CHAPTER III Recovery of lipids from fish processing byproduct through enzymatic and fermentation techniques and their characterization.

Transcription

1 CHAPTER III Recovery of lipids from fish processing byproduct through enzymatic and fermentation techniques and their characterization Introduction Fish is a good source of proteins, vitamins, minerals and lipids. World fish production today stands leveled off around 140 MMT with the growth specially in capture fisheries is stagnant or negative in the last five years (FAO, 2010). In addition, fish is also produces by aquaculture practices all over the world. The main species produced through aquaculture are carps and tilapia (fresh water fishes) followed by mollusk, shellfish and marine crustaceans (marine fishes). In India, fresh water fishery is one of the major contributors to the animal proteins. Indian major carps (catla, rohu, mrigal), common carp and tilapia are the major fresh water fishes having economical value in India. In 2008, more than 115MMT of global fish production was used for direct consumption while approximately 27MMT ended up for non-food purpose particularly, manufacture of fish oil (FAO, 2010). Fish as such and fish oil (FO) have been considered as important sources of omega-3 fatty acids (Gbogouri et al. 2006), especially DHA and EPA which reduce the risk of coronary heart diseases (Kang and Leaf, 1996). Economic value of FO had a record increase in the past few years due to its demand and, is expected to continue to increase in the near future (Turchini et al. 2009). During fish processing considerable quantity of byproducts are generated and disposed off. Fish processing is an economical activity across the globe and it generates substantial quantity of solid waste that poses disposal problems. Considering 45% of the live weight to be waste, it was estimated that nearly 63.6 MMT of waste is generated globally out of which 2.8 MMT in India (FAO, 2010). As constrains related to environmental issues are becoming quite stringent, it is necessary to develop an optimized system for the utilization of these fishery byproducts for value addition like converting into protein and lipids. Further, in order to offset the higher demand of fish oil and meal, there is need to find sustainable alternatives resource to produce them. Fisheries byproducts (fish gut, head, viscera, blood, skin, bones, liver) from fish processing plants have been considered of low value and are disposed (Turchini et al. Page 47

2 2009). The protein and oil recovered from these byproducts can be better alternative to commercial fish meal and oil and this can minimize disposal / pollution problems associated with such wastes (Fiori et al. 2012). Reports are available on nutritional quality particularly fatty acid composition of some fresh water and marine fishes of India (Ghosh and Dua, 1997; Ackman et al. 2002; Rao et al. 2010; Sharma et al. 2010). Most of the studies on lipids are based on fish muscle, which forms a part of a staple diet ( Das and Sahu, 2001; Ackman et al. 2002; Sharma et al. 2010). Lipids mainly stored in fish in the subcutaneous tissue, belly flap, muscle tissue, liver, mesenteric tissue, and the head (Ackman 1994). In the lipids, saturated fatty acids (SFA) are dominated by palmitic (C16:0) and myristic (C14:0) acids followed by stearic acid, whereas the major monounsaturated fatty acids MUFA) are oleic and palmitoleic acids (Kolakowska et al. 2002). Meat of Indian fresh water and marine fishes have been reported to have n-3 PUFA mainly EPA, DHA and linolenic acid (Nazeer et al. 2009; Ackman et al. 2002; Sharma et al. 2010). Information on lipid content, lipid class composition and fatty acid profile in byproduct of commercial fishes is need to be carried out for their exploitation as alternative sources of fish oils. Hence, it is important to analyze the type and quantity of lipids in fisheries byproducts. Proteins, chitin and carotenoids from fish industry waste is generally recovered by employing chemical (acidic and alkali) and/or biological (enzymatic) methods (Rao et al. 2000; Healy et al. 2003; Rao and Stevens, 2006). As chemical methods are not ecofriendly, biological methods using lactic acid bacteria and enzymes are gaining importance. In case of lipid recovery, use of chemical ( acids or alkali) and physical methods (thermal processing) may destroy the cis n-3 PUFA by oxidation. Fermentation and enzymatic hydrolysis can be effective and eco-friendly approaches for simultaneous recovery of lipids and proteins from fish processing waste. Further, the conventional method is an energy intensive process compared to biotechnological approach. Utilization of these lipids and proteins may provide a means for value addition to the fish processing waste. Enzymatic hydrolysis of fresh water fish viscera has been done for obtaining protein hydrolysate with higher degree of hydrolysis (Bhaskar et al. 2008). There are some reports on recovery of chitin, protein and carotenoid from shrimp processing waste by enzymatic hydrolysis (Synowiechi and Al-Khateeb, 2000; Holanda Page 48

3 and Netto. 2007) and fermentation (Rao and Stevens, 2006; Bhaskar et al b; Khanfari et al. 2008). Available literature demonstrate not many studies are available on the effect of fermentation and enzymatic hydrolysis on recovery of lipids from the fresh water fisheries byproducts. Few reports are available on recovery of lipids by enzymatic hydrolysis of fish viscera (Egidijus et al. 2005, Rasa et al. 2005). B y adopting fermentation approach for recovery of oil from fish waste, it would also be possible to recover other functional ingredients such as protein hydrolysate and collagen, which have numerous biomedical applications. Lactic acid fermentation of fish visceral waste may also contribute to probiotic properties in the protein hydrolysate fraction. Hence, in this study the following aspects are carried out : (i) Elucidation of lipid classes and their fatty acid composition of marine and fresh water fish industry waste (head and visceral waste) in comparison to meat, and (ii) Simultaneous recovery of lipids and proteins from fish industry waste by fermentation and enzymatic hydrolysis. Results Characterization of lipids from different body components (meat, head and waste) of selected Indian fresh and marine water fishes of commercial importance Head, meat and waste (viscera) of five different fresh water [Rohu (Labeo rohita), Catla ( Catla catla), Comman carp ( Cyprinus carpio), Mrigal ( Cirrhinus mrigala) and Tilapia (Oreochromis niloticus)] and marine fishes - pink perch (Nemipterus japonicus), Indian mackerel (Rastrelliger kanagurta) and Indian Oil Sardine (Sardinella longiceps), Tuna ( Thunnus albacares) and Sear fish ( Scomberomorous commerson) were analyzed for total lipids and their lipid classes (neutral, glycol and phospho lipid) by open column chromatography. Total lipids and their classes were also analyzed for fatty acid composition Lipid class and fatty acid composition of fresh water fishes Total lipids : In case of fresh water fishes, the total lipid content (wet weight basis) varied from % in meat; % in head and % in visceral waste (Table 3.1). Higher levels of lipid content were observed in rohu head (17.8%) and visceral waste (27.8%) compared to other fresh water fishes. Visceral waste from all the Page 49

4 fishes except tilapia had fat content more than 9.0%. In all the fishes, the fat content was higher in head and viscera compared to meat and differed significantly between fishes (p < 0.05) and different body portions (p < 0.001). The results indicated that the lipid content in the head and viscera of fresh water fishes was in the range of 30-35% on wet weight basis. It is evident from the results that they could serve as good source of lipids. Lipid classes : The extracted lipid from head, meat and visceral waste of different fresh water fishes were separated into different lipid classes (neutral-, glyco- and phospholipid) by silica gel column chromatography and the composition is presented in Table 3.1. Neutral lipids constituted the major portion in the total lipid extract of head ( %), meat ( %) and visceral waste ( %) (Table 3.2). No significant difference (p > 0.05) was observed in neutral lipid co ntent in lipid extract from different fishes except for common carp, but was significant (p < 0.05) between different body portions except mrigal. Glycolipids level did not differ between different body portions and was in the range of % of total lipids. Phospholipids content was, in general higher in total lipids from meat ( %) compared to head ( %) and visceral waste ( %) but was not different (p > 0.05) between different species of fishes. The results show that neutral lipids are found to be the major lipid classes in fresh water fishes analyzed. As neutral lipid was the major lipid class in all the body component of fresh water fishes, it was further classified into different fractions (Hydrocarbons, sterolesters, triglycerides, free fatty acids, di-acylglycerol and mono-acyl glycerols) (Table 3.2). Hydrocarbons ( %) and triglycerides ( %) were the major components of neutral lipids from different fishes and constituted more than 70% of total neutral lipids (Table 3.2). Hydrocarbons content in meat, head and visceral waste ranged between %, %, % respectively, of neutral lipids and differed significantly between different fishes (p < 0.001) and body portions (p < 0.01). Page 50

5 Table 3.1. Composition of lipid classes (%) in total lipids from fresh water fishes Fishes Body parts Total lipid Neutral lipid Glycolipid Phospholipid Rohu Meat 2.9 ± 0.57 a a a a Head 17.8 ± 1.00 b b b b Waste* 27.8 ± 1.67 c c c b Catla Meat 1.2 ± 0.19 a a a a Head 8.9 ± 1.48 b b b b Waste* 9.8 ± 2.21 b b b b Mrigal Meat 1.8 ± 0.34 a a a a Head 4.5 ± 0.82 b a b b Waste* 12.0 ± 0.40 c a b b Tilapia Meat 0.8 ± 0.10 a a a a Head 5.7 ± 0.77 b b a b Waste* 4.3 ± 0.73 b b b b C. carp Meat 3.8 ± 0.22 a a a a Head 10.4 ± 0.60 b b b b Waste* 9.4 ± 0.82 b b b b C. Carp comman carp; * waste includes only visceral waste; Values are mean ± SD (n=5); Values not sharing common alphabets within the column of Individual fishes are significantly different (p<0.05). Triglycerides content in meat, head and visceral waste of fishes ranged between %, % and %, respectively. However, sterolesters and free fatty acids content did not differ between different body portions and content of triglycerides did not differ between different fishes. Page 51

6 Table 3.2. Subclasses of neutral lipids (% of neutral lipids) of fresh water fishes Fishes Parts HC SE/FAME TAG FFA DAG/Chol MAG Rohu Meat a a a a a a Head b a b a a a Waste* c a b b b a Catla Meat a a a a a a Head a a b b b b Waste* c a a a b c Mrigal Meat a a a a a a Head b a b b a b Waste* c b b b b c Tilapia Meat a a a a a a Head b a a a b b Waste* b a b a c a C carp Meat a a a a a a Head b b b b b b Waste* c a c a c c HC hydrocarbons; SE sterolesters; FAME fatty acid methyl esters; TAG triglycerides; FFA free fatty acids; Chol cholesterol; DAG diacylglyceride ; MAG monoacylglyceride,; C carp comman carp; Values are mean ± SD (n=5); Values not sharing common alphabets within the column of Individual fishes are significantly different (p<0.05); * waste includes only visceral waste. Fatty acid composition of TL, NL, GL and PL Fatty acid composition of TL from head, viscera and meat portions of five different fresh water fishes reveals the dominance of unsaturated fatty acids (USFA) (Table 3.3). Page 52

7 The fatty acid composition of visceral waste and head from these fishes shows palmitic acid (16:0) followed by oleic acid (18:1; n-9) are the dominant fatty acids (14.2 to 34.1% and 3.0 to 29.7%), respectively. Similarly, USFA in these waste components varied from to 56.36% in head and to 48.28% in viscera (Table 3.3). In case of lipids from head, palmitic acid accounted 60% of saturated fatty acids content. Lipid from head of different fishes revealed that PUFA viz., EPA ( 20:5n-3) and DHA (22:6n-3) were relatively higher in case of common carp and catla. DHA content in meat of all the freshwater fishes analyzed was >3% with highest being in common carp followed by tilapia. DHA levels in head and visceral waste of fresh water fishes ranged from and %, respectively. The fatty acid profile of the lipid extracts indicated the dominance of n-3 fatty acids over n-6 fatty acids except for the lipid extract from meat and head from rohu (Table 3.3). Further, the fatty acid composition of different body components of freshwater fishes, especially head and visceral waste indicate that these can be effectively utilized for beneficial purposes if recovered. Fatty acid compositions of different classes (NL, GL and PL) of total lipid extract from different portion of fishes were analyzed and are presented in Table 3.4 (head), Table 3.5 (meat) and Table 3.6 ( visceral waste). The results reveal the presence of higher amount of DHA in phospholipid fraction. DHA content in phospholipid fraction ranged from , % in head, meat and visceral waste respectively. Highest amount of DHA was found in phospholipid fraction of lipid extract in rohu meat (19.18%) (Table 3.6). n-3 PUFA (EPA+ DHA + linolenic acid) content were similar in visceral waste and meat of catla, tilapia and mrigal whereas higher in rohu visceral waste (16.4%) compared to its meat (10.9%). In case of common carp n-3 PUFA was lower in visceral waste (5.3%) compared to meat (7.6%). Result indicates the presence of unsaturated fatty acids in viscera and head of fresh water fishes. Hence, efforts should be made to recover the lipids from these tissues and utilize them. Utilization of visceral waste by eco-friendly methods will minimize the pollution problems associated by processing waste of fresh water fishes Page 53

8 Table 3.3. Fatty acid composition (%) of lipid extracts of body component of fresh water fishes. Fatty acids Catla Tilapia Rohu Mrigal Comman Carp H M W H M W H M W H M W H M W 14: ND : ND ND ND ND : : :0 ND ND 1.5 ND ND ND ND ND ND ND 16: :1n :1n :2 n :3 n :5 n :5 n ND :6 n SFA USFA SFA saturated fatty acid; USFA unsaturated fatty acid, H- head, M meat, W viscera waste; ND not detected; Values are mean for 5 different batches. Page 54

9 Table 3.4. Fatty acid composition (% of lipid class) of neutral, glycol- and phospholipids of head of fresh water fishes. Fatty acids Catla Tilapia Rohu Mrigal Comman Carp NL GL PL NL GL PL NL GL PL NL GL PL NL GL PL 14: ND ND : : : : ND : :1n :1n :2 n :3 n ND 0.4 ND ND ND :5 n :5 n ND ND :6 n SFA USFA SFA saturated fatty acid; USFA unsaturated fatty acid, Values are mean for 5 different batches; NL - Neutral lipid, GL - Glycolipids, PL Phospholipids, ND not detected. Page 55

10 Table 3.5. Fatty acid composition (% of lipid class) of neutral, glycol- and phospholipids of meat of fresh water fishes. Fatty acids Catla Tilapia Rohu Mrigal Comman Carp NL GL PL NL GL PL NL GL PL NL GL PL NL GL PL 14: ND ND ND ND 15:0 ND ND : : : ND ND : ND :1n :1n :2 n :3 n ND ND ND ND 0.4 ND ND ND 20:5 n :5 n :6 n SFA USFA SFA saturated fatty acid; USFA unsaturated fatty acid; NL - Neutral lipid, GL - Glycolipids, PL Phospholipids, Values are mean for 5 different batches; ND not detected. Page 56

11 Table 3.6. Fatty acid composition (% of lipid class) of neutral, glycol- and phospholipids of visceral waste of fresh water fishes. Fatty acids Catla Tilapia Rohu Mrigal Comman Carp NL GL PL NL GL PL NL GL PL NL GL PL NL GL PL 14: : ND : : : ND ND ND 16: ND ND 18:1n :1n ND :2 n :3 n ND ND ND :5 n ND :5 n ND ND 2.4 ND ND ND ND 22:6 n ND SFA USFA SFA saturated fatty acid; USFA unsaturated fatty acid; NL - Neutral lipid, GL - Glycolipids, PL Phospholipids, Values are mean for 5 different batches; ND not detected. Page 57

12 Lipid and its composition in marine fishes Total lipids : In case of marine fishes, total lipids (TL) was extracted from meat, head and visceral wastes except scales of five commercial fishes (mackerel, sardine pink perch, tuna and sear fish) and analyzed for their lipid class, neutral lipid subclasses and their fatty acid composition. TL content (wet weight basis; wwb) varied from 3.8 to 14.1% in head, 2.53 to % in meat and 2.9 to 15.10% in visceral waste (Table 3.7). Higher lipid content was found to be in Sardine waste (15.1 % on wet weight basis) whereas lowest in the case of pink perch (2.70%). In case of head portion, the lowest TL content (3.8%) was in pink perch head and highest in mackerel head (14.1%). The lipid content in waste of marine fishes was found to be lesser compared to fresh water fishes. Lipid classes : The composition of different lipid classes obtained from TL of these different fishes is presented in Table 3.7. NL constituted the major portion ( %) of the TL. Amongst all the fishes, NL was higher in head ( %) followed by waste ( %). The content of GL varied depending on the body component and it was in the range of 7.86 to %, with the highest quantity being in tuna waste. Phospholipid content was lowest among the lipid classes which ranged from %, with highest in case of tuna waste. NL are found to be the major lipid classes in fishes analyzed, as observed in lipids most of the aquatic animals. Hydrocarbons were the major constituent of NL in all the body components of marine fish analyzed except tuna and sear fish (Table 3.8). Hydrocarbons varied from 6.8 to 78.6% of NL depending on the body components and species of fish analyzed. The other two major fractions were sterol esters/fame and TAG/tocopherol which ranged from 2.03 to 20.51% and 2.67 to 20.31%, percent respectively. In case of tuna and sear fish, triglycerides were the major fraction among the neutral lipid subclasses which varied from % and % in sear fish and tuna respectively. Page

13 Table 3.7. Composition of different lipid classes (%) in total lipids from marine fishes. Fishes Parts Total lipid Neutral lipid Glycolipid Phospholipid Sardine Meat a a a a Head b a a b Waste* c a a c Mackerel Meat a a a a Head b a b b Waste* c b c b Pink perch Meat a a a a Head b a a a Waste* a b b a Tuna Meat 3.1 ± 0.5 a 62.6 ± 3.0 a 31.3 ± 4.1 a 6.0 ±1.8 a Head 6.3 ± 0.3 b 78.7 ±5.1 b 19.2 ± 2.2 b 2.0 ±0.9 b Waste* 4.1 ± 0.5 a 58.8 ±2.2 a 34.8 ± 2.4 a 6.3 ±1.3 a Seer fish Meat 6.2 ± 0.4 a 62.0 ±2.8 a 34.0 ± 4.5 a 4.0 ± 0.4 a Head 8.9 ± 1.1 b 74.5 ±4.4 b 22.4 ± 2.6 b 3.2 ± 0.2 b Waste* 3.4 ± 0.3 c 61.0 ±4.7 a 33.7 ± 0.8 a 4.4 ± 0.5 a * visceral waste; Values are mean ± SD (n=5); Values not sharing common alphabets within the column of Individual fishes are significantly different (p<0.05). Page

14 Table 3.8. Subclasses of neutral lipids (% of total neutral lipids) of different marine fishes. Fishes Parts HC SE/FAME TAG FFA DAG/Chol MAG Sardine M a a a a a a H b b a a W c a a a Mackerel M a a a a a a H a b b b a a W b a c c a a Pink perch M a a a a a a H a b a a a W b a a a a Tuna M 10.3±1.0 a 5.1 ± 0.4 a 61.1± 4.2 a 15.4±1.4 a 3.2 ± 0.9 a 4.4±0.7 a H 10.5±1.5 a 18.2 ± 1.2 a 51.2± 3.7 a 12.2±1.9 a 3.5 ± 0.7 a 4.3±0.5 a W 4.2 ± 1.1 b 11.6 ± 1.6 a 56.0±4.9 a,b 14.2±1.5 a 5.8 ± 0.3 b 7.9±1.9 b Sear fish M 6.8 ± 1.3 a 0.7 ± 0.1 a 55.6±4.5 a 27.3±2.8 a 8.0 ± 1.1 a 2.1±0.3 a H 6.8 ± 0.9 a 2.2 ± 0.5 b 56.7±4.2 a 25.4±2.2 a 7.2 ± 0.7 a 2.2± 0.3 a W 39.5±2.8 b 1.8 ± 0.6 b 29.3±1.9 b 12.6±2.3 b 8.9 ± 0.6 a 8.0±0.5 b HC hydrocarbons; SE sterolesters; FAME fatty acid methyl esters; TG triglycerides; FFA free fatty acids; DAG diacylglyceride ; MAG monoacylglyceride; Chol cholesterol; M meat; H head; W visceral waste; Value are mean ± SD (n=5). Values not sharing common alphabets within the column of individual fishes are significantly different (p<0.05). Page

15 Fatty acid composition of TL, NL, GL and PL Fatty acid composition of TL, NL, GL and PL indicated the presence of higher levels of EPA and DHA in meat, head and visceral portions (Table ) of the marine fishes analyzed. In the marine fishes analyzed, like freshwater fishes, palmitic acid (16:0, %) was the major saturated fatty acid followed by stearic acid (18:0, %) and myristic acid (14:0, %). Amongst the monounsaturated fatty acids, oleic acid (18:1; n-9, %) and palmitoleic (C16:1, %) were dominant fatty acids. Fatty acid composition of TL and various lipid classes from different species of marine fishes revealed the presence of higher amount of n-3 fatty acids (Tables ). The major n-3 fatty acids present in different component of the marine fishes analyzed were EPA (20:5n -3, %), DHA (C22:6n -3, %) and lower concentrations of linolenic acid (18:3n-3, 0 1.5%) and docosapentanoic acid (22:5n-3, 0 3.2%) (Tables ). Among n-6 fatty acid arachidonic acid (20:4n-6, %) and linoleic acid (18:2n -6, 0 1.6%) were the fatty acids present in different body component of marine fishes. Further, DHA levels were higher in PL fractions ( %) compared to NL ( %) and GL ( %) in the marine fishes analyzed. DHA content was higher in phospholipids fractions and ranged from %, %, % in head, meat and waste respectively. Results further show lipids from different body component of marine fishes found to be rich in EPA and DHA compared to fresh water fishes which are rich in linolenic acid (18:3n-3) and lower in EPA and DHA. However, the total lipid content is lower in marine fishes compared to fresh water fishes. The potential of non-meat components of marine fishes as sources of recoverable EPA and DHA, is evidenced in this study. Page

16 Table 3.9. Fatty acid composition (% of total lipids) of lipids extract of body component of marine fishes. Fatty acids Sardine Mackeral Pink perch Tuna Seer fish H M W H M W H M W H M W H M W 14: : : : : :1n :1n :2n :3n ND ND ND ND ND ND 0.6 ND 20: ND ND ND 0.5 ND 20:4n-6 ND ND ND ND ND ND ND ND ND :5n :5n ND :6n SFA USFA SFA saturated fatty acid; USFA unsaturated fatty acid, H- head, M meat, W waste mainly viscera; Values are mean for 5 different batches; ND not detected Page 62

17 Table Fatty acid composition (% of lipid class) of neutral, glycol- and phospholipids of head of marine fishes. Fatty acids Sardine Mackeral Pink perch Tuna Seer fish NL GL PL NL GL PL NL GL PL NL GL PL NL GL PL 14: ND 8.4 ND :0 ND ND ND 16: : : :1n ND ND :1n ND ND :2n ND :3n-3 ND ND ND ND 0.9 ND ND ND ND ND ND ND 20: ND 0.7 ND ND 1.7 ND ND ND 20:4n-6 ND ND ND ND ND ND ND ND ND :5n ND :5n ND :6n SFA USFA SFA saturated fatty acid; USFA unsaturated fatty acid; NL - Neutral lipid, GL - Glycolipids, PL Phospholipids, ND not detected; Values are mean for 5 different batches. Page 63

18 Table Fatty acid composition (% of lipid class) of neutral, glycol- and phospholipids of meat of marine fishes. Fatty acids Sardine Mackeral Pink perch Tuna Seer fish NL GL PL NL GL PL NL GL PL NL GL PL NL GL PL 14: : ND ND 16: : : :1n :1n ND :2n :3n-3 ND ND ND ND ND ND ND 1.7 ND ND ND ND ND 20: ND ND ND ND ND ND ND ND 20:4n-6 ND ND ND ND ND ND ND ND ND :5n :5n ND 1.4 ND ND 1.4 ND ND :6n SFA USFA SFA saturated fatty acid; USFA unsaturated fatty acid; NL - Neutral lipid, GL - Glycolipids, PL Phospholipids, ND not detected; Values are mean for 5 different batches. Page 64

19 Table Fatty acid composition (% of lipid class) of neutral, glycol- and phospholipids of waste of marine fishes. Fatty acids Sardine Mackeral Pink perch Tuna Seer fish NL GL PL NL GL PL NL GL PL NL GL PL NL GL PL 14: : : : : :1n :1n :2n ND 1.2 ND :3n-3 ND ND ND ND ND ND ND ND ND 20: ND ND ND ND ND ND ND 20:4n-6 ND ND ND ND ND ND ND ND ND :5n :5n-3 ND ND ND ND ND ND ND ND :6n SFA USFA SFA saturated fatty acid; USFA unsaturated fatty acid; NL - Neutral lipid, GL - Glycolipids, PL Phospholipids, Values are mean for 5 different batches; ND not detected. Page 65

20 Application of native LAB for simultaneous recovery of lipids and proteins from fish processing waste Results have shown that various marine and fresh water fish processing byproducts contain a reasonable quantity of lipids and hence can serve a potential source of fish oil. Further, compared to marine fishes, lipid content was significantly higher in fresh water fish visceral waste. Among fresh water carps, rohu (highest lipid content) and catla, commercially the most dominant species. Also the lipid content was higher in rohu viscera followed by catla and mrigal. As EPA+DHA content was higher in catla (8.2%) compared to mrigal (2.3%), it was taken in combination with rohu f or fermentation and enzymatic hydrolysis experiments. Hereafter, rohu and catla (1:1) mixed visceral waste used as byproduct for detail studies. Lactic acid bacteria (LAB) isolated from naturally fermented fish visceral waste (FVW) was used for fermentation of FVW for 3 days at 37±2ºC for simultaneous recovery of lipids and protein. The LAB selected for the fermentation experiments were on the basis of their proteolytic properties and antimicrobial spectrum towards wide range of pathogens. The protease produced by the LAB is capable of hydrolysing FVW to recover protein hydrolysate and lipids simultaneously. The antibacterial properties (due to bacteriocins) provide probiotic properties in the fermented protein hydrolysate. The four native LAB selected were proteolytic and had antibacterial properties against various Gram +ve and Gram -ve bacteria. They were compared with a proteolytic LAB isolated from naturally fermented tannery fleshings having broad spectrum of antibacterial properties (Amit et al. 2009). Fermentation efficiency of potential native isolates as mentioned above ( Pediococcus sp FM37, P. acidilactici NCIM5368, Enterococcus sp MW2, E. fecalis NCIM5367) were comparatively evaluated vis-à-vis with a known LAB isolate viz., Enterococcus faecium NCIM5335. Fermentation with native LAB was done after cooking the fish visceral waste to inactivate the indigenous flora and enzymes (particularly lipases and proteases). Fermentation was also carried out without adding any LAB for comparison and was considered as control. Fermentation efficiency of different LAB was analyzed with their ability to decrease the ph of the medium. There was significant decrease in ph of fermented FVW Page 66

21 using both native and standard LAB compared to control (without LAB) (Figure 3.1). Homogenized and cooked FVW (Fresh) had an initial ph of 6.3 which reduced to on fermentation depending on the LAB inoculums used (Figure 3.1 ) and highest reduction in ph was in case of P. acidilactici NCIM5368 (3.81). The decrease in ph will reduce the count of spoilage causing bacteria and also help in hydrolysis of protein. In comparison to LAB fermented FVW, reduction in ph of control (without added inoculum) was the least (5.6). However, there was no significant difference in ph reduction between native LAB and standard LAB used. The recovered oil was analyzed for their acid value (mg KOH per g oil) to evaluate their quality. It was found that acid value increased to in LAB fermented FVW from the initial value of 9 in Fresh FVW (Figure 3.1). Acid value (mg KOH/gm oil) a a b b acid value b b b b ph b b b b b c ph 0 3 Fresh FM37 MW2 Control LAB Figure 3.1. Acid value of the recovered oil and ph of the fermented mass on fermentation using different LAB. Standard LAB [ E faecium NCIM5335]; Native LAB [5368: P acidilactici NCIM5368, 5367: E faecalis NCIM5367, MW2: P acidilactici, FM37: P acidilactici FM37], Control no LAB. Values are mean ± SD (n=5); Values not sharing common alphabets within same pattern are significantly different (p<0.05). Similar to ph reduction, acid value also did similar in FVW fermented with native LAB ( ) and standard LAB (22.7). The increased acid value may possibly be Page 67

22 due to the acids formed on fermentation as well as lipolysis by indigenous/bacterial lipases that results in free fatty acids. The role of indigenous lipases in increase in acid value of recovered oil was confirmed from the fact that the acid values of oil recovered from cooked and fermented fish waste. On fermentation using different LAB, recovery of lipids (FO-LAF) ranged from % (Figure 3.2) considering the lipids extracted from fresh FVW using solvents as 100%. FVW fermented with P. acidilactici NCIM5368 and E. faecalis NCIM5367 resulted in higher recovery of FO-LAF of and % respectively (Figure 3.2). % oil recovery / DH of protein a a b a b Oil recovery a,b a,b a a,b c Degree of hydrolysis a,b c d 30 e FM37 MW2 Control Figure 3.2. Percent oil recovery and degree of hydrolysis on fermentation of fish visceral waste using different LAB. Standard LAB [ E faecium NCIM5335]; Native LAB [5368: P acidilactici NCIM5368, 5367: E faecalis NCIM5367, MW2: P acidilactici, FM37: P acidilactici FM37], Control no LAB. Values are mean ± SD (n=5); Values not sharing common alphabets within same pattern are significantly different. Page 68

23 Recovery of FO-LAF was significantly higher in native LAB ( P. acidilactici NCIM5368) compared to standard LAB (E faecium NCIM5335). In addition to recovery of oil from FVW, the process of fermentation also resulted in an aqueous portion rich in protein hydrolysate as well as a residue portion rich in collagen. Degree of protein hydrolysis ( DH) correlated well with that of FO-LAF recovery as it was higher in the case of FVW fermented with P. acidilactici NCIM5368 (58.83 %) and E. faecalis NCIM5367 (58.33 %) compared to DH in case of other native LAB and standard LAB (Figure 3.2). The least recovery of FO-LAF (71.32 %) and degree of hydrolysis (26.04 %) was in control, where no LAB was added. Fermentation with added LAB increased the oil recovery and degree of protein hydrolysis by and %, respectively, compared to control (Figure 3.3) Oil recovery Degree of hydrolysis % FM37 MW2 LAB Figure 3.3. Percentage increase in oil recovery and degree of protein hydrolysis on lactic acid fermentation of fish visceral waste as compared to control; LAB [ E faecium NCIM5335, 5368: P acidilactici NCIM5368, 5367: E faecalis NCIM5367, MW2: P acidilactici, FM37: P acidilactici FM37], LAB presumably hydrolyzes the protein-lipid complex, releasing the oil and protein. The protein released probably is further hydrolysed due to the proteolytic action of LAB as well as the lactic acid produced therein. Similar to FO-LAF recovery and DH, FVW fermented with native LAB E. faecalis NCIM5367 (66.83%) and P. acidilactici Page 69

24 NCIM5368 (62.33%) had higher protein extractability compared to standard LAB E. faecium NCIM5335 (58.16 %) (Figure 3.4). The collagen content (% of total collagen) in the residue on fermentation indicates highest recovery in case of fermentation with E. faecalis NCIM5367 (55.47%) and least in case of in E. faecium NCIM5335 (33.22%). However, maximum collagen was found in case of control (59.55%). This again clearly shows that there was least hydrolysis in case of control and that collagen to an extent can also be hydrolyzed by the proteolytic LAB. % Extractibility/ collagen a a a,b b b b a Extractibility c b a Collagen d c FM37 MW2 Control Figure 3.4. Percent extractability of protein and collagen on fermentation of fish visceral waste using different LAB. Standard LAB [ E faecium NCIM5335]; Native LAB [5368: P acidilactici NCIM5368, 5367: E faecalis NCIM5367, MW2: P acidilactici, FM37: P acidilactici FM37], Control no LAB. Values are mean ± SD (n=5); Values not sharing common alphabets within same pattern are significantly different (p<0.05). Fatty acid composition of FO-LAF There were no significant changes in fatty acid composition of FO-LAF recovered on fermentation of fish visceral wastes with native LAB compared to standard LAB (Table 3.13). The recovered FO-LAF had % of saturated, % monounsaturated and % polyunsaturated fatty acids (PUFA) (Table 3.13). Page 70

25 The major fatty acids in oil recovered on fermentation of fish visceral waste were palmitic (14:0), palmitoic (14:1), stearic (18:0), oleic (C18:1), linoleic (C18:2 n-6) and linolenic acid (C18:3 n-3). There was no significant change in fatty acid composition between FO-LAF and by solvent extraction (Fresh) (Table 3.3). The process of fermentation did not significantly affect the quality of oil recovered thereby reinforcing the value of fermentation for effective recovery of fish waste lipids. Table Fatty acid composition (%) of oil recovered on fermentation of fish visceral waste by native lactic acid bacteria. Fatty acids NCIM5335 MW2 NCIM5367 FM37 NCIM5368 Fresh 14:0 2.0 ± 0.3 a 1.4 ± 0.6 a 2.0 ± 0.4 a 2.2 ± 0.4 a 1.9 ± 0.5 a 1.5 ± 0.5 a 15:0 0.8 ± 0.2 a 1.2 ± 0.3 a 0.9±0.2 a 0.8 ± 0.2 a 0.9 ± 0.1 a 1.0 ± 0.2 a 16: ± 3.8 a 29.7 ± 3.0 a 31.4 ± 3.3 a 29.1 ± 2.9 a 30.7 ± 3.5 a 28.0 ± 3.4 a 16:1 9.9 ± 1.2 a 10.9 ± 1.3 a 10.2 ± 1.4 a 10.7 ± 0.9 a 10.1 ± 1.6 a 10.8 ± 1.2 a 18:0 5.9 ± 0.8 a 5.5 ± 0.4 a 6.0 ± 0.7 a 5.5 ± 0.6 a 5.8 ± 0.4 a 5.2 ± 0.8 a 18:1n ± 2.2 a 15.7 ± 1.2 a 15.0 ± 1.8 a 15.1 ± 1.6 a 14.7 ± 2.1 a 15.3 ± 1.9 a 18:1n ± 0.6 a 3.5 ±0.7 a 4.1 ± 0.2 a 4.2 ± 0.3 a 3.8 ± 0.4 a 3.7 ± 0.3 a 18:2n ± 1.4 a 10.2 ± 1.3 a 10.2 ± 0.9 a 10.5 ± 1.5 a 9.9 ± 1.2 a 10.7 ± 1.8 a 18:3n ± 0.4 a 7.1 ± 0.8 a 7.4 ± 1.1 a 7.2 ± 0.9 a 7.7 ± 1.2 a 7.8 ± 1.0 a 20:5n ± 0.3 a 2.5 ± 0.2 a 2.6 ± 0.4 a 2.6 ± 0.2 a 2.7± ± :5n ± 0.3 a 0.8 ± 0.1 a 0.6 ± 0.2 a 0.7 ± 0.2 a 0.8 ± 0.2 a 0.7 ± 0.1 a 22:6n ± 0.2 a 2.5 ± 0.3 a 2.7 ± 0.3 a 2.6 ± 0.2 a 2.7 ± 0.3 a 2.6 ± 0.3 a NCIM E faecium NCIM5335, NCIM Pediococcus acidilactici NCIM5368, FJ1- Enterococcus faecalis NCIM5367, FM37 - Pediococcus sp FM37, MW2 Enterococcus sp MW2; Fresh oil recovered by solvent extraction; Values are mean ± SD (n=5); Values not sharing common alphabets within same row are significantly different (p<0.05). In-vitro antioxidant properties of fermented liquor (protein hydrolysate) portion The fermentation process resulted in three distinct portions viz., lipids (top layer), aqueous middle layer and lower collagenous residue (Figure 3.5). The middle layer rich in hydrolysed portion was considered for evaluation of its antioxidant properties and the Page 71

26 same is presented in Table The protein hydrolysate from fermentation process also exhibited antibacterial properties against human pathogens like Escherichia coli, Micrococcus luteus and Salmonella enteritides as shown in Table Antioxidant activity varied with different cultures used for fermentation. The protein hydrolysate by native LAB isolates exhibited higher DPPH radical scavenging activity and total antioxidant activity. DPPH radical scavenging activity as expressed as EC 50 indicates the hydrolysate proteins in milligram that is required to scavenge 50% of DPPH radicals produced. EC 50 for DPPH scavenging was lower (1.8 mg) in case of protein hydrolysate prepared using E. faecalis NCIM5367 followed by P. acidilactici NCIM5368 (EC 50 value of 1.9 mg) and relatively higher in case of P. acidilactici MW2 (EC 50 of 2.1 mg). Table DPPH radical scavenging and total antioxidant activity of protein hydrolysate Culture code recovered on fermentation of fresh water fish visceral waste Antioxidant activities Antibacterial activities DPPH SRSA TAO ML EC SI Enterococcus sp MW2 2.1±0.2 a 1.88±0.3 a 34.9±1.8 a P. acidilactici NCIM ±0.2 a 1.15±0.2 b 41.9 ±2.9 b E faecalis NCIM ±0.1 a 1.68±0.2 a 41.4 ±2.7 b Pediococcus sp FM37 2.0±0.2 a 1.61±0.2 a 31.4 ±1.6 a E. faecium NCIM ±0.2 a 1.89±0.3 a 41.4 ±2.2 b * : ascorbic acid equivalents in microgram per mg of protein in the hydrolysate, EC 50 : effective concentration for 50% activity; Values are mean ± SD (n=5) ; ML : M luteus ; EC : E coli ; SI : S itridicus; + : > 6-mm inhibition zone ; ++ : > 12-mm inhibition zone ; +++ : > 18mm inhibition zone ; Values not sharing common alphabets within same row are significantly different (p<0.05). Total antioxidant activity ( g ascorbic acid equivalents (AAE) per mg of hydrolyzed protein) in case of protein hydrolysate from LAB fermentation ranged 34.9 to 41.9 cooked visceral waste. Superoxide radical scavenging activity was found to be higher in protein hydrolysate of P. acidilactici NCIM5368 (native LAB) which had the lowest EC 50 value of 1.15 mg protein. Superoxide radical scavenging activity was significantly higher in case hydrolysate of P. acidilactici NCIM5368 compared to standard LAB ( E faecium NCIM5335). This fraction has a potential to be used as an Page 72

27 ingredient in livestock or aquaculture feeds. Also, it can be used as a probiotic; and, use of probiotic as feed additives is been preferred over that of antibiotics as they do not exhibit any of the undesirable effect associated in the use of antibiotics viz, toxicity, allergy, residues in food, bacterial drug resists. A B Lipids Protein hydrolysate Residue Figure 3.5. Difference between fermented and non fermented FVW using LAB showing distinct phase separation of lipid and protein; A not fermented; B fermented Optimization of fermentation conditions for recovery of lipids and protein by response surface methodology Studies on fermentation ensilaging of fish visceral waste indicated that Pediococcus acidilactici NCIM5368 was found to be the better starter culture for the simultaneous recovery of lipids and protein. Hence, conditions were optimized by using response surface methodology for obtaining higher oil recovery and degree of protein hydrolysis. Three independent variables viz., inoculum level (X1), glucose lev el (X2) and fermentation time (X3) were optimized using response surface method considering Page 73

28 degree of hydrolysis (DH; %) and ph as main response variables. The other response variables analysed were oil recovery, total titratable acidity (TTA) and antiox idant activities (DPPH radical scavenging activity and total antioxidant activity). The optimized conditions were found to be 15% (v/w) inoculum, 12 % (w/w) glucose and 48 h of fermentation at 37 ± 1 C to obtain a maximum DH%. As higher fermentation efficiency correlates with decrease in ph and increase in degree of hydrolysis, these were considered the best response variables for optimization. Oil recovery and antioxidant activities of the protein hydrolysate is due to hydrolysis of protein (increase in DH %) due to decrease in ph and protease produced by lactic acid bacteria (LAB). The FVW, used for the study had protein and fat of 8.43 %, and 24.12%, respectively, on wet weight basis and moisture content of 65 %. The fresh material had a ph 6.2 which is favorable for the proliferation of LAB. The influence of three independent factors viz., inoculum level (X1; % v/w), glucose level (X2; % w/w) and incubation time (X3; h) on the fermentation of cooked FVW was determined using full factorial design as mentioned previously. The observed values for the response variable (ph, TTA, oil recovery %, DH% and total antioxidant activity and DPPH radical scavenging activity) at different combinations of the independent variables are presented in Table The response surface graph for ph, as a function of sugar level and inoculum, sugar and time, time and inoculum is presented in Figure 3.6. ph decreased with increase in sugar concentration till 12% and inoculum of 15% after which there was no reduction in ph. Lower decrease in ph on increase inoculum and sugar can be due to alterations in substrate concentrations as well as inhibition of growth of LAB by excess production of end products of fermentation. With regard to time, ph reduced till 2 nd day of fermentation after which the decrease was not significant probably due to buffering of the medium by the peptides released on hydrolysis of collagen rich protein. The response surface graph for DH %, as a function of sugar level and time, sugar and inoculum, time and inoculum is presented in Figure 3.7, which increased with fermentation time of 48 h DH increases with the increase in glucose level till 12% (w/w) beyond which DH% reduces, which may be due to substrate inhibition due to excess of sugar. Oil recovery (% ) was found to be highest (87%) in optimized condition (inoculum - 15% v/w, sugar - 12% w/w and time 48 h). Page 74

29 Table Actual level of Independent variables and observed value of response variables during optimization experiment. Run # Independent Variables Response Variables X1 X2 X3 Y1 Y2 Y3 Y4 Y5 Y X1- Inoculum, X2 time, X3 sugar, Y1 ph, Y2 - total titrable acidity (TTA); Y3 oil recovery; Y4 DPPH radical scavenging activity ; Y5 degree of protein hydrolysis Page 75

30 Oil recovery is correlated with hydrolysis of protein as on degradation of protein in the lipid protein complex release the oil which floats on fermentation of cooked fish visceral waste. The fermented protein hydrolysate fraction had a reasonably good antioxidant activity which is attributed to higher degree of hydrolysis. Antioxidant property of a peptide depends on the intrinsic characteristics of peptides. The study points out the fact that such radical scavenging activity rich fraction can possibly be used in animal feed formulations as a supplement to relieve oxidative stress. Figure 3.6. Response surface graphs of ph as function of (A) sugar (%, w/w) and inoculum (%, v/w); (B) sugar (%, w/w) and Time (hours); (C) fermentat ion time (h) and inoculum (%, v/w) during fermentation of cooked homogenized FVW at 37 ± 2 C The regression equation for ph and DH% of the fermented of fish visceral waste, as a function of the three independent variables (X1, X2 and X3) and their linear a nd quadratic interactions is presented in the following equation: Page 76

31 ph = *x *x *y *y *x*y *x*y *x 2 *y *x 2 *y *12.*x *144.*x *12.*x *144.*x *12.*y *144.*y *12.*y *144.*y DH = *x *x *y *y *x*y *x*y *x 2 *y *x 2 *y *12.*x *144*x *12.*x *144.*x *12.*y *144.*y-.0265*12.*y *144.*y Figure 3.7. Response surface graphs of DH as function of (A) sugar (%, w/w) and Time (hours); (B) sugar (%, w/w) and inoculum (%, v/w); (C) fermentation time (h) and inoculum (%, v/w) during fermentation of cooked homogenized FVW at 37 ± 2 C. The optimized levels of variables (X1, X2 and X3) were determined u sing desirability profiles (Figure 3.8) for DH and TTA by assigning 0 to least desired level of Page 77

32 response and 1 to most desired response. The optimized factors for obtaining the highest DH% on fermentation using Pediococcus acidilactici NCIM5368 as starter culture were 15% Inoculum, 12% glucose and 96 h of fermentation at 37 ± 1 C. The predictability of the model was evaluated by comparing observed and predicted value of DH%. The model was further validated by employing random combinations (independent of runs used in experimental runs) of the independent variables (Table 3.16). The study reveals that fermentation of FVW with native LAB ( Pediococcus acidilactici NCIM5368) can be an effective method for simultaneous recovery of lipids and protein. Table Degree of protein hydrolysis (DH %) and ph observed during validation experiment and corresponding values predicted by the model. Run # Independent variables Response variables Inoculum Time Sugar ph obs ph pred DH obs DH-pred Obs observed; pred predicted; DH - degree of protein hydrolysis Page 78

33 Figure 3.8. Desirability profiles for degree of protein hydrolysis (DH) and ph at optimized levels of independent variables. Page 79

34 Large scale fermentation using Pediococcus acidilactici NCIM5368 of fish visceral waste for simultaneous recovery of lipids and protein Fish visceral waste used for large scale fermentation had a ph of 6.17 and protein and lipid content of 8.4 and 24.1% respectively on wet weight basis (Table 3.17). Large scale fermentation (20kg x 5) on optimized conditions of cooked homogenized fish visceral waste resulted in simultaneous recovery oil (22.5 %, w/w) and protein (6.21 %, w/w) on wet weight basis (Figure 3.9). The nitrogen recovery was >74% of the total nitrogen present in the original material (Table 3.17). Protein and fat content in the lyophilized fermented liquor powder were 36.27% and 1.1%, respectively. Oil recovery was > 93% of the total oil present in the fish visceral waste. The yield of lyophilized powder and its nitrogen content was 6.21 and 74.59% respectively. Results show that fermentation of FVW with native LAB (Pediococcus acidilactici NCIM5368) can be an effective approach for simultaneous recovery of lipids and protein. The lipids recovered from fish visceral waste through lactic acid fermentation need to be evaluated for its effect on biochemical parameters and bioefficacy in comparison to cod liver oil so that it can be used in aquaculture, life feed stock and pharmaceutical industries. Lipids Protein Figure 3.9. Fermentation of fresh water fish visceral waste using Pediococcus acidilactici NCIM5368 in optimized conditions Page 80

35 Table Yield (%) and content of lipid (%, wwb) and protein at different stages during lactic acid fermentation of fresh water fish visceral waste at large scale Moisture Protein (%, wwb) Fat (%, wwb) ph Yield of protein (%) Yield of Oil (%) A B A B FVW a a a a Cooked FVW a a a a FPH Extract b b b Lyophilized FPH c c b Recovered Oil FVW: Fresh fish visceral waste; FPH fermented protein hydrolysate; wwb wet weight basis; A: as percentage (w/w) of the wet fish visceral waste. B: nitrogen or oil recovery as percentage of nitrogen or lipid content, respectively, in fresh FVW. Values are mean SD Page 81

36 Enzymatic hydrolysis for simultaneous recovery of lipids and protein from fresh water fish visceral waste Four different commercial proteases (Fungal protease, Protex, Alcalase and Nutrase) were evaluated for recovering lipids and protein simultaneously by hydrolysis. On enzymatic hydrolysis of homogenized FVW using different commercial proteases, recovery of lipids ranged from 42-74% (Fig ure 3.10) considering the lipids extracted using solvents as 100%, depending on the type of enzyme. FVW hydrolyzed with fungal protease resulted in highest recovery of lipids (74.9%) as well as higher DH (49.1%) when compared to recovery of lipids and DH in case of other enzymes. The results indicate the utility of commercial proteases in providing an ecofriendly and feasible solution for reducing problems associated with disposal of fish processing waste. Enzymatic hydrolysis with added proteases to fish visceral waste increased the oil recovery and degree of protein hydrolysis by and %, respectively, compared to control (Figure 3.11) a DH % Oil Recovery % b b b % a a b b c 30 c 15 0 F protease Protex Alcalase Neutrase Control Figure Percent oil recovery and degree of protein hydrolysis on enzymatic hydrolysis of fish visceral waste using different commercial proteases. Values are mean ± SD (n=5); Values not sharing common alphabets within same pattern are significantly different (p<0.05). Page 82

37 % DH % Oil Recovery % F protease Protex Alcalase Neutrase Proteases Figure Percentage increase in oil recovery and degree of protein hydrolysis (DH %) on enzymatic hydrolysis of fish visceral waste as compared to control Extractability of the fish waste protein hydrolysate ( FWPH) also followed the similar pattern of degree of protein hydrolysate wherein FWPH of fungal protease (55.41%) was found to be highest. Extractability was in the range of % (Figure 3.12), least being in control (no added enzyme). The DH, extractability and oil recovery observed in case of control could possibly be due to the heat treatment step; and, partly due to the activity of endogenous enzymes before or after and heat treatment, as fish are known to contain some heat activated proteases. Among different commercial proteases used, Protex 7L showed highest recovery (49.5%), least being in alcalase (35.9%). However, maximum collagen recovery was found in case of control (76.5%). This again clearly shows there was least hydrolysis in case of control and also that part of the collagen could be hydrolyzed by the proteases used. Page 83

38 90 75 Extractibility Collagen content c % a a b b c a c a 30 d 15 0 F Protease Protex Alcalase Neutrase Control Figure Percent extractability (%) and collagen content (%) on enzymatic hydrolysis of fish visceral waste using different commercial proteases. F Protease Fungal protease; Values are mean ± SD (n=5); Values not sharing common alphabets within same pattern are significantly different (p<0.05). Control no added enzyme. The in-vitro antioxidant properties - total antioxidant activity (TAO) and DP PH radical scavenging activity of hydrolysates obtained using different commercial proteases are presented in Table Control had lower TAO as well as DPPH scavenging activity when compared to hydrolysate samples. TAO (expressed as of mg of ascorbic acid equivalents per micrograms of protein in hydrolysate; AAE/mg) was found to be highest in hydrolysate obtained using neutrase (34.4) and least was in case of hydrolysate obtained with alcalase (22.1). The TAO exhibited by the hydrolysate could possibly be due to the peptides and amino acids resulting from the hydrolysis of FVW. DPPH radical scavenging activity (%) in the protein hydrolysate samples was used for determining EC 50, the concentration of hydrolysate protein (mg) that scavenges 50% of DPPH radicals. EC 50 in case of protein hydrolysate prepared using fungal protease was lowest (3.47 mg protein) and highest in case of neutrase (3.99 mg protein) indicating the effectiveness of fungal protease to scavenge free radicals. Page 84

39 Table Total antioxidant activity (μg of ascorbic acid equivalents/mg of protein) and EC 50 (mg protein) values of DPPH and superoxide radical scavenging activity of fish waste protein hydrolysate on enzymatic hydrolysis using different commercial proteases Antioxidant activities Protease used DPPH SRSA TAO Protex 7L 3.77±0.2 a,b 2.88±0.2 a,c ±1.8 a Nutrase 3.99±0.2 a 2.35±0.3 b ±2.9 a Alcalase 3.85±0.1 a 2.68±0.2 c ±2.7 b Fungal protease 3.47±0.2 b 2.26±0.3 b ±1.6 a Control 4.01±0.2 a 3.20±0.2 a ±2.2 b DPPH DPPH radical scavenging activity; SRSA superoxide radical scavenging activity; TAO total antioxidant activity; Control no added enzyme; Values are mean ± SD (n=5); Values not sharing common alphabets within same patt ern are significantly different. The quality of oil recovered on enzymatic hydrolysis with respect to acid value and peroxide value (Table 3.19). In case of acid value, oil recovered using neutrase showed highest acid value (mg KOH g -1 oil) of 22. No significant changes were observed in acid value of oil recovered by different protease. Peroxide value (meq O 2 /kg of oil) in the oil recovered after hydrolysis was in case of control (no added enzyme) compared to 8.7 in oil extracted by solvent from fresh fish viscera. Peroxide value in oil recovered on hydrolysis of FVW by Fungal protease (32.5) and alcalase (28.9) were significantly lower compared to other proteases. With regards to fatty acid composition of oil recovered by hydrolysis using different enzymes, no change was observed irrespective of the type of enzyme as compared to the oil recovered by solvent extraction (Table 3. 20). Further, there was no significant change in the fatty acid profile in the oil recovered by enzymatic hydrolysis using different proteases to that of fresh oil. Thus, enzymatic hydrolysis can be an effective approach for the recovery of oil without affecting the quality of oil Page 85

40 Table Acid value and peroxide value in oil recovered by enzymatic hydrolysisfrom fresh water fish viscera. Enzyme used Acid value (mg KOH/gm of oil) Peroxide value (meq O 2 /kg of oil) Fresh ± 1.1 a 8.7 ± 0.9 a Protex 7L ± 1.2 b ± 2.8 b F protease ± 0.8 b ± 3.1 c Neutrase ± 1.1 b ± 1.2 b Control ± 1.0 b ± 2.4 b Alcalase ± 0.5 b ± 2.5 c Values are mean ± SD (n=5); Values not sharing common alphabets within same pattern are significantly different; Fresh extracted by solvent; Control no enzyme added. Table Fatty acid composition (%) of recovered lipids on enzymatic hydrolysis of cooked fresh water fish viscera. Fatty acids Fresh Alcalase Protex 7L Fungal Protease Nutrase Control 14:0 3.3±0.3 a 2.9 ± 0.3 a 3.1 ± 0.4 a 2.8 ± 0.3 a 2.9 ± 0.3 a 2.8 ± 0.3 a 15:0 1.2±0.2 a 1.0 ± 0.1 a 1.2 ± 0.2 a 1.1 ± 0.2 a 1.0± 0.2 a 1.3±0.2 a 16:0 30.8±3.1 a 29.5 ± ± 3.2 a 28.7± ±2.9 a 28.3±3.5 a 16:1 7.5±1.2 a 7.9 ± 1.3 a 7.8 ± 0.9 a 7.4±1.4 a 7.8 ± 1.1 a 7.3 ± 0.9 a 18:0 8.8± ± 0.8 a 8.0 ± 1.1 a 7.9±1.2 a 7.3 ± 0.8 a 7.7 ± 0.8 a 18:1n ±3.0 a 23.2±2.5 a 22.9±2.3 a 23.1±2.6 a 23.0±2.2 a 22.8±2.4 a 18:1n-7 3.3±0.4 a 3.4 ± 0.3 a 3.3 ± 0.4 a 3.4±0.5 a 3.3 ± 0.4 a 3.2 ± 0.3 a 18:2n-6 9.7±1.4 a 10.6± 1.2 a 10.6± 1.6 a 9.9±1.2 a 10.8± 0.9 a 10.6±1.3 a 18:3n-3 2.7± ±0.3 a 2.9±0.3 a 2.6±0.4 a 3.0 ± 0.3 a 2.9 ± 0.3 a 20:4n-6 0.8±0.2 a 1.0±0.2 a 0.9±0.2 a 0.8±0.2 a 0.9 ± 0.3 a 1.0 ± 0.3 a 20:5n ± 0.2 a 2.4 ± 0.3 a 2.5 ± 0.2 a 2.6±0.4 a 2.5 ± 0.3 a 2.6 ±0.3 a 22:5n ± 0.1 a 0.6 ± 0.2 a 0.7 ± 0.1 a 0.8±0.2 a 0.6 ± 0.2 a 0.9 ± 0.3 a 22:6n-3 2.8±0.3 a 2.9±0.3 a 2.8 ± 0.4 a 2.9±0.3 a 2.9 ± 0.4 a 2.8 ± 0.3 a SFA saturated fatty acid, USFA unsaturated fatty acid; Values are mean ± SD (n=5); Values not sharing common alphabets within same row are significantly different (p<0.05), Fresh extracted by solvent; Control no enzyme added. Page 86

41 Optimization of enzymatic hydrolysis conditions for the simultaneous recovery of lipids and protein from fish processing waste Experiments on enzymatic hydrolysis of fish visceral waste using different proteases indicated that Fungal Protease was found to be the better protease for the simultaneous recovery of lipids and protein. Hence, conditions were optimized by response surface methodology for obtaining higher oil recovery and degree of protein hydrolysis. Three independent variables viz., enzyme concentration (X1; %w/w), Time (X2; min) and enzyme: substrate ratio (X3) were optimized using response surface method considering oil recovery (%) and degree of hydrolysis (DH; %) as main response variables. The other response variables analyzed were antioxidant activities (DPPH radical scavenging activity and total antioxidant activity). The optimized conditions were found to be 0.5% (w/w) inoculum, 1:3 wastes: water concentration and 2 h of hydrolysis at 40 ± 1 C to obtain a maximum oil recovery. As higher hydrolysis correlates with increase in degree of hydrolysis and oil recovery so these was the best response variables considered for optimization. Antioxidant activities of the protein hydrolysate are due to hydrolysis of protein (increase in DH %) by the action of Fungal proteases. The fish visceral waste (FVW), used for th e study had protein and fat of 8.71%, and 23.14%, respectively, on wet weight basis and moisture content of %. The fresh material had a ph 6.2 which is favorable for the proliferation of LAB. The influence of three independent factors viz., enzyme concentration (X1; %w/w), Time (X2; min) and enzyme: substrate ratio (X3) on the enzymatic hydrolysis of cooked FVW was determined using full factorial design as mentioned previously. The observed values for the response variable (oil recovery (%), DH (%), total antioxidant activity and DPPH radical scavenging activity) at different combinations of the independent variables are presented in Table The response surface graph for oil recovery, as a function of substrate: water ratio and enzyme concentration (% w/w), substrate: water ratio and time (min), time (min) and enzyme concentration (%, w/w) is presented in Figure The response surface graph for degree of hydrolysis (DH %), as a function of time (min) and enzyme concentration (%, w/w), substrate : water ratio and enzyme concentration (% w/w) is presented in Figure 3.14 which increased with Page 87

42 hydrolysis time of 120 mins beyond which it reduces. DH decreases as the enzyme concentration increases after 0.5% (w/w). Table Actual level of Independent variables and observed value of response variables during optimization experiment. X1 X2 X3 Y1 Y2 Y3 Y X1- Enzyme concentration (w/w, %), X2- Time (min), X3 Substrate: water ratio; Y1 oil recovery; Y2 Degree of hydrolysis; Y3 DPPH radical scavenging activity ; Y4 Total antioxidant activity (TAO) Oil recovery (%) also followed the same trend as degree of hydrolysis which was found higher at 0.5% (w/w). The oil recovery and degree of protein hydrolysis was found Page 88

43 to be highest in optimized condition (enzyme concentration 0.5% w/w, substrate: water ratio 1:3 and time 120 min). Oil recovery is correlated with hydrolysis of protein as on degradation of protein in the lipid protein complex release the oil which floats on fermentation of cooked fish visceral waste. The protein hydrolysate fraction had reasonably good antioxidant activities (DPPH radical scavenging activity and total antioxidant activity) which is attributed to higher degree of hydrolysis. Antioxidant property of a peptide depends on the intrinsic characteristics of peptides. The study points to the fact that such radical scavenging activity rich fraction can possibly be used in animal feed formulations as a supplement to relieve oxidative stress. Figure Response surface graphs of oil recovery as function of (A) substrate : water ratio and Enzyme (% w/w); (B) substrate : water ratio and Time (min); (C) Ti me (min) and enzyme (%, w/w) during hydrolysis of cooked homogenized FVW at 40 ± 2 C. Page 89

44 Figure Response surface graphs of Degree of hydrolysis (DH) as function of (A) Time (min) and enzyme (%, w/w); (B) substrate : water ratio and enzyme (% w/w) (C) DH and enzyme (%w/w) during hydrolysis of cooked homogenized FVW at 40 ± 2 C The optimized levels of variables (X1, X2 and X3) were determined using desirability profiles (Figure 3.15) for Degree of hydrolysis (%) and oil recovery (%) by assigning 0 to least desired level of response and 1 to most desired response. The optimized factors for obtaining the highest DH% and Oil recovery on enzymatic hydrolysis using fungal protease as proteolytic enzyme were 0.5% (w/w) enzyme concentration, 1:3 - substrate: water ratio and 2 hours of hydrolysis time at 40 ± 1 C. The regression equation for Oil recovery and degree of protein hydrolysis on enzymatic Page 90

45 hydrolysis of fish visceral waste, as a function of the three independent variables (X1, X2 and X3) and their linear and quadratic interactions is presented in the following equation: Degree of hydrolysis = *x 68.75*x *y *y *x*y *x*y *x 2 *y *x 2 *y *2*x *4*x *2.0*x *4.*x *2.*y *4*y *2*y *4*y Oil recovery = *x *x *y *y *x*y *x*y *x 2 *y *x 2 *y *2*x *4*x *2*x *4*x *2*y *4*y *2*y *4*y Page 91

46 Figure Desirability profiles for degree of protein hydrolysis and oil recovery at optimized levels of independent variables Page 92