Antioxidant activity of protein hydrolysates obtained from discarded Mediterranean fish species

ANTIOXIDANT ACTIVITY OF PROTEIN HYDROLYSATES 1 OBTAINED FROM DISCARDED MEDITERRANEAN FISH 2 SPECIES 3 Pedro J. García-Moreno1, Irineu Batista2, Carla Pires2, Narcisa M. Bandarra2, F. Javier 4 Espejo-Carpio1, Antonio Guadix1, Emilia M. Guadix1 5 1Department of Chemical Engineering, University of Granada, 18071 Granada, Spain 6 2Division of Aquaculture and Upgrading, IPMA, 1449-006 Lisbon, Portugal 7 8 9 DOI: 10.1016/j.foodres.2014.03.061 10 11  Corresponding author: Tel.: +34 958 241329; Fax: +34 958 248992; E-mail: [email protected] ABSTRACT 12 In this study, five discarded species in the Mediterranean Sea, namely sardine, horse 13 mackerel, axillary seabream, bogue and small-spotted catshark, were evaluated as raw 14 material for obtaining fish protein hydrolysates exhibiting antioxidant activity. The DH of the 15 hydrolysates ranged from 13.2 to 21.0 %, with a protein content varying from 60.7 to 89.5 %. 16 The peptide profile of all hydrolysates was very similar, except for the hydrolysate of small- 17 spotted catshark. Their lipid content was found to be between 4.6 and 25.3 %. The highest 18 DPPH scavenging activity was found for the hydrolysates of sardine and horse mackerel with 19 EC50 values varying from 0.91 to 1.78 mg protein/mL. Sardine and small-spotted catshark 20 hydrolysates exhibited the highest ferrous chelating activity with an EC50 value of 0.32 mg 21 protein/mL. Moreover, sardine and bogue hydrolysates presented the highest reducing power. 22 Finally, a total of six antioxidant peptides were theoretically identified within the structure of 23 myosin and actin proteins from sardine and small-spotted catshark. The potential antioxidant 24 activity exhibited by the hydrolysates suggests that it is feasible to obtain added-value 25 products such as natural antioxidants from these discarded species. 26 Keywords: discards, fish protein hydrolysates, antioxidant activity, DPPH, Fe2+ chelating 27 activity, reducing power 28 1. INTRODUCTION 29 Marine discards are that portion of total fish catch which is not retained for sale and returned 30 to the sea. It comprises non-target species with low commercial value, fish below minimum 31 commercial size, fish caught in excess of individual quota and damaged fish which is not 32 worthy for fishermen to keep on board (Kelleher, 2005). In the Alboran Sea, the portion of the 33 West Mediterranean Sea lying between the Spanish southern coast and the north of Morocco, 34 discards are mainly composed of commercial species such as sardine (Sardina pilchardus), 35 horse mackerel (Trachurus mediterraneus) and axillary seabream (Pagellus acarne) which 36 are dumped at the sea due to high-grading practices, quota restriction and minimal 37 commercial-size requirements. Other species such as bogue (Boops boops) and small-spotted 38 catshark (Scyliorhinus canicula) are discarded due to their reduced commercial value (García- 39 Moreno, Pérez-Gálvez, Morales-Medina, Guadix A, & Guadix EM, 2013a). 40 These practices represent an important underutilization of marine resources. Since discards 41 are generally dead or dying when returned to the sea, they also cause significant 42 environmental problems such as alterations on marine trophic chains (Bozzano & Sardà, 43 2002). In order to ensure the sustainability of EU fisheries, the EU Commission is in the 44 process of implementing a reformed Common Fisheries Policy which aims to gradually 45 implement a practice of zero-discards (EU, 2011). Nevertheless, technical measures should 46 also be applied in order to successfully meet discards bans, since discards can be reduced (i.e. 47 by improving the selectivity of the fishing gears) but cannot be completely eliminated. In this 48 sense, it seems to be of special importance the development of up-grading processes which 49 permit to obtain added-value products from this underutilized raw material. 50 In this context, discarded species in the Alboran Sea are good sources of protein, with protein 51 contents ranging from 17 to 23 % depending on the species (García-Moreno et al., 2013a). 52 Thus, enzymatic hydrolysis of their protein fraction is a convenient method for the production 53 of bioactive compounds that could be utilised in the nutraceutical and pharmaceutical fields. 54 In this regard, several fish protein hydrolysates have shown numerous bioactivities such as 55 antioxidant, antihypertensive, antithrombotic, immunomodulatory, antimicrobial, among 56 others (Je, Lee, Lee, & Ahn, 2009; Kim & Wijesekara, 2010). 57 Due to the increasing interest in finding antioxidants from natural sources which may have 58 less potential hazard than synthetic ones, research on fish protein hydrolysates exerting 59 antioxidant activity has gained an increased interest. Antioxidants are generally employed to 60 prevent lipid oxidation in foods in order to avoid the formation of toxic compounds and 61 undesirable odours and flavours (Lin & Liang, 2002). Furthermore, oxidative stress has also 62 been involved in the occurrence of several diseases such as hypertension, cancer, diabetes, 63 Alzheimer’s and aging (Hajieva & Behl, 2006). 64 In the last decade, several authors have reported a strong antioxidant activity for fish protein 65 hydrolysates obtained from different species such as black scabbardfish (Aphanopus carbo) 66 (Batista, Ramos, Coutinho, Bandarra, & Nunes, 2010), sardinelle (Sardinella aurita) 67 (Bougatef et al., 2010), saithe (Pollachius virens) (Chabeaud, Dutournié, Guérard, 68 Vandanjon, & Bourseau, 2009), yellowfin sole (Limanda aspera) (Jun, Park, Jung, & Kim 69 2004), mackerel (Scomber austriasicus) (Wu, Chen, & Shiau, 2003), and herring (Clupea 70 harengus) (Sathivel et al., 2003). However, there is little information about the production of 71 fish protein hydrolysates with antioxidant activity from discarded species in the Alboran Sea. 72 Another important aspect to consider is the choice of the enzyme since it has a great impact 73 on the release of antioxidant peptides by hydrolysis of fish protein (Laroque, Chabeaud, & 74 Guérard, 2008). The endoproteases subtilisin (EC 3.4.21.62) and pancreatic trypsin (EC 75 3.4.21.4) have previously shown good results in the production of fish protein hydrolysates 76 exhibiting antioxidant activity (Amarowicz & Shahidi, 1997; Rajapakse, Mendis, Byun, & 77 Kim, 2005; Thiansilakul, Benjakul, & Shahidi., 2007). Nevertheless, only a few studies have 78 addressed the effect of a combination of these enzymes (García-Moreno et al., 2013b). 79 Therefore, the objective of this study was to investigate the potential of five discarded species 80 in the Alboran Sea (Sardina pilchardus, Trachurus mediterraneus, Pagellus acarne, Boops 81 boops and Scyliorhinus canicula) as raw material for the production of fish protein 82 hydrolysates exhibiting antioxidant activity. For that purpose, the combined effect of 83 subtilisin and trypsin as enzymatic treatment was evaluated. 84 2. MATERIALS AND METHODS 85 2.1 Raw material 86 Raw sardine (Sardina pilchardus), horse mackerel (Trachurus mediterraneus), bogue (Boops 87 boops), axillary seabream (Pagellus acarne) and small-spotted catshark (Scyliorhinus 88 canicula), was provided by the fishing harbour of Motril (Spain) in September 2011. All fish 89 were kept in ice during the transportation and pressed in the same day. 90 2.2 Separation of protein fraction 91 The whole fish, included viscera and gonads, was preheated at 40ºC for 30 min (Digiterm 92 100, Selecta, Barcelona, Spain). Then, it was fed into an electric press (ESP-K, Sanahuja, 93 Castellón, Spain) where it was subjected to three consecutive pressing steps until attaining a 94 final pressure of 150 bar. The pressing stage permitted to reduce the moisture content and the 95 volume of the protein rich material, which also implies a diminution of the handling and 96 insulation costs. The cakes obtained from the pressing operation were grinded in a cutter (SK- 97 3, Sammic, Sevilla, Spain) and then frozen at -20ºC prior to their use as substrate for protein 98 hydrolysis. For small-spotted catshark, the grinding and homogenization of the press cake was 99 not possible due to the high resistance of its skin. Thus, muscle of this species was employed 100 as substrate for protein hydrolysis. It was obtained by de-heading, de-gutting and removing 101 the skin from the whole fish. 102 2.3 Hydrolysis procedure 103 For the enzymatic hydrolysis, two serine endoprotease enzymes were employed; one of 104 bacterial origin (subtilisin, EC 3.4.21.62) and other from an animal source (pancreatic trypsin, 105 EC 3.4.21.4), both provided by Novozymes (Denmark) as Alcalase 2.4L and PTN 6.0S, 106 respectively. The following hydrolysis conditions were studied: (a) 2 h hydrolysis with 107 subtilisin followed by 2 h hydrolysis with trypsin; (b) 2 h reaction with trypsin followed by 2 108 h incubation with subtilisin and; (c) 4h hydrolysis with simultaneous addition of both 109 enzymes. The first enzyme utilized in treatments a) and b) was not inactivated prior to 110 addition of the second enzyme. 111 The protein content of the raw material for the hydrolysis was determined by using the 112 Kjeldahl method (AOAC, 2006), with a nitrogen-to-protein conversion factor of 6.25. The 113 results, expressed as % wet base, were as follows: sardine 19.2 %, horse mackerel 20.2 %, 114 axillary seabream 23.4 %, bogue 21.9 % and small-spotted catshark 27.4 %. 115 Then, a given mass of grinded press cake was homogenised with demineralised water until 116 reaching a final volume of 200 mL. This suspension, having a protein concentration of 25 g/L, 117 was then transferred into a jacketed reactor of volume capacity 250 mL. The experiments 118 were conducted at pH 8 and 50ºC, while enzyme-protein ratio was set at 3 % (w/w) for both 119 enzymes. Protein was considered as substrate. 120 The degree of hydrolysis, defined as the percentage of the number of peptide bonds cleaved 121 compared to the total number of peptide bonds in the substrate studied, was calculated as a 122 function of the base consumption throughout the reaction employing an automatic titrator 123 (718 Stat Titrino, Metrohm, Herisau, Switzerland) (Camacho, González-Tello, Páez-Dueñas, 124 Guadix EM, & Guadix A, 2001). According to this method, the degree of hydrolysis (DH) 125 can be related to the amount of base (NaOH, 1 N) consumed to keep the pH constant during 126 the reaction, as follows (Eq. 1): 127   b P TOT DH = B·N α·m ·h ×100 (1) 128 where B (mL) is the amount of base consumed, Nb (eq/L) is the normality of the base, α is the 129 average degree of dissociation of the α-NH2 amino groups released during the hydrolysis, 130 which is dependent on the temperature and the pH, mP (g) is the mass of protein in the 131 substrate and hTOT (meq/g) is the number of equivalents of peptide bonds per gram of protein. 132 At pH 8 and temperature of 50ºC, the 88.5% of the amino groups are dissociated, while hTOT 133 was assumed to be 8.6 meq/g of protein, as reported in literature (Adler-Nissen, 1986). 134 A set of 250 hydrolysates, originated from the five species and three enzymatic treatments 135 studied and drawn at different times of reaction (0, 5, 10, 20, 30, 45, 60, 90, 120, 125, 130, 136 140, 150, 165, 180, 210 and 240 min), were evaluated in order to determine the influence of 137 DH on the DPPH scavenging activity. The samples were heated in a boiling water bath for 15 138 min to inactivate the enzyme and were filtered in order to remove the remained solids. They 139 were kept at -20 ºC until performing the analyses. 140 After completion of the hydrolysis, the final hydrolysates were also heated in a boiling water 141 bath for 15 min and filtered. Then, they were lyophilized and stored at -20 ºC until analyses 142 were performed. 143 2.4 Characterization of the hydrolysates 144 2.4.1 Protein content 145 The protein content of the lyophilized hydrolysates was determined using a FP-528 LECO 146 nitrogen analyser (LECO, St Joseph, MI, USA) calibrated with ethylenediaminetetraacetic 147 acid according to the Dumas method (Saint-Denis & Goupy, 2004). 148 2.4.2 Lipid content and lipid classes 149 The lipid content of the lyophilized hydrolysates was determined according to the method 150 described by Folch et al. (1956). Lipid classes were determined by thin-layer chromatography 151 (TLC) using hexane/diethyl ether/acetic acid (65:35:1, v/v/v) as the developing solvent 152 system. The developed plates were sprayed with 10% phosphomolybdic acid in ethanol and 153 heated at 120 ºC for 5 min. The identification of the different classes was done by comparison 154 with the standards from Sigma. For quantification purposes, the TLC plates were scanned 155 (GS-800 densitometer, Bio-Rad, Alcobendas, Spain) and analyzed with Quantity One analysis 156 software (Bio-Rad, Alcobendas, Spain). 157 2.4.3 Molecular mass distribution of hydrolysates 158 The molecular mass distribution of the fish protein hydrolysates was estimated by gel 159 filtration chromatography with a FPLC ÄKTA (Amersham Biosciences, Uppsala, Sweden) 160 using a Superdex Peptide 10/300 GL column with a UV detector at 254 nm. The eluent was 161 30% acetonitrile with 0.1% trifluoroacetic acid at a flow rate of 0.5 ml/min. A molecular mass 162 calibration curve was prepared using the following standards: ribonuclease A (13,700 Da), 163 aprotinin (6500 Da), angiotensin I (1296 Da), bradykinin (1060 Da) and triglycine (189 Da). 164 2.5 Determination of antioxidant activity 165 2.5.1 1,1-Diphenyl-2-picrylhydrazyl (DPPH) radical scavenging activity 166 DPPH radical scavenging activity of the hydrolysates was determined by two different 167 methods depending on the objective pursued. 168 In order to evaluate the evolution of the DPPH scavenging activity of the hydrolysates with 169 the degree of hydrolysis, the method reported by Brand-Williams et al. (1995) which requires 170 a low amount of sample was employed. Briefly, an aliquot of each sample (50 L) was mixed 171 with 100 L of Tris buffer solution (50 mM, pH 7.4) and with 850 L of a daily-prepared 172 solution of 1,1-diphenyl-2-picrylhydrazyl (DPPH) at 0.1 mM in methanol. The mixture was 173 then kept at room temperature in the dark for 30 min, and the reduction of DPPH radical was 174 measured at 515 nm. A blank was run in the same way by using distilled water instead of 175 sample, and sample control was also made for each sample by adding methanol instead of 176 DPPH solution. Then, DPPH radical scavenging activity was calculated according to Eq. 2: 177 sample sample_control blank (A -A ) DPPH inhibition (%) = 1- × 100 A    (2) 178 For the determination of the EC50 value of the final hydrolysates, which is defined as the 179 concentration of hydrolysate (mg protein/mL) needed to inhibit DPPH activity by 50%, the 180 method described by Picot et al. (2010) was used. This method is in the line of most of the 181 methods used to measure the DPPH scavenging activity because they employ the same 182 amount of sample as volume of DPPH solution. It was used for the final hydrolysates since 183 enough volume of these samples was available. In short, a volume of 1 mL of each protein 184 hydrolysate having different protein concentrations (0.2–5 mg/mL) was added to 1 mL of 0.1 185 mM DPPH in methanol. The mixture was shaken for 1 h at 25 ºC in the dark. Then, the 186 absorbance of the reaction mixture was measured at 517 nm. A blank was run in the same 187 way by using distilled water instead of sample, and sample control was also made for each 188 sample by adding methanol instead of DPPH solution. Triplicate measurements were carried 189 out for each sample and DPPH scavenging activity was also calculated by Eq. 2. 190 2.5.2 Reducing power 191 The reducing power of fish protein hydrolysates samples was determined according to the 192 method of Oyaizu (1992). Two mL of each hydrolysate at different protein concentrations (3– 193 20 mg/mL) were added to 2 mL of 0.2 mM phosphate buffer (pH 6.6) and 2 mL of 1% 194 potassium ferricyanide. The reaction mixture was incubated at 50 ºC for 20 min and then 2 195 mL of 10% TCA were added. The mixture was centrifuged at 1500×g for 10 min. A 2 ml 196 aliquot of the supernatant was mixed with 2 mL distilled water and 0.4 mL of 0.1% ferric 197 chloride. The absorbance of the resulting solution was recorded at 700 nm after 10 min. An 198 equivalent volume of distilled water instead of sample was used as control. Analyses were 199 carried out in triplicate. 200 fact to the formation of more or less peptides with accessible hydrophobic regions which are 342 more lipid-binding than those with the hydrophobic regions embedded in the interior. 343 In order to better characterize the hydrolysates, the lipid class composition was determined for 344 the lyophilized samples obtained with the sequential treatment subtilisin plus trypsin . For the 345 hydrolysates produced from press cakes, it was found that triacylglycerols represented the 346 major lipid class. They were around 75.0 % for sardine and horse mackerel and 45.0 % for 347 axillary seabream and bogue. In the hydrolysates from these species free fatty acids were the 348 second most important lipids. Its content was species dependent, but it was considerably 349 higher in axillary seabream and bogue hydrolysates, approximately 35.0 %. This high content 350 of free fatty acids is detrimental for the oxidative stability of the hydrolysates, since free fatty 351 acids are even more prone to oxidation than esterified fatty acids (Aidos, Van der Padt, Boom 352 & Luten, 2001). Considering cholesterol content, it ranged from 3.9 % for sardine to 14.7 % 353 for bogue. Phospholipids content of these hydrolysates was practically constant and it was 354 around 9.0 %. On the other hand, phospholipids were the most abundant lipids for small- 355 spotted catshark muscle hydrolysate, 54.4 %, followed by cholesterol, 35.2 %. This finding 356 may be due to the fact than in lean fish, phospholipids make up most of the lipids of the cell 357 (Liang & Hultin, 2005). Although antioxidant activity has been previously described for 358 phospholipids, a recent study reported that it was insignificant when compared to other 359 antioxidants such as BHT, EDTA and ascorbic acid (García-Moreno, Horn & Jacobsen, 360 2014). Conversely, triacylglycerols and free fatty acids contents of this hydrolysate were 361 practically negligible, 5.0 and 5.4 % respectively. Similarly, Daukšas et al. (2005) reported 362 that triacylglycerols content of cod hydrolysates depended on the raw material and enzyme 363 employed, varying from 37 to 88 %. These authors also reported phospholipid content for cod 364 hydrolysates of 59 %, which is in the range of the content found in small-spotted catshark 365 hydrolysate. 366 The molecular mass size distribution of the hydrolysates obtained from the press cakes by the 367 three enzymatic treatments was very similar (Fig. 3). In general, the gel filtration profile of 368 the hydrolysate samples indicated hydrolysis of the fish proteins into small molecular mass 369 peptides (<1000 Da) and free amino acids. With regard to the small-spotted catshark 370 hydrolysate, it showed a rather different peptide profile from the hydrolysates prepared from 371 the press cake of the bony fish. Although it also had most of the peptides with a molecular 372 mass below 1000 Da, it presented two narrow peaks at short (7-8.5 mL) and long (23.5-25.5 373 mL) elution volume which were not found in the other hydrolysates (dotted squares in Fig. 3). 374 3.4 Antioxidant activity of final hydrolysates 375 It is widely known that antioxidants can act by different mechanisms. Therefore, the use of 376 various methods for the evaluation of the antioxidant activity is recommended (Frankel & 377 Meyer, 2000). 378 The DPPH radical scavenging and the Fe2+ chelating activities of the 15 final hydrolysates 379 were measured as a function of the protein concentration in order to determine their EC50 380 values (Table 3). It was observed that sardine and horse mackerel hydrolysates presented the 381 highest DPPH scavenging activity with EC50 values ranging from 0.91 to 1.78 mg protein/mL. 382 They were followed by axillary seabream and bogue hydrolysates with EC50 values varying 383 from 1.94 to 2.91 mg protein/mL, whereas small-spotted catshark hydrolysates presented the 384 lowest DPPH scavenging-activity with EC50 values in the range of 3.82 - 4.45 mg protein/mL. 385 These results are in line with the EC50 value reported for rohu egg protein hydrolysates, 1.5 386 mg hydrolysate/mL, (Chalamaiah et al., 2013), but are considerable lower than the EC50 value 387 found for protein hydrolysates from toothed ponyfish muscle, 25 mg hydrolysate/mL, 388 (Klomklao, Benjakul, & Kishimura, 2013) and for hake protein hydrolysates when 389 concentrations up to 30 mg hydrolysate/mL did not reach a DPPH inhibition of 50 % (Pires, 390 Clemente, & Batista 2013). 391 In most of the cases, significant differences were found in the DPPH scavenging activity of 392 the hydrolysates when employing different enzymatic treatments (Table 3). In general, the 393 sequential addition of trypsin plus subtilisin resulted in the hydrolysates with the lowest 394 EC50 values. It may be due to the fact that adding subtilisin as second enzyme may favor the 395 cleavage at the C-terminal of hydrophobic residues which can contribute to the DPPH 396 inhibitory activity (Je, Lee, Lee, & Ahn, 2009). 397 The capacity to bind transition metals is also a useful indication of antioxidant activity. It is 398 due to the fact that transition metal ions, Fe2+ and Cu2+, catalyze the generation of reactive 399 oxygen species such as hydroxyl radical (OH·) which initiates lipid peroxidation (Stohs & 400 Bagchi, 1995). Sardine and small-spotted catshark hydrolysates exhibited the highest ferrous- 401 chelating activity with EC50 values of 0.32 mg protein/mL (Table 3). The hydrolysates 402 prepared from horse mackerel, axillary seabream and bogue presented slightly higher EC50 403 values ranging from 0.42 to 0.63 mg protein/mL. They showed higher ferrous binding 404 capacity than those of silver carp hydrolysates produced with Flavourzyme, which exhibited 405 only 60 % at a concentration of 5 mg hydrolysate/mL (Dong et al., 2008). Nonetheless, they 406 were in the line of EC50 values obtained by Ktari et al. (2012) for zebra blenny protein 407 hydrolysates, which ranged from 0.15 to 0.25 mg hydrolysate/mL depending on the 408 enzymatic treatment used. 409 The metal binding capacity of protein hydrolysates is generally attributed to their content in 410 effective sites capable of chelating metal ions (Ovissipour et al., 2013). In this sense, histidine 411 containing peptides have been reported to exhibit metal chelating activity through their 412 imidazole ring (Bougatef et al., 2009). Considering the different enzymatic treatments 413 assayed, it was revealed that for axillary seabream, bogue and small-spotted catshark, the 414 simultaneous addition of subtilisin and trypsin led to the hydrolysates with the lowest metal 415 chelating activity (Table 3). This fact may be caused by differences in the structure of the 416 peptides in these hydrolysates (Thiansilakul, Benjakul, & Shahidi., 2007). However, further 417 investigations are required in order to obtain more information about the amino acid 418 sequences of the active peptides which can confirm it. 419 The reducing power assay is another common test employed to determine the antioxidant 420 activity of fish protein hydrolysates. Particularly, this method evaluates the capacity of fish 421 protein hydrolysates to act as a reducing agent (Batista, Ramos, Coutinho, Bandarra, & 422 Nunes, 2010). As observed in Fig. 4, the reducing power of all the hydrolysates increased 423 with the concentration of protein. A similar trend was reported in previous studies (Ktari et 424 al., 2012; Pires, Clemente, & Batista 2013). From Fig. 4, it was observed that sardine and 425 bogue hydrolysates exhibited the highest reducing power independently of the enzymatic 426 treatment, whereas small-spotted catshark hydrolysates presented the lowest. This fact may be 427 probably due to the differences observed in the molecular mass profile of these hydrolysates 428 (Fig. 3), as well as the amino acid composition of their peptides (Theodore, Raghavan, & 429 Kristinsson, 2008). It also should be mentioned that the simultaneous addition of subtilisin 430 plus trypsin was the least appropriate enzymatic treatment for the production of fish protein 431 hydrolysates with a high reducing power (Fig. 4). The results obtained in this work were 432 similar to that reported for black scabbardfish (Batista, Ramos, Coutinho, Bandarra, & Nunes, 433 2010), higher than those found for hake by products (Pires, Clemente, & Batista 2013), but 434 considerably lower than those obtained for zebra blenny (Ktari et al., 2012) and sardinelle 435 (Bougatef et al., 2010). 436 3.5 Theoretical identification of antioxidant peptides 437 A total of six sequences of previously reported antioxidant peptides, namely YA, PR, HH, 438 EL, VKV and KD, were identified within the proteins studied. The dipeptides YA and PR 439 previously identified by Tang et al. (2010) and Wang et al. (2008), respectively, showed good 440 DPPH scavenging properties. Both peptides had been found in the primary structure of 441 sardine myosin, which point toward both peptides could contribute to the good values of 442 DPPH radical scavenging activity found for sardine hydrolysate. Other dipeptides with good 443 radical scavenging properties previously found in the sardine proteins studied were HH 444 encrypted also in myosin (Chen, Muramoto, Yamauchi, Fujimoto & Nokihara, 1998) and EL 445 which was found in both myosin and beta-actin (Suetsuna, Ukeda, & Ochi, 2000). Apart from 446 these peptides with scavenging properties, two more sequences, VKV and KD, are encrypted 447 in myosin and beta-actin of the sardine. These peptides were identified as antioxidant peptides 448 (Suetsuna & Ukeda 1999) using the methyl linoleate model. With respect to small-spotted 449 catshark, only four of the previous sequences (KD, YA, HH and EL) where identified within 450 the Beta-actin sequence. Almost all the peptides identified in sardine and small-spotted 451 catshark proteins, presented a size 300-200 Da. 452 Although these myosin and actin-derived peptides may be present 453 within the sardine and small spotted catshark hydrolysates, further 454 studies (e.g. in silico studies) are required in order to confirm that 455 these peptides are the responsible of the observed antioxidant 456 activity.4. CONCLUSIONS 457 This work denoted that it is feasible to produce fish protein hydrolysates exhibiting strong in 458 vitro antioxidant activity when employing press cakes of discarded species as raw material. 459 These hydrolysates, with less potential hazard than synthetic antioxidants, are suitable 460 products to be used by the food industry in order to prevent lipid oxidation. They presented a 461 varying protein (60.7-89.5 wt%) and lipid (4.6-25.3 wt%) content depending on the species. 462 The molecular mass profiles indicated that the hydrolysates were mainly constituted of small 463 peptides, below 1000 Da, which have been reported to contribute to their antioxidant activity. 464 The highest DPPH scavenging activity was found for the hydrolysate of sardine produced by 465 the enzymatic treatment trypsin plus subtilisin , with an EC50 value of 0.91 mg protein/mL. 466 Sardine and small-spotted catshark hydrolysates exhibited the highest ferrous chelating 467 activity, with EC50 value of 0.32 mg protein/mL, except for the small-spotted catshark 468 hydrolysate obtained by the simultaneous addition of subtilisin and trypsin. In terms of 469 reducing power, sardine and bogue hydrolysates presented the highest electron donating 470 capacity, with an absorbance higher than 0.8 at a concentration of 20 mg protein/mL. 471 Therefore, it can be concluded that, among the five species evaluated, sardine is the discarded 472 fish species with the highest potential for the production of fish protein hydrolysates with 473 antioxidant activity. Besides, six sequences of antioxidant peptides previously described in 474 literature (YA, PR, HH, EL, VKV and KD) were also found to be encrypted in the structure 475 of myosin and actin proteins from sardine. 476 5. ACKNOWLEDGEMENTS 477 This work was supported by the Spanish National Plan I+D+i (projects CTQ2008-02978 and 478 CTQ2011-23009) and by Santander Bank. P.J. García-Moreno acknowledges a FPI grant 479 from the Spanish Ministry of Science and Innovation and an international exchange grant for 480 young researchers from the Santander Bank. 481 6. REFERENCES 482 Adler-Nissen, J. (1986). Enzymic hydrolysis of food proteins. London: Elsevier Applied 483 Science Publishers. 484 Aidos, I., Van der Padt, A., Boom, R.M. & Luten, J.B. (2001). Upgrading of maatjes herring 485 byproducts: production of crude fish oil. Journal of Agriculture and Food Chemistry, 49, 486 3697-3704.Amarowicz, R., & Shahidi, F. (1997). Antioxidant activity of peptide fractions of 487 capelin protein hydrolysates. Food Chemistry, 58, 355-359. 488 AOAC. (2006). Official Methods of Analysis. Washington DC: Association of Official 489 Analytical Chemists. 490 Batista, I., Ramos, C., Mendonca, R., & Nunes, M. L. (2009). Enzymatic hydrolysis of 491 sardine (Sardina pilchardus) by-products and lipid recovery. Journal of Aquatic Food 492 Product Technology, 18, 120-134. 493 Batista, I., Ramos, C., Coutinho, J., Bandarra, N. M., & Nunes, M. L. (2010). 494 Characterization of protein hydrolysates and lipids obtained from black scabbardfish 495 (Aphanopus carbo) by-products and antioxidative activity of the hydrolysates produced. 496 Process Biochemistry, 45, 18-24. 497 Bougatef, A., Hajji, M., Balti, R., Lassoued, I., Triki-Ellouz, Y., & Nasri, M. (2009). 498 Antioxidant and free radical-scavenging activities of smooth hound (Mustelus mustelus) 499 muscle protein hydrolysates obtained by gastrointestinal proteases. Food Chemistry, 114, 500 1198-1205. 501 Bougatef, A., Nedjar-Arroume, N., Manni, L., Ravallec, R., Barkia, A., Guillochon, D., & 502 Nasri, M. (2010). Purification and identification of novel antioxidant peptides from 503 enzymatic hydrolysates of sardinelle (Sardinella aurita) by-products proteins. Food 504 Chemistry, 118, 559-565. 505 Bozzano, A., & Sardà, F. (2002). Fishery discards consumption rate and scavenging activity 506 in the northwestern Mediterranean Sea. ICES Journal of Marine Science, 59, 15-28. 507 Brand-Williams, W., Cuvelier, M.E., & Berset, C. (1995). Use of a free radical method to 508 evaluate antioxidant activity. Lebensmittel,Wissenschaft und Technologie, 28, 25-30. 509 Camacho, F., González-Tello, P., Páez-Dueñas, M. P., Guadix, E. M., & Guadix, A. (2001). 510 Correlation of base consumption with the degree of hydrolysis in enzymic protein hydrolysis. 511 Journal of Dairy Research, 68, 251-265. 512 Chabeaud, A., Dutournié, P., Guérard, F., Vandanjon, L., & Bourseau, P. (2009). Application 513 of response surface methodology to optimise the antioxidant activity of a saithe 514 (Pollachius virens) hydrolysate. Marine Biotechnololgy, 11, 445-455. 515 Chalamaiah, M., Jyothirmayi, T., Bhaskarachary, K., Vajreswari, A., Hemalatha, R., & 516 Dinesh Kumar, B. (2013). Chemical composition, molecular mass distribution and 517 antioxidant capacity of rohu (Labeo rohita) roe (egg) protein hydrolysates prepared by 518 gastrointestinal proteases. Food Research International, 52, 221-229. 519 Chen, H. M., Muramoto, K., Yamauchi, F., Fujimoto, K., & Nokihara, K. (1998). 520 Antioxidative properties of histidine containing peptides designed from peptide fragments 521 found in the digests of a soybean protein. Journal of Agricultural and Food Chemistry, 46, 522 49–53. 523 Daukšas, E., Falch, E., Šližyte, R., & Rustad, T. (2005). Composition of fatty acids and lipid 524 classes in bulk products generated during enzymic hydrolysis of cod (Gadus morhua) by- 525 products. Process Biochemistry, 40, 2659-2670. 526 Decker, E. A., & Welch, B. (1990). Role of ferritin as a lipid oxidation catalyst in muscle 527 food. Journal of Agriculture and Food Chemistry, 38, 674-677. 528 Dong, S., Zeng, M., Wang, D., Liu, Z., Zhao, Y., & Yang, H. (2008). Antioxidant and 529 biochemical properties of protein hydrolysates prepared from silver carp 530 (Hypophthalmichthys molitrix). Food Chemistry, 107, 1485-1493. 531 EU (2011). Proposal for a Regulation of the European Parliament and of the Council on the 532 Common Fisheries Policy COM (2011) 425. 533 Folch, J., Lees, M., & Stanley, G. H. S. (1957). A simple method for the isolation and 534 purification of total lipids from animal tissues. The Journal of Biological Chemistry, 226, 535 497-509. 536 Frankel, E. N., & Meyer, A. S. (2000). The problems of using one-dimensional methods to 537 evaluate multifunctional food and biological antioxidants. Journal of the Science of Food 538 and Agriculture, 80, 1925-1941. 539 García-Moreno, P. J., Pérez-Gálvez, R., Morales-Medina, R., Guadix, A. & Guadix, E.M. 540 (2013a). Discarded species in the West Mediterranean Sea as sources of Omega-3 PUFAs. 541 European Journal of Lipid Science and Technology, 115, 982-989. 542 García-Moreno, P.J., Pérez-Gálvez, R., Espejo-Carpio, F.J., Muñío, M.M., Guadix, A. & 543 Guadix, E.M. (2013b). Lipid Characterization and Properties of Protein Hydrolysates 544 Obtained From Discarded Mediterranean Fish Species. Journal of the Science of Food and 545 Agriculture, 93, 3777-3784. 546 García-Moreno, P.J., Horn, A.F. & Jacobsen, C. (2014). Influence of casein-phospholipid 547 combinations as emulsifier on the physical and oxidative stability of fish oil-in-water 548 emulsions. Journal of Agricultural and Food Chemistry, 62, 1142-1152. 549 Guerard, F., Guimas, L., & Binet, A. (2002). Production of tuna waste hydrolysates by a 550 commercial neutral protease preparation. Journal of Molecular Catalysis B-Enzymatic, 19, 551 489-498. 552 Hajieva, P., & Behl, C. (2006). Antioxidants as a potential therapy against age-related 553 neurodegenerative diseases: Amyloid beta toxicity and Alzheimer's disease. Current 554 Pharmaceutical Design, 12, 699-704. 555 Jao, C., & Ko, W. (2002). 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical scavenging by 556 protein hydrolyzates from tuna cooking juice. Fisheries Science, 68. 430-435. 557 Je, J. Y., Lee, K. H., Lee, M. H., & Ahn, C. B. (2009). Antioxidant and antihypertensive 558 protein hydrolysates produced from tuna liver by enzymatic hydrolysis. Food Research 559 International, 42, 1266-1272. 560 Jeon, Y. J., Byun, H. G., & Kim, S. K. (2000). Improvement of functional properties of cod 561 frame protein hydrolysates using ultrafiltration membranes. Process Biochemistry, 35, 562 471–478. 563 Jun, S. Y., Park, P. J., Jung, W. K., & Kim, S. K. (2004). Purification and characterization of 564 an antioxidative peptide from enzymatic hydrolysate of yellowfin sole (Limanda aspera) 565 frame protein. European Food Research and Technology, 219, 20-26. 566 Kelleher, K. (2005). Discards in the world’s marine fisheries. Rome: FAO Fisheries 567 Department. 568 Khantaphant, S., Benjakul, S., & Ghomi, M.R. (2011). The effects of pretreatments on 569 antioxidative activities of protein hydrolysate from the muscle of brownstripe red snapper 570 (Lutjanus vitta). LWT - Food Science and Technology, 44, 1139-1148. 571 Kim, S. K., Kim, Y. T., Byun, H. G., Nam, K. S., Joo, D. S., & Shahidi, F. (2001). Isolation 572 and characterization of antioxidative peptides from gelatin hydrolysate of Allaska Pollack 573 skin. Journal of Agriculture and Food Chemistry, 49, 1984–1989. 574 Kim, S. K., & Wijesekara, I. (2010). Development and biological activities of marine-derived 575 bioactive peptides: A review. Journal of Functional Foods, 2, 1-9. 576 Klomklao, S., Benjakul, S., & Kishimura, H. (2013). Functional properties and antioxidative 577 activity of protein hydrolysates from toothed ponyfish muscle treated with viscera extract 578 from hybrid catfish. International Journal of Food Science and Technology, 48, 1483- 579 1489. 580 Ktari, N., Jridi, M., Bkhairia, I., Sayari, N., Ben Salah, R., & Nasri M. (2012). Functionalities 581 and antioxidant properties of protein hydrolysates from muscle of zebra blenny (Salaria 582 basilisca) obtained with different crude protease extracts. Food Research International, 49, 583 747-756. 584 675 1, ribonuclease A (13,700 Da); 2, aprotinin (6500 Da); 3, angiotensin I (1296 Da); 4,bradykinin (1060 Da); 5, 676 triglycine (189 Da). 677 Fig. 3. Gel filtration chromatograms showing the molecular mass distribution of the final 678 hydrolysates obtained by the enzymatic treatment subtilisin + trypsin. 679 680 681 (a) 0 0.2 0.4 0.6 0.8 1 0 5 10 15 20 A 700 nm Concentration of protein, mg mL-1 Sardine Horse mackerel Axillary seabream Bogue Small-spotted catshark (b) 0 0.2 0.4 0.6 0.8 1 0 5 10 15 20 A 700 nm Concentration of protein, mg mL-1 Sardine Horse mackerel Axillary seabream Bogue Small-spotted catshark (c) 0 0.2 0.4 0.6 0.8 1 0 5 10 15 20 A 700 nm Concentration of protein, mg mL-1 Sardine Horse mackerel Axillary seabream Bogue Small-spotted catshark Fig. 4. Reducing power of the final hydrolysates produced by: (a) subtilisin + trypsin, (b) 682 trypsin + subtilisin and (c) subtilisin + trypsin (simultaneous). Results are average of triplicate 683 determination ± standard deviation. 684 685 Table 1. Degree of hydrolysis and protein and lipid content of the final hydrolysates 686 Hydrolysate DH, % Protein, % Lipid, % Sardine S+T 14.9 61.5 ± 0.4a 22.3 ± 0.2a T+S 13.2 60.7 ± 0.4b 25.3 ± 0.0a (S+T)o 13.7 66.4 ± 0.4c 19.7 ± 0.9b Horse mackerel S+T 19.7 67.8 ± 0.2d 17.4 ± 0.3c T+S 18.2 67.1 ± 0.2e 17.6 ± 0.3c (S+T)o 21.0 62.5 ± 0.2f 21.0 ± 0.1ab Axillary seabream S+T 17.2 73.0 ± 0.3g 8.8 ± 0.9d T+S 16.0 73.5 ± 0.3g 8.0 ± 0.6d (S+T)o 16.3 74.9 ± 0.2h 8.4 ± 0.9d Bogue S+T 17.6 75.7 ± 0.7i 8.3 ± 0.1d T+S 17.0 76.8 ± 0.2j 8.1 ± 0.0d (S+T)o 15.3 76.4 ± 0.2j 8.4 ± 0.2d Small-spotted catshark S+T 19.2 87.0 ± 0.2k 6.5 ± 1.9e T+S 18.3 89.5 ± 0.2l 4.6 ± 1.3f (S+T)o 17.3 88.7 ± 0.5m 4.8 ± 1.8f DH: degree of hydrolysis; S+T: subtilisin plus trypsin; T+S: trypsin plus subtilisin; (S+T)o: simultaneous 687 addition of subtilisin and trypsin. 688 Protein and lipid data are means of triplicate determinations ± standard deviation. Mean values within a column 689 followed by different letter mean significant differences (p<0.05). 690 691 Table 2. Lipid classes presented in the final hydrolysates produced by the enzymatic treatment 692 subtilisin+trypsin 693 Lipid class, % Sardine Horse mackerel Axillary seabream Bogue Small-spotted catshark Triacylglycerols 74.9 ± 6.6a 73.9 ± 3.7a 45.0 ± 0.5b 43.8 ± 3.5b 5.0 ± 1.1c Free fatty acids 12.2 ± 2.3a 6.5 ± 0.9b 36.2 ± 0.4c 33.8 ± 3.7c 5.4 ± 1.0b Cholesterol 3.9 ± 2.3a 9.7 ± 1.0b 10.2 ± 0.1b 14.7 ± 0.1c 35.2 ± 1.7d Phospholipids 9.0 ± 2.0a 9.9 ± 1.9a 8.6 ± 0.2a 7.7 ± 0.3a 54.4 ± 1.9b Data are means of triplicate determinations ± standard deviation. Mean values within a row followed by different 694 letter mean significant differences (p<0.05). 695 696 Table 3. EC50 values of the final hydrolysates 697 Hydrolysate EC50 (mg protein/mL) DPPH radical scavenging Fe2+ Chelating activity Sardine S+T 1.30 ± 0.12a 0.32 ± 0.01a T+S 0.91 ± 0.02b 0.32 ± 0.01a (S+T)o 1.75 ± 0.05c 0.32 ± 0.01a Horse mackerel S+T 1.63 ± 0.03d 0.42 ± 0.03b T+S 1.47 ± 0.01e 0.49 ± 0.01c (S+T)o 1.78 ± 0.08c 0.46 ± 0.01d Axillary seabream S+T 2.56 ± 0.02f 0.45 ± 0.01d T+S 2.34 ± 0.05g 0.5 ± 0.03c (S+T)o 2.44 ± 0.03g 0.51 ± 0.01c Bogue S+T 2.91 ± 0.06h 0.51 ± 0.01c T+S 2.84 ± 0.05h 0.50 ± 0.01c (S+T)o 1.94 ± 0.02i 0.63 ± 0.03e Small-spotted catshark S+T 4.45 ± 0.06j 0.32 ± 0.02a T+S 3.82 ± 0.06k 0.32 ± 0.02a (S+T)o 4.35 ± 0.11l 0.51 ± 0.01c S+T: subtilisin plus trypsin; T+S: trypsin plus subtilisin; (S+T)o: simultaneous addition of subtilisin and trypsin. 698 Data are means of triplicate determinations ± standard deviation. Mean values within a column followed by 699 different letter mean significant differences (p<0.05). 700 701