an increase in protein concentration with respect to the fixed amount of enzyme (5 μg), there was a reduction in the size of the zone of clearance for both trypsin and α-chymotrypsin enzymes. Complete reduction of the zone was observed at E/S ratio of 3.33:1 (wt/wt) for trypsin in comparison to that of the standard soybean trypsin inhibitor (Fig 2A and 2B) while the complete zone reduction was observed for the α-chymotrypsin incubated plates at enzyme/ substrate ratio of 1:1 (wt/wt) as indicated in Fig 2C. Results of the spectrophotometric assays further confirmed both trypsin and α-chymotrypsin enzyme inhibitory activities of the stem proteins. 50% inhibition of trypsin was observed at a protein concentration of 72.71 micrograms (i.e. at the enzyme/protein (wt/wt) ratio of 6.9:1) higher than the standard soybean trypsin inhibitor (STI) (S2 Table, Fig 3A). 50% inhibition of α-chymotrypsin was observed at a protein concentration of 275.17 micrograms (i.e. at the enzyme/protein (wt/wt) ratio of 1.82:1) as indicated in S2

DPPH radical scavenging activity of T. cordifolia proteins and protein digests
This method involves measuring antioxidant potential by calculating the free-radical scavenging ability of antioxidants using stable free radical like 1,1-diphenyl-2-picrylhydrazyl (DPPH). Overall, increase in percentage reduction of DPPH radical with increasing concentration of T. cordifolia stem proteins (0.75 to 3 mg/ml) was observed (p = 0.000687 < 0.05) with the highest scavenging activity at the highest amount tested (3 mg/ml) (Fig 4A). The same result was shown for wheat germ protein and alfalfa leaf protein hydrolysates [41,42]. For papain hydrolysates, a decrease in DPPH radical scavenging activity compared to unhydrolyzed protein was observed but there was no change in the scavenging activity with the changes in hydrolysis time. There was a slight decrease in DPPH radical scavenging activity of proteins after hydrolysis with trypsin enzyme (79%-68%) thus, maintaining the scavenging activity even after digestion. A slight decrease was also observed in DPPH radical scavenging activity using proteins with enhancing time of pepsin hydrolysis (51%-43%) (S3 Table, Fig 4B). Of all the hydrolysates, trypsin hydrolysate was noticed to show the enhanced DPPH radical scavenging activity at 30 min of hydrolysis time (79.04%); the value is comparable with previous studies on alfalfa protein hydrolysates generated by commercial proteases [42].

ABTS radical scavenging activity of T. cordifolia proteins and its enzyme hydrolysates
The ABTS decolorization assay is suitable for determining the antioxidant capacities of both compounds (lipophilic and hydrophilic) due to its solubility in both aqueous and organic solvents. It has been used predominantly for assessing antioxidant activity of food antioxidants and plant and animal protein hydrolysates [35,[43][44][45][46][47]. All enzyme digests of proteins are demonstrated to have antioxidant activity. The concentration of the digests was kept constant throughout the assay (2 μg/μl, 10 μl). The absorbance of the reaction mixtures (both test samples and control) for 10 minutes was monitored to determine the decoloration rate to attain a plateau after a reaction period (result not shown). Before hydrolysis, the T. cordifolia stem proteins showed increasing radical scavenging activity with an increase in the TEAC values (from 4.17 to 22 μM) with increasing assay concentrations of proteins (from 0.1 to 3mg/ml) ( Fig 5A). After hydrolysis, the values were increased notably for α-chymotrypsin hydrolysates at 30 minutes of hydrolysis followed by a slight decrease with subsequent digestion times. As indicated Bioactive peptides from stem of Tinospora cordifolia (Willd.) Miers from the p-value, trypsin hydrolysates showed a mild but significant reduction in ABTS •+ scavenging activity with increasing hydrolysis times (p = 0.027 < 0.05) (S4 Table, Fig 5B). For the pepsin hydrolysates, there was a drastic reduction in TEAC value at 30 min digestion (0.43 ± 0.02 mM TEAC/mg protein) with subsequent recovery of ABTS free radical scavenging activity with increasing digestion time. The α-chymotrypsin hydrolysates were the most efficient in scavenging the ABTS •+ . There was no linear relation between hydrolysis time and radical scavenging activity for pepsin (p = 0.49 >0.05) and α-chymotrypsin digests (p = 0.45 >0.05). Papain digest was not analyzed for ABTS scavenging activity because it formed a turbid solution with the ABTS reaction mixture.

Ferrous ion chelating activity of T. cordifolia proteins and protein hydrolysates
The catalyzing activity of transition metal ions like cuprous ions (Cu 2+ ) and ferrous ions (Fe 2+ ) in the generation of reactive oxygen species in vivo has been well documented [48]. Fe 2+ ions can catalyze the Haber-Weiss reaction to induce superoxide formation which results in the formation of hydroxyl radical. Hydroxyl radicals can oxidize many vital biomolecules that lead to lipid peroxidation and DNA damage. Fe 2+ ion is considered the primary molecule involved in the Fenton reaction which also produces the highly oxidizing hydroxyl radicals [49]. Therefore, it can be inferred that the chelation of transition metal ions by antioxidant proteins or peptides would reduce the oxidation reaction and could go a long way in ameliorating the oxidative damage. The chelating activity of the T. cordifolia stem proteins is  Fig 6A. An increase in the Fe 2+ chelating ability of the proteins (7-29%) was noticed with increasing concentrations (0.6 mg/ml-10 mg/ml) as indicated by the p-value (p = 0.000024 < 0.05). The ferrous ion chelation by the protein digests was found to be related to the enzyme used for digestion. Of all the enzyme digests, α-chymotrypsin had the highest chelating activity (4 mg/ml α-chymotrypsin digests showed 33.92% ferrous ion chelation at 120 min hydrolysis). The pepsin digestion resulted with a decrease in the Fe 2+ ion chelating activity when compared with the reaction control, this kind of activity having been reported for buckwheat protein in vitro digests [35]. For trypsin, there was a slight and continuous decrease in activity with increasing digestion time (p = 0.026 < 0.05) in comparison to the control reaction. There was no linear correlation between the time of hydrolysis and the metal ion chelating activity of chymotrypsin digest as indicated by the p-value (p = 0.79 > 0.05) (Fig 6B). The changes in the rate of ferrous ion chelation could probably because of the disruption of the iron-binding structure of the proteins due to hydrolysis or the trypsin and α-chymotrypsin inhibitory activities of the T. cordifolia proteins resulting in a decrease in a number of iron-binding sites available in the protein structure [35]. The availability of the metal ion binding functional amino acid residues like N-terminal histidine at different times of hydrolysis plays a major role in metal ion chelation [50]. The increase in ferrous ion chelating activity of the heated reaction controls of both hydrolysates) might be because of the denaturation of protein after boiling resulted in the formation of active protein fractions capable of binding metal ions. Consequently, the metal chelating ability of the protein digests may influence other mechanisms of antioxidant activity [42,51,38]. The metal chelating activity of T. cordifolia protein digests was compared with the strong chelator EDTA, whose chelating capacity was 89.94% at 0.16 mg/mL. Papain hydrolyzate formed a turbid mixture with the ferrous ion chelation reaction mixture and hence could not be tested for antioxidant activity.

Purification of proteins/peptides
The fractions obtained by FPLC (size exclusion chromatography) of papain digested T. cordifolia stem proteins were pooled into 18 fractions (S5 Table based on the protein and peptide peaks obtained in the chromatogram (S1 Fig) and were lyophilized.

Radical scavenging activity of FPLC fractions of papain digests
The strong scavenging of the relatively stable free radical, ABTS is a strong indicator of the high antioxidant effectiveness of a molecule or compound. All FPLC fractions of the papain digests exhibited ABTS free radical scavenging activity except fraction 18. Among all the fractions, fraction 9 (18.20 ± 1.37 mM TEAC/mg protein) showed the highest scavenging activity, while the TEAC values were similar for fractions 11 (4.43 ± 0.21 mM TEAC/mg protein), 15 (4.57 ± 0.02 mM TEAC/ mg protein), and 17 (4.38 ± 0.07 mM TEAC/mg protein) indicating the occurrence of similar ABTS •+ quenching activity (S5 Table, Fig 7A). The ABTS radical scavenging of the mentioned fractions (9,11,15,17) is considerably higher than that of zein hydrolysate (2.5 mM TEAC/mg protein) [52] and buckwheat protein in vitro digests (2 mM TEAC/ mg protein) [35]. The toxic superoxide radical is formed in many autoxidation reactions in vivo and the electron transport chain. It participates in numerous biological reactions generating many reactive oxygen species(ROS) like hydrogen peroxide, peroxynitrite (ONOO -), hypochlorous acid (HOCl), and the most reactive ROS in vivo, the hydroxyl radical (OH • ) via the Haber-Weiss reaction [5]. All generated free radicals are major perpetrators of oxidative damage [53]. Studies on the scavenging of these free radicals are important in understanding ROS defense mechanisms and developing more effective antioxidant drugs. Fractions 9,11,15, and 17 showed a strong superoxide radical scavenging activity, while fractions 1 and 2 demonstrated a moderate activity (S5 Table, Fig 7B). Fraction 9 has the highest superoxide radical scavenging activity completely scavenging the free radical at both 0.1 mg/ml and 0.05 mg/ml. Fraction 9 at 0.0125 mg/ml showed 64.15 ± 6.51% superoxide scavenging activity, making it a more effective radical scavenger when compared with the reported 69.15% at 2 mg/ml superoxide scavenging activity of purified chickpea hydrolysate fractions [54], and 45% at 1 mg/ml scavenging activity of mushroom-derived antioxidant peptides [55]. Fractions 15 (86.46 ± 2.51%) and 17 (91.01 ± 1.69%) had similar free radical quenching activity followed by fraction 11 (79.09 ± 2.95%), all were at 0.05 mg/ml. The superoxide radical scavenging activity of the fractions 9, 11, 15, and 17 are higher than that of alfalfa leaf peptides (67% at 0.9 mg/ml); purified whey protein fraction (50% at 9.3 mg/ml); and pepsin hydrolysate of barley hordein (20% at 0.5 mg/ml) [42,56,57]. obtained displayed that fraction 9 has strong free radical scavenging activities and has the advantage to be used as a natural antioxidant source. The SDS-PAGE analysis of the active fractions 9, 11, 15, and 17 was done to check the profile but the proteins/peptides were not detected in 14% SDS-PAGE because of the low molecular weight of these proteins/peptides (S2 Fig). Amino acid of radical scavenging peptides from active fraction 9 Numerous peptides were expected to be produced on papain hydrolysis of T. cordifolia proteins since the papain enzyme has both endo-and exopeptidase activities. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) was used to recognize the peptides present in the FPLC purified fractions of the papain digest ( S3 Fig). Although to a lesser extent, biologically active food-derived peptides are widely described by means of MALDI-MS. Antioxidant peptides in the mass range of m/z 4000-7000 have been detected in the WSE of Mungoong, a traditional Thai fishery product produced by the cephalothorax of shrimp [58]. Lunasin, a promising chemo-preventive peptide from soybean was characterized by MALDI-MS [59].  Table). Antioxidant activity of amino acids has been attributed to their structure, composition, and hydrophobicity [50,60]. Several studies have shown a high relation with certain amino acid residues and the antioxidant activity of peptides. Hydrophobic amino acids like Ala, Pro, and Phe might cause for having high free radical scavenging activity with peptides. Hydrophobic amino acids like Leu or Val, if located at the N-terminus of a peptide, are important contributors to its antioxidant activity, with the assumption that Leu has long aliphatic side-chain group favoring interactions with acyl chains of susceptible fatty acids [50,61,62]. Amino acids having aromatic residues (Trp, Tyr) can act as direct radical scavengers of electron-deficient radicals with their phenolic and indolic groups donating protons to them [63,64]. Peptides containing His and Lys have also been noted to be a high antioxidant activity [50,65]. Histidine-containing peptides have antioxidant activity through different mechanisms like hydroxyl radical scavenging, transition-metal ion chelation, and quenching of singlet oxygen and other active oxygen species via decomposition of the imidazole ring with Histidine [50,66,67]. The radical scavenging hexapeptide with the sequence Tyr-Phe-Tyr-Pro-Glu-Leu was isolated from casein hydrolysate by Suetsuna et al. [68]. The high activity was attributed to the Glu-Leu sequence (C-terminal amino acid) in the peptide. A peptide (His-Gly-Pro-Leu-Gly-Pro-Leu) was isolated from fish-skin gelatin by Mendis et al. [69] and its high antioxidant activity attributed the presence of Leu, and His and the repeating amino acid residues of Gly-Pro are the possible reason for its high activity. In addition, a low molecular weight fraction with strong superoxide radical scavenging activity was isolated from chickpea protein hydrolysate by Li et al. [54]. It was noticed that the fraction had the significantly enhanced concentrations of Phe, Ile, Leu, and Val when compared to other fractions and so it explained that the superoxide scavenging activity might be related to the amino acids with hydrophobicity. Fraction 9 has a very high superoxide and ABTS radical scavenging activity. This may be correlated to the increased levels of total hydrophobicity of amino acids in the peptides of active fraction 9. Two prominent peptides in fraction 9 (VLYSTPVKMWEPGR and VITVVA TAGSETMR) contained Val at their N-terminus, in which VLYSTPVKMWEPGR comprising the antioxidant amino acids Pro, Gly, Leu, and Trp. The purified peptide with sequence HIGI NINSR contained three residues of the hydrophobic amino acid Isoleucine. Although a definite relationship between the structure and activity cannot be established for the T. cordifolia peptides, it can be explained that the amino acids in the peptide sequences are having low molecular weight, especially in fractions 9, 11, 15, and 17 that had a strong radical scavenging activity with the hydrophobic amino acids responsible for both superoxide and ABTS radical scavenging activity.

Conclusion
T. cordifolia stem proteins showed a strong trypsin inhibitory activity (greater than standard soybean trypsin inhibitor), while it also displayed α-chymotrypsin inhibition. Both the protein extracts and protein hydrolysates showed considerable DPPH and ABTS radical scavenging activities, and moderate ferrous ion chelating activity. The strong gastrointestinal enzyme inhibition coupled with a high antioxidant activity suggests a probable prolonged antioxidant effect of the stem proteins after ingestion. The extremely high ABTS and superoxide scavenging activities of papain digest fractions indicated that the lower molecular weight peptides were efficient free radical scavengers than the higher proteins and peptides, owing to the hydrophobic and aromatic amino acids composition. Since T. cordifolia is already an important composition of many traditional Indian medicine formulations, both its purified stem proteins and the derived peptides by enzyme hydrolysis could be incorporated into food products or nutraceuticals or developed to be a safe and efficient drug for treating oxidative stress and related disorders. Gopinath.