Food Chemistry 257 (2018) 302–309

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Use of fish trypsin immobilized onto magnetic-chitosan composite as a new T

tool to detect antinutrients in aquafeeds
Rafael D.S. Azevedoa, Ian P.G. Amaralb, Amália C.M. Ferreiraa, Talita S. Espósitoc,
⁎
Ranilson S. Bezerraa,
a
Laboratório de Enzimologia – LABENZ, Departamento de Bioquímica, Universidade Federal de Pernambuco, Av. Prof. Moraes Rego, Campus Universitário, Recife 50670-
420, PE, Brazil
b
Centro de Biotecnologia, Universidade Federal da Paraiba, Campus I, Cidade Universitária, João Pessoa, PB, Brazil
c
Laboratório de Biotecnologia de Organismos Aquáticos, Universidade Federal do Maranhão, Departamento de Oceanografia e Limnologia, Av. dos Portugueses, 1966,
Bacanga, São Luis, MA, Brazil

A R T I C LE I N FO A B S T R A C T

Keywords: The unplanned inclusion of antinutrients in fish food affects many biological processes, such as digestibility of
Protease inhibitors amino acids and diet conversion, resulting in undesirable effects on body growth. Thus, the objective of this
Fish processing residues research was to propose the use of immobilized fish proteases in the detection of protease inhibitors, one of the
Immobilization and enzyme reuse most important antinutrients. In order to evaluate the detection of antinutritional factors through the im-
Aquaculture
mobilized trypsin, the enzyme was incubated with eight diets developed for commercial fish, and residual ac-
Chemical compounds utilized in this article: tivity was measured. Comparatively, the tilapia trypsin showed an inhibition of antinutrients (protease in-
Ammonium sulfate (PubChem CID: 6097028) hibitors), present in the eight studied diets, up to 48% greater than the porcine trypsin immobilized in magnetic
Benzamidine (PubChem CID: 80289)
chitosan. Thus, it is possible to suggest the use of immobilized derivatives containing specific proteases of the
Tris 2-Amino-2-hydroxymethyl-propane-1,3-
target organism in the detection of antinutritional factors that reduce animal’s digestive capacity and negatively
diol (PubChem CID: 1531)
Hydrochloric acid (PubChem CID: 313) influence their growth during husbandry.
Sodium chloride (PubChem CID: 5234)
Benzoyl-arginine-p-nitroanilide (PubChem CID:
2724371)
Dimethyl sulfoxide (PubChem CID: 679)
Acrylamide (PubChem CID: 6579)
Sodium hydroxide (PubChem CID: 14798)
Ferrous chloride (PubChem CID: 24458)
Ferric chloride (PubChem CID: 24380)
Glutaraldehyde (PubChem CID: 3485)
Glycine (PubChem CID: 750)
Citric acid (PubChem CID: 311)
Monosodium phosphate (PubChem CID:
23672064)
Sodium phosphate dibasic (PubChem CID:
24203)
Potassium permanganate (PubChem CID:
516875)
Sodium bisulfite (PubChem CID: 23665763)
Acetic acid (PubChem CID: 176)

1. Introduction the diet. Detection and removal of antinutritional factors is an im-

portant aspect in the formulation of fish feed. Among the main sub-
The formulation of feed that promotes accelerated growth of the stitutes of animal protein in the formulation of feeds, proteins from soy,
target organism is one of the strategies used for rapid aquaculture de- pea, or canola are examples of ingredients used in the formulation of
velopment, or for minimizing the costs of including animal protein in feed for aquatic organisms. Current discussions are concerned with

⁎
Corresponding author.
E-mail address: ransoube@uol.com.br (R.S. Bezerra).

https://doi.org/10.1016/j.foodchem.2018.03.034
Received 16 August 2017; Received in revised form 8 March 2018; Accepted 8 March 2018
Available online 10 March 2018
0308-8146/ © 2018 Elsevier Ltd. All rights reserved.
R.D.S. Azevedo et al.
bioaccumulation and the increasing presence of these antinutrients
(vegetable by-products), as ingredients in fish diets. Additionally, they
warn about histological alterations and the induction of antigens
through the feed’s protein source, or even about damage to growth
performance due impair of dietary use and conversion (Couto et al.,
2014). Indeed, negative physiological and morphological effects have
already been associated with the intake of antinutritional factors for
fish species, such as the Atlantic salmon or Rainbow trout (Chikwati
et al., 2012; Knudsen et al., 2008; Kortner et al., 2012; Yamamoto et al.,
2012).
Antinutrients can reduce functional or nutritional properties of feed
for aquatic organisms, and a variety of compounds derived from ve-
getable by-products can be classified as antinutrients, for instance:
phytosterols, phytic acid, saponins, tannins and protease inhibitors of
protein origin. Protease inhibitors are considered the most important
antinutritional factor, because they affect protein digestion and amino
acid assimilation (Bajpai, Sharma, & Gupta, 2005). In this context,
trypsin inhibitors have been indicated as responsible for pancreatic
hypertrophy and digestive enzyme hypersecretion (Guillamon et al.,
2008). These characteristics provide reflections about the interaction
between fish digestive proteases and their possible inhibitors present in
the feed, with the objective of improving protein digestion. Among the
digestive proteinases, trypsin plays a fundamental role in the digestion
of other zymogens and, thus, is referred to as a key enzyme in protein
digestion. As such, trypsin is a good target for detecting its inhibitors
present in fish feed.
Magnetic composites are well known for facilitating many biolo-
gical, biochemical and biotechnological processes. In this way, sophis-
ticated processes have demonstrated the ubiquity of magnetic sensors,
such as redox protein immobilization (Bagheri, Asgharinezhad, &
Ebrahimzadeh, 2016), extraction and preconcentration of ions on food
samples (Bagheri et al., 2016), enzyme purification (Alves et al., 2017)
and even the delicate sensitive determination of potassium channel
blockers (Hashemi, Bagheri, Afkhami, Amidi, & Madrakian, 2018).
Chitosan magnetization brings to very useful polymers the ease of
maximizing their use in a myriad of biotechnological reactions, parti-
cularly those involved in food/feed hydrolysis through immobilized
enzymes.
Enzymatic immobilization shows a series of advantages when
compared to soluble enzymes, such as enzyme reuse, high efficiency in
the removal of reaction products, and high stability in the face of
temperature and pH variations. In this context, natural polymers such
as chitosan can be used for immobilization due to characteristics such
as elevated biocompatibility, low molecular weight and high adsorption
capacity, in addition to other sophisticated functions associated to these
properties (Jayakumar, Prabaharan, Kumar, Nair, & Tamura, 2011;
Kumirska, Weinhold, Thoming, & Stepnowski, 2011). Many studies on
cellular and enzymatic immobilization in chitosan have emerged since
the 1980s, describing simple methodologies for protein and enzyme
immobilization (e.g. pectinases, chitinases and proteases) (Liu, Li, Li,
He, & Zhao, 2010; Seo, Jang, Park, & Jung, 2012; Singh, Singh, Suthar,
& Dubey, 2011), which showed good results for biotechnological ap-
plication. Its use in antinutrient detection enables the elucidation of
current issues, such as the difficulty for simple separation of specific
antinutrients in feed used in aquaculture, and provides advances in the
comprehension of molecular interactions between digestive enzymes
and feed for aquatic organisms. In this study, intestinal proteases from
Nile tilapia were immobilized in magnetic chitosan support for anti-
nutrient detection in fish feed.
2. Materials and methods
2.1. Raw materials
Processing residue of the seabob shrimp, Xiphopenaeus kroyeri
(heads, tails and shells), were collected in local fishing community from
303
Food Chemistry 257 (2018) 302–309
the State of Alagoas (Brazil). Samples were sorted manually to remove
contaminants (leaves, small fish and debris in general). Processing re-
sidues of the fish Oreochromis niloticus (carcass and entrails) were ob-
tained from Noronha Pescados, a fishing company from the state of
Pernambuco (Brazil). Residues from fish and shrimp were transported
on ice to LABENZ – Federal University of Pernambuco. The reagents
used were acquired from Merck (Darmstad, Germany) and Sigma (St.
Louis, MO, USA). The column for affinity chromatography was acquired
from GE Healtcare™. A schematic diagram attached on supplementary
data C illustrates all procedures.
2.2. Enzyme extraction and partial purification
Fish intestines (1g) were homogenized in refrigeration with 25 mL
of phosphate buffer 0.05 M pH 7.8 using a thread homogenizer. The
resulting product was centrifuged at 10,000ɡ for 10 min at 4 °C to re-
move cellular and nuclear residues. The supernatant (crude extract)
was used to partially purify the trypsin through heating and saline
fractioning with ammonium sulfate, according to the method of Bezerra
et al. (2005). The precipitates from fractions 0–30% (F1) and 30–70%
(F2) were resuspended with 10 ML of phosphate buffer at 0.05 M pH
7.8.
Fraction F2 showed greater specific proteolytic activity and was
applied in a benzamidine column HiTrap Benzamidine FF (GE
Healtcare™) that was previously balanced with buffer tris-HCl 0.01 M
pH 7.8 with a flow of 1 mL per minute and the collection of 1 mL
fractions that were monitored in a spectrophotometer at 280 nm, in
order to monitor protein presence. After the application of fraction F2,
the column was washed with 10 mL of buffer tris-HCl 0.01 M pH
7.8 + 1 M NaCl for removal of weakly-bound proteins, until absorbance
at 280 nm was close to zero. Elution of trypsin from the resin was
conducted with 0.01 M HCl + 0.5 M NaCl in accordance to factory in-
structions. After chromatography, fractions that showed proteolytic
activity against BApNA (benzoyl-arginine-p-nitroanilide) were put to-
gether in a “pool”.
2.3. Trypsin activity
The catalytic activity of the enzyme was determined using BApNA at
8 mM dissolved in dimethyl sulfoxide (DMSO). All purification stages
(F1, F2) were diluted (1:10) with 0.05 M phosphate buffer pH 8.0 and
the activity was evaluated in triplicates. The resulting product (p-
Nitroaniline) was quantified at 405 nm in a microplate reader (Bio-Rad
X-Mark spectrophotometer, California, USA) after 15 min of reaction at
25 °C. An enzyme unit was defined as the amount of enzyme needed to
hydrolyze 1 µmol of BApNA per minute.
2.4. Electrophoresis (SDS-PAGE)
Samples from each purification step were applied on Sodium do-
decyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using a
4% concentration gel and 12% separation gel. Electrophoresis was
conducted in a vertical electrophoresis system (Bio-Rad Laboratories,
Inc) at 11 mA. Molecular weight standards containing bovine serum
albumin (66 kDa), ovalbumin (45 kDa), glyceraldehyde 3-phosphate
dehydrogenase from rabbit muscle (36 kDa), bovine carbonic anhy-
drase (29 kDa), bovine pancreatic trypsinogen (24 kDa), soybean
trypsin inhibitor (20.1 kDa), alpha-lactalbumin and bovine milk
(14.2 kDa) were used to estimate the molecular weight of the samples.
Gel coloring was done with Silver-BULLit™ (Amresco®) following fac-
tory instructions.
2.5. Recovery of chitosan
Processing residues from Xiphopenaeus kroyeri (0.5 kg) were homo-
genized with distilled water (w/v) in a food processor (LB – 15PMB
R.D.S. Azevedo et al. Food Chemistry 257 (2018) 302–309

Table 1
Formulation and proximate analysis of experimental diets studied (% dry matter).

INGREDIENTS DIET 1 DIET 2 DIET 3 DIET 4 DIET 5 DIET 6 DIET 7 DIET 8

Fish meal 66% 31.50 27.00 21.00 13.00 20.28 23.92 27.05 30.17
Corn meal 34.00 34.00 34.00 34.00 50.00 40.00 30.00 20.00
Poultry by-product 67% 15.00 11.00 9.00 7.00 5.70 10.00 15.00 20.00
Soybean meal 34% 17.00 21.00 21.00 25.00 20.80 23.02 25.10 27.12
Vitamin and Minerals 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00
NaCl 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25
Antioxidant 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25
Vitamin C – 5.00 10.00 15.00 0.00 0.00 0.00 0.00
Methionine – – 4.00 – 0.67 0.50 0.35 0.19

Proximate composition (%)

Moisture 1.50 2.19 2.10 1.23 2.34 3.09 1.46 8.19
Crude protein 45.44 42.15 39.39 34.65 28.20 36.44 40.71 45.25
Ether extract 5.02 9.87 11.07 19.19 5.82 4.94 5.98 7.70
ENN** 25.10 24.45 30.00 21.00 48.36 37.42 32.94 29.87
Ash 11.74 10.49 9.53 8.55 9.70 10.61 12.34 12.50
Crude Fiber 7.89 12.01 7.39 13.72 5.58 7.46 6.57 7.74
Gross Energy (kJ.g−1) 20.10 19.94 20.48 20.43 18.57 17.25 19.16 17.83

*
Rovimix fish: vit. A: 5.000.000 UI; vit. D3: 200.000 UI; vit. E: 5.000 UI; vit. K3: 1.000 mg; vit. B1: 1.500 mg; vit. B2: 1.500 mg; vit. B6: 1500 mg; vit. B12: 4.000 mg; vit. C: 15.000 mg;
folic acid: 500 mg; pantothenic acid: 4.000 mg; B.H.T.: 12.25 g; biotin: 50 mg; inositol: 1.000 mg; nicotinamide: 7.000 mg; choline: 40 g; cobalt: 10 mg; copper: 500 mg; iron: 5.000 mg;
iodine: 50 mg; manganese: 1.500 mg; selenium: 10 mg; zinc: 5.000 mg; q.s.q.: 1.000 g.
**
ENN% = 100−(crude protein% + crude lipid% + moisture% + ash% + fiber%).

Siemsem Ltda.) at room temperature, followed by enzymatic autolysis
at 40 °C in a water bath for 2 h. The enzymes present in the solution
were deactivated at 100 °C. The liquid phase was separated from the
solid part with the help of a sifter, and solid material was dried in an
aerated stove at 40 °C. This material was submitted to the stages of
decalcification, deproteinization, pigment removal and deacetylation,
as described by (Cahu et al., 2012). The deacetylation process was
conducted through alkaline treatment with 500 mL of NaOH 50% (w/
v−1) at 65 °C for 24 h. The chitosan pH was neutralized, washed with
distilled water and dried in an aerated stove at 50 °C. In an attempt to
achieve a chitosan with a high degree of deacetylation, the purification
methodology adapted from (Weska, Moura, Batista, Rizzi, & Pinto,
2007) was applied. Recovered chitosan samples were lyophilized and
used for magnetization.
2.6. Chitosan magnetization
Magnetization was carried out through sample co-precipitation in
the presence of FeCl2 and FeCl3 at 0.6 and 1.1 M, respectively (Amaral,
Carneiro-da-Cunha, Carvalho, & Bezerra, 2006). Chitosan (1 g) was
solubilized in 3% acetic acid under smooth agitation. FeCl2 and FeCl3
solutions were added (1:1) and pH was adjusted to 11.0 with NaOH 1 M
(Amaral et al., 2006). The mixture was put in a water bath for 30 min at
80 °C under constant agitation. Magnetic chitosan particles were col-
lected with the help of a magnet and were then washed with distilled
water until pH neutralization. Samples were centrifuged for 20 min at
8.000ɡ and were lyophilized.
2.7. Fish and porcine protease immobilization
Magnetic chitosan (10 mg) was activated for immobilization with
glutaraldehyde at 12,5% diluted in phosphate buffer at 0.05 M pH 8.0
for two hours under smooth agitation at 25 °C. Magnetic chitosan was
recovered with a magnet and then washed 10 times to remove excess
glutaraldehyde. To evaluate the best incubation time for tilapia en-
zymes, magnetic chitosan was incubated with 1 mg/mL of porcine
trypsin (Sigma-Aldrich Ltda.) for 1, 2, 3, 4, 6, 8, 10, 12, 14 and 16 h.
Tilapia proteases were incubated with magnetic chitosan for 8 h
under smooth agitation at 4 °C. The same procedure was used to im-
mobilize porcine trypsin starting with 1 mg solution of porcine trypsin/
mL of buffer. The supernatant of porcine trypsin and purified fish
proteases were collected, and the support was washed 10 times with
1 mL of phosphate buffer 0.05 M pH 8.0. To evaluate the quantity of
immobilized enzymes, protein content in the supernatant and washings
were estimated through absorbance at 260 nm and 280 nm in a spec-
trophotometer. Remaining active groupings of glutaraldehyde were
blocked with glycine at 0.1 M for 16 h at 4 °C. To determine im-
mobilized enzyme activity, 150 µL of BApNA at 8 mM dissolved in
DMSO and 850 µL 0.05 M phosphate buffer pH 8.0 were added to the
tube containing the immobilized enzyme and, after 30 min under mild
agitation at 25 °C, 200 µL were transferred to a 96-well plate and read at
405 nm.
2.8. Effects of pH, temperature and storage time of immobilized enzymes
The effects of pH on the residual activity of immobilized enzymes
were analyzed on a scale from 4.5 to 9.0 with different buffers (citrate-
phosphate, pH 4.5–5.5; phosphate, pH 6.0–8.0; tris-HCl, pH 8.5; and
NaOH-Glycine, pH 9.0). Thermal stability of immobilized enzymes was
evaluated at temperatures between 25 °C and 70 °C for 15 min and re-
peated twice as described in Amaral, Carneiro-da-Cunha, Carvalho, and
Bezerra (2006). The viability of immobilized enzymes stored at −20 °C
was evaluated at 0, 30, 60 and 90 days after the immobilization process.
Residual activity during storage was quantified using BApNA 8 mM
diluted in DMSO and phosphate buffer 0.05 M pH 8.0, as previously
described.
2.9. Antinutrients detection in commercial fish feed
Eight different types of fish diets were used (Table 1). The diets
tested were selected based on two features: (1) The search for alter-
native ingredients that replace animal protein is growing, especially
through vegetable protein inclusion. (2) Crops-processing waste as a
meal source may be a viable alternative to replace this animal protein
(tested in the present work). Each feed was homogenized with Tris-HCl
pH 8.0 buffer (1 g of feed for 10 mL of buffer) at 25 °C and then cen-
trifuged at 8.000ɡ for 5 min. Supernatants were collected for detection
of antinutritional factors and 1 mL of this solution was incubated with
magnetic chitosan support with the immobilized enzyme for 30 min.
After that, the supernatant was removed and the residual activity was
quantified as previously described (Supplementary data B). The dif-
ference of immobilized enzyme activity was calculated by the differ-
ence between the substrate control absorbance (405 nm) and the ab-
sorbance (405 nm) of the test activity divided by the absorbance

304
R.D.S. Azevedo et al. Food Chemistry 257 (2018) 302–309

Table 2
Residual catalytic activity for immobilized proteases on chitosan magnetic support, and for free proteases after incubation with the eight studied diets. BApNA 8 mM was used as a
substrate and the experiment was conducted at 25 °C. Results are shown in average% ± standard deviation compared to a control assay without diet (n = 4).

Diet Residual activity of immobilized enzymes (% ± SD) Residual activity of free enzymes (% ± SD)

Porcine trypsin Tilapia trypsin p value Porcine trypsin Tilapia trypsin p value

Diet 1 90.9 ± 3.3 63.3 ± 1.0 0.003 59.4 ± 5.0 33.6 ± 26.6 0.002
Diet 2 90 ± 1.0 77.2 ± 1.0 0.000 53.4 ± 6.7 48.0 ± 8.0 0.109
Diet 3 89.3 ± 2.6 82.5 ± 2.1 0.000 43.7 ± 2.9 28.3 ± 12.1 0.017
Diet 4 87.8 ± 3.4 58.9 ± 0.4 0.004 50.1 ± 9.4 38.8 ± 5.4 0.087
Diet 5 99.7 ± 1.8 71.9 ± 1.6 0.000 56.6 ± 11.8 34.6 ± 10.9 0.054
Diet 6 74.2 ± 2.0 61.1 ± 0.2 0.007 44.4 ± 7.1 41.9 ± 5.0 0.197
Diet 7 79 ± 3.3 89.7 ± 3.4 0.000 77.1 ± 25.9 57.0 ± 26.7 0.164
Diet 8 99.2 ± 2.7 5.3 ± 0.1 0.000 66.8 ± 4.9 36.3 ± 26.3 0.000

(405 nm) of the control * 100. The best catalytic activity was observed in fraction F2 (30–70%
saline saturation) of protease purification. A result that goes along with
previous studies (Khangembam and Chakrabarti, 2015; Silva et al.,
2.10. Data and statistical analysis
2011), where fish proteases are precipitated, for the most part, in
concentrations above 30% ammonium sulfate. The greatest catalytic
All collected data was tested for normal distribution using the
activity for F2 was 0.365 ± 0.0052 µg/mL, compared with values
Kolmogorov-Smirnov test. All results showed normal distribution and
below 0.159 ± 0.0046 µg/mL observed for other stages of the pur-
the difference between averages was analyzed using the t-test (two
ification process by ammonium sulfate, a result that indicated greater
groups) or ANOVA followed by the Tuckey post hoc test (more than two
protease concentration in this fraction. The affinity chromatography
groups). Results were considered significant when p < 0.05.
process and purification performance showed a satisfactory purification
degree (Supplementary data A). This result supports its application in
3. Results and discussion the food industry because the majority of food and detergent industries
do not require proteases with elevated purification degrees (Freitas
3.1. Enzyme extraction and partial purification et al., 2012). The satisfactory enzyme activity confirms the already
known discussions about the viability of similar methodologies for re-
All of the activities in the purification and recovery stages of fish covery and application of fish enzymes, especially when application is
residue processing proteases are shown in Table 2. Similar results were idealized for the feed industry for aquatic organisms, which was the
obtained by other authors, especially on the purification yield (Freitas objective of this research.
et al., 2012; Esposito, Marcuschi, Amaral, Carvalho, & Bezerra, 2010). All fractions from purification stages were separated by electro-
Fig. 1 shows chromatogram with benzamidine agarose carried out in phoresis, as shown in Fig. 2, where one can observe the success of
about an hour. The speed of this purification process improves one of partial protease purification by identifying a greater concentration of
the points discussed by Esposito et al. (2010) about how much high serine proteases (larger band) in fraction F2. Small differences are ob-
time consumption when purifying proteins can render application of served when this data is compared with data from partially-purified
these processes in semi-industrial and industrial scales unfeasible. As
such, this speed in the process and the low cost of fish residue pro-
cessing reinforce the feasibility for application of fish enzymes in in-
dustrial and biotechnological processes.

10
3,0

2,5 8
Absorbance 405nm
Absorbance 280 nm

2,0
6
Buffer + 0.5M NaCl

Buffer + 1M NaCl

1,5

4
1,0
Elution

0,5 2

0,0
0
-5 0 5 10 15 20 25 30 35 40 45 50
Sample (1 mL)
Fig. 2. SDS-Page electrophoresis profile: (lane 1) Molecular weight standards containing
Fig. 1. Affinity chromatogram with HiTrap Benzamidine FF (GE Healtcare™) column. bovine serum albumin (66 kDa), ovalbumin (45 kDa), glyceraldehyde 3-phosphate de-
This experiment was performed following manufacturer's instructions (5 mL of crude hydrogenase from rabbit muscle (36 kDa), bovine carbonic anhydrase (29 kDa), bovine
extract at 636 mg/mL were loaded into the column). Elution was conducted with 0.01 M pancreatic trypsinogen (24 kDa), soybean trypsin inhibitor (20.1 kDa), alpha-lactalbumin
HCl + 0.5 M NaCl. Absorbance at 280 nm was used to estimate protein content. and bovine milk (14.2 kDa). Purification steps: (lane 2a) crude extract of fish trypsin and
Absorbance at 405 nm is a result of specific activity of trypsin-like with 8 mM BApNA. (lane 2b) pool from affinity chromatography with HiTrap Benzamidine FF (GE
Experiment performed in quadruplicate (n = 4). Healtcare™).

305
R.D.S. Azevedo et al. Food Chemistry 257 (2018) 302–309

trypsins of other aquatic organisms, such as Lutjanus synagris, Thunnus

tongool, Araipama gigas, and Theragra chalcogramma; (Esposito,
Marcuschi et al., 2010; Freitas et al., 2012; Kishimura, Klomklao,
Benjakul, & Chun, 2008), which show molecular weights of 28.4 kDa,
24.0 kDa, 28.0 kDa, and 24 kDa, respectively, compared to the
∼23 kDa obtained in this study.

3.2. Trypsin simile immobilization in the magnetic chitosan

Evaluation of the best time for contact between the enzyme and
magnetic chitosan support was considerably variable over the time
range studied, therefore, a significant difference (p = 0.34) was not
observed between established conditions. Although there are evidences
that an incubation time longer than 8 h saturates the support and
consequently wastes the activity (Xu, Deng, Yang, & Zhang, 2007), in
addition to a possible deformation of reactive sites as shown by Simons,
Mosisch, Torda, and Hilterhaus (2013), this characteristic was not ob-
served in this study. Thus, it is possible to use any time between one and
sixteen hours in the immobilization, without quantitatively affecting
the immobilized enzymatic activity.
Chitosan is already a recognized support because of its protein im-
mobilization capacity and its advantage when compared to other sup-
ports (Vieira et al., 2013), and also due to features such as its elevated
biocompatibility, heavy metal chelation, hydrophilicity, the fact that it
is not toxic, and for displaying an elevated affinity to proteins
(Krajewska, 2004). Chitosan magnetization, on the other hand, can add
other advantages such as low preparation costs and the possibility of
easy separation of reaction products. Lastly, it reinforces the knowledge
that immobilized enzymes can be considered the most promising
technique for high-efficiency biotechnological processes, mainly for
environmental investigation, diagnostics, biotransformation, pharma-
ceutical and food industries (Datta, Christena, & Rajaram, 2013).

3.3. Effects of temperature, pH and storage time

Fig. 3 summarizes the results obtained for temperature and pH ef-

fects on the immobilized enzyme. The maximum catalytic activity ob-
served was 50 °C for porcine and tilapia trypsins. However, tilapia
proteases showed a little more stability, especially above 50 °C in
comparison to commercial trypsins. Similar results were previously
observed for catalytic activity and optimum temperature of im-
mobilized tilapia trypsin (Amaral et al., 2006). Analysis of these para-
meters is a fundamental stage for comprehending the best functioning
conditions of the immobilized enzyme, in addition to directing a future
application of the immobilized enzyme. It is important to note that the
purified enzyme possesses the same temperature profile compared to
commercial proteases in tested situations. On the other hand, pH var-
iations appear different than those previously observed (Amaral et al.,
2006). In the results of this research, it was observed that activity in-
creased rapidly at pH 9, and statistical differences were found between
tilapia enzyme and porcine enzyme activity. This can be explained by
the difference in the nature of the support used in this study. It should
be noted that tilapia trypsin showed significantly greater catalytic pH
activity between 8.5 and 9.0. The continuity of the evaluation at higher
pH severely compromised the stability of the support. In these situa-
tions, the magnetic chitosan composite started an unmaking process.
In the storage experiment, immobilized tilapia trypsin showed high
activity retention (Fig. 4). After 30 days of storage, a non-significant
loss of 5% was observed in enzymatic activity (p = 0.49). Storage of
tilapia trypsin for 60 days only resulted in a 13% loss in activity, and
this level of activity was maintained when trypsin was stored for
90 days. Small initial losses might represent weakly-bound enzyme
leaching from the support (Kondyurin, Nosworthy, & Bilek, 2011), and
retained maintenance level after 60 days of storage shows stabilization
of the immobilized enzyme. The stability observed effectively enables
application of this derivative in case that storage is absolutely necessary
Fig. 3. (A) Effect of temperature variation in tilapia and immobilized porcine trypsin.
Results are shown as relative activity (%) compared to the highest activity recorded
during the experiment. Identical uppercase letters represent statistically non-significant
tilapia trypsin (α = 0.05). Identical lowercase letters represent statistically non-sig-
nificant averages for porcine trypsin (α = 0.05). The symbol # represents statistically-
significant averages (α = 0.05) between trypsins. Bars represent standard area for each
test carried out. Red lines represent porcine trypsin while black lines represent tilapia
trypsin. (B) Relative activities (%) in pH variation. Same uppercase letters represent
statistically non-significant averages for tilapia trypsin (α = 0.05). Identical lowercase
letters represent statistically non-significant averages for porcine trypsin (α = 0.05). Bars
represent standard error for each test performed. (C) Relative activity (%) effect of storage
time on immobilized enzyme. Letters a and b represent a statistic (α = 0.05) between
considered groups. Experiment performed in quadruplicate (n = 4).
for commercialization (Datta et al., 2013).
The immobilized enzyme showed 128 ± 6.64% of relative average
activity after 20 reuses. This increase in catalytic activity is probably
associated with enzyme stabilization in the support. Similar conditions
were registered by Singh et al. (2011), where there is considerable in-
crease in catalytic activity, approximately 55%, with immobilized en-
zyme reuses. This difference can be related to greater enzyme stability
in the support due to the presence of cysteine joined with histidine in
the immobilized protease. The increase recorded in the reuse test is
considerably higher than that reported by Andre et al. (2016). Viability
after 20 reuses is also higher than that observed by Maciel et al. (2012)
which, under similar conditions, used Levan magnetic particles as
support for trypsin immobilization and obtained enzyme activity of
84.43 ± 5.42% after 10 reuses.

306
R.D.S. Azevedo et al.
a ingredient due to excellent amino acids profile, and apparently does not
100 ab
b b
90
80
70
Relative Activity (%)
60
50
40
30
20
10
0
0 20 40 60 80
Storage Time (Days)
Fig. 4. Storage time for the immobilized enzyme in chitosan-based magnetic support.
Results are shown as% relative to the activity prior to storage. Bars represent standard
deviation. Identical letters represent statistically non-significant averages (α: 0.05).
BApNA 8 mM was used in each test as a substrate and the activity was conducted at 25 °C.
Experiment performed in triplicate (n = 3).
3.4. Antinutrients detection
Relative activity (%) of free and immobilized protease inhibition
after antinutrients incubation is shown in Table 2. Firstly, immobilized
tilapia trypsin (ITT) and porcine trypsins (IPT) have more capacity for
enzymatic activity maintenance in comparison to free trypsins. Second,
enzyme immobilization has a more precise tool for detection of differ-
ences between the enzymatic activity of tilapia and porcine trypsins
p < 0.007. This difference cannot always be observed in assays with
free proteases. Interestingly, in both situations tilapia trypsins showed
more damage in catalytic property when compared to porcine trypsins
for practically all tested diets. This finding suggest that trypsin from
tilapia can be more affected by the possible presence of antinutrients in
the studied diets compared to porcine trypsin. This feature reinforces
the peculiarities of molecules obtained from aquatic organisms as pre-
viously mentioned (e. g. (Ferraro et al., 2010) and provides a warning
about possible mistakes in scientific studies that use only one source of
animal enzyme to investigate digestive physiology. Especially when the
goal is the extrapolation of results found with mammalian trypsin to
other animal groups.
After incubation with diet 1, ITT lost more than 30% of enzymatic
activity compared to IPT (p = 0.003), a difference of 56% between
these two trypsins. Here IPT lost only 9.1% after incubation with diet 1,
while ITT lost 36.7% of enzymatic activity. The inclusion of vegetable
protein in diet 1 is around 51%, and it is well-known that vegetal in-
gredients are responsible for inhibiting digestive proteases of Sparus
aurata, Oreochromis niloticus and Solea senegalensis (Moyano López,
Martı́nez Dı́az, Dı́az López, & Alarcón López, 1999; Perera and Yúfera,
2017). For that reason, even with a low inclusion of soybean meal,
which is an antinutrients recognized source in aquafeeds, other vege-
table ingredients may reduce digestive capacity. This fact can also
compromise nutrient bioavailability in addition to fish feed conversion.
As such the possibility of error in estimative of feed conversion by fish
perhaps high when based on mammalian proteases activity.
The ITT maintained 77.2% of initial activity after diet 2 incubation.
This result represents 85% of IPT activity. A slight difference between
immobilized trypsins were also observed after diet 3 incubation
(p = 0.000). However, they represent only 7% among immobilized
trypsins. Remarkably, diet 3 contains more animal protein compared to
other studied diets, especially fish meal. This feature may explain the
slight differences observed. Not least since fish meal is a better protein
Food Chemistry 257 (2018) 302–309
induce intestinal inflammation or inhibit digestive enzymes (Bajpai
et al., 2005; Chikwati et al., 2012).
After incubation with diet 4, ITT kept 58.9% of its initial activity. In
this case, the difference between immobilized proteases (tilapia and
pork) reaches 32% (p = 0.04). The higher difference among ITT and
IPT at diet 4 can be related to the lowest animal protein inclusion
(around 20%) of the aquafeeds studied. This result reinforces discus-
sions about beneficial effects of animal protein (specially fish meal) on
growth performance (Martínez-Alvarez, Chamorro, & Brenes, 2015).
The best activity for immobilized porcine trypsin was observed after
incubation with diet 5 (99.7%). At the same conditions, ITT showed
71.9% of residual activity. The lowest difference between porcine and
tilapia immobilized trypsin in response to aquafeed was found in diet 6
(p = 0.007 to immobilized enzymes and p = 0.197 to free enzymes).
On the other hand, the highest activity of all tests (immobilized and free
trypsins) was observed for immobilized tilapia trypsin and free tilapia
100
trypsin after diet 7 incubation. Lastly, the results for enzymatic activity
after diet 8 incubation was different in both experimental conditions
(p = 0.000). In this test, diet 8 strongly inhibits ITT and free tilapia
trypsin. At this condition, ITT kept only 5.3% while IPT kept 99.2% of
initial activity.
Interestingly, diet 5 contains more corn meal (50% of dry matter)
compared to other diets studied. The satisfactory hydrolysis by porcine
trypsins after diet 5 incubation shows that corn meal can be a low-cost
alternative ingredient to animal feed. Especially in countries like Brazil,
where corn production is high. However, the final amount of vegetable
ingredients in diet 5 is much higher, where corn and soybean meal are
responsible for 70% of dry matter, followed only by diet 6, with 63% of
vegetable ingredients. IPT holds 99.7% of initial activity while ITT
holds just 72% after diet 5 incubation. After diet 6 incubation, IPT lost
only 25.8%, while ITT lost 39% of activity. These results are very si-
milar to those of diet 2 (which is the most balanced diet in terms protein
sources among the diets studied). This suggests that corn meal inclusion
can minimize aquafeeds’ production costs and antinutrients effects.
Unfortunately, however, diet 5 contains a considerable inclusion of
lipid and carbohydrates in its composition, and toxic effects of these
ingredients have been identified to Gadus morhua larvae (Saele et al.,
2013). There are also evidences that these ingredients influence the
immune response of grass carp (Jin et al., 2013). For Nile Tilapia, it is
known that lipids accumulate in viscera and carcasses, affecting their
commercialization. Correlating this information with an elevated pre-
sence of lipids/carbohydrates and ITT enzymatic activity, these points
lead us to hypothesize that these ingredients can strongly delay fish
digestive processes.
After diet 7 incubation, it becomes clear that the immobilization
process resulted in more stability to enzymatic catalysis. Thus, it is
possible to assume that free trypsins are more susceptible to inhibition
than ITT and IPT. Since this is the physiological condition, greater at-
tention is necessary on the possible deleterious effects of this inhibition,
including due to the pathophysiological potential of antinutrients like
protease inhibitors found in dietary components (see (Chikwati et al.,
2013; Krogdahl, Penn, Thorsen, Refstie, & Bakke, 2010).
Diet 8 has the most inhibitor potential to tilapia proteases, espe-
cially to ITT, which lost around 94.7% of initial activity, while IPT lost
only 0.8%. The highest difference observed is probably due to the sy-
nergic soybean meal inclusion in diet 8. Indeed, diet 8 had the highest
inclusion of soybean meal (27.12%) among all aquafeeds studied. This
percentage is more than 40% of that observed in diet 1 (lower soybean
meal inclusion). Therefore, attention should be given to this data be-
cause soymeal is currently the largest source of vegetable protein in
Nile tilapia diets (Plaipetch, 2013), and soybean meal induced negative
effects on growth, feed utilization and immunological status in models
fishes like Danio rerio, Salmo salar and Oncorhynchus mykiss (Baeverfjord
and Krogdahl, 1996; Hedrera et al., 2013; Refstie et al., 2000).
In addition, is well-know that vegetable ingredients are rich in
307
R.D.S. Azevedo et al.
protease inhibitors. About this, Kunitz and Bowman-Birk families, are
easily found in these vegetable ingredients and are recognized for de-
crease in activity of digestive enzymes such as trypsin (Becker-Ritt,
Mulinari, Vasconcelos, & Carlini, 2004). To O. niloticus this protease
inhibition is a very important aspect because can impair the adequate
fish growth (due impair feed utilization) and consequently affects
economic aquaculture.
The immobilization process minimized deleterious effects of anti-
nutrients to ITT and IPT compared to free trypsins. This feature shows
that the immobilization at magnetic chitosan support improved cata-
lytic properties of trypsins, besides providing higher stability to tilapia
and porcine trypsins. Thus, this result corroborates the knowledge that
the enzymatic immobilization improves catalysis property
(Bayramoglu, Doz, Ozalp, & Arica, 2017; Mohamad, Marzuki, Buang,
Huyop, & Wahab, 2015).
Hence, once we have demonstrated that immobilized trypsins are
more appropriate to detect differences between tilapia and porcine
trypsins. About this, Outzen, Berglund, Smalås, & Willassen, 1996 also
showed that another fish trypsin (Atlantic salmon) works differently
against enzymatic inhibitors when compared to bovine trypsin. In their
work, theses authors showed that chelators used like metal trypsin in-
hibitor strongly reduce salmon trypsin (−49%) while bovine trypsin
remains with more than 95% of activity (Outzen et al., 1996). There-
fore, notes that fish trypsins apparently can be easily inhibited com-
pared to porcine or bovine trypsins. Remarkably, fish proteases are
well-known by sophisticate features of pH and temperature stability,
low molecular weight and high oxidizing properties (Bezerra et al.,
2005; Esposito et al., 2010; Freitas et al., 2012). These particularities
are useful examples to explain that fish enzymes can respond in another
way against external factors. Summarily, the results obtained suggest
that the use of specific enzymes (target animal) is a more accurate way
to estimate the true nutrient bioavailability to feeding aquatic organ-
isms in antinutrients context. Moreover, antinutrients effects on free
enzymes also deserve attention for being found at physiological con-
ditions. Indeed, this characteristic was previously demonstrated (Bajpai
et al., 2005; Perera and Yúfera, 2017) and leads to damage to biological
fitness, a fundamental feature for right fish growth.
4. Conclusions
Based on stability to temperature and pH variations, as well con-
tinuous enzymatic activity during storage time experiment, it is possible
to suggest that magnetic chitosan composite can be a useful tool for
enzyme immobilization, particularly considering the use of fish pro-
teases. Thus, bioelectrochemical characterization of enzyme support
may expand possibilities to magnetic-chitosan composite applications.
Besides that, the results obtained suggest that fish trypsin can be more
sensitive to the antinutrients’ presence compared to commercial trypsin
due to a strong inhibition after contact with tested diets. Lastly, im-
mobilized fish trypsin is a viable way to better understand how anti-
nutrients can impair digestive enzymes of aquatic organisms.
Acknowledgments
The authors are grateful to Conselho Nacional de Desenvolvimento
Científico e Tecnológico – CNPq (Processes 135143/2011-0; 459809/
2014-8), Financiadora de Estudos e Projetos – Finep/RECARCINA
(Agreement: 01.13.0220.00) and FAPEMA for the financial support,
and to Dr. Daniela Ferraz Baconi Campeche who provided the aquafeed
used in this research.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.foodchem.2018.03.034.
308
Food Chemistry 257 (2018) 302–309
References
Alves, M. H. M. E., Nascimento, G. A., Cabrera, M. P., Silvério, S. I.d. C., Nobre, C.,
Teixeira, J. A., & de Carvalho, L. B. (2017). Trypsin purification using magnetic
particles of azocasein-iron composite. Food Chemistry, 226, 75–78.
Amaral, I., Carneiro-da-Cunha, M., Carvalho, L., & Bezerra, R. (2006). Fish trypsin im-
mobilized on ferromagnetic Dacron. Process Biochemistry, 41, 1213–1216.
Andre, J., Saleh, D., Syldatk, C., & Hausmann, R. (2016). Effect of spacer modification on
enzymatic synthetic and hydrolytic activities of immobilized trypsin. Journal of
Molecular Catalysis B: Enzymatic, 125, 88–96.
Baeverfjord, G., & Krogdahl, A. (1996). Development and regression of soybean meal
induced enteritis in Atlantic salmon, Salmo salar L., distal intestine: A comparison
with the intestines of fasted fish. Journal of Fish Diseases, 19, 375–387.
Bagheri, H., Asgharinezhad, A. A., & Ebrahimzadeh, H. (2016). Determination of trace
amounts of Cd(II), Cu(II), and Ni(II) in food samples using a novel functionalized
magnetic nanosorbent. Food Analytical Methods, 9, 876–888.
Bajpai, S., Sharma, A., & Gupta, M. (2005). Removal and recovery of antinutritional
factors from soybean flour. Food Chemistry, 89, 497–501.
Bayramoglu, G., Doz, T., Ozalp, V. C., & Arica, M. Y. (2017). Improvement stability and
performance of invertase via immobilization on to silanized and polymer brush
grafted magnetic nanoparticles. Food Chemistry, 221, 1442–1450.
Becker-Ritt, A. B., Mulinari, F., Vasconcelos, I. M., & Carlini, C. R. (2004). Antinutritional
and/or toxic factors in soybean (Glycine max (L) Merril) seeds: Comparison of dif-
ferent cultivars adapted to the southern region of Brazil. Journal of the Science of Food
and Agriculture, 84, 263–270.
Bezerra, R., Lins, E., Alencar, R., Paiva, P., Chaves, M., Coelho, L., & Carvalho, L. (2005).
Alkaline proteinase from intestine of Nile tilapia (Oreochromis niloticus). Process
Biochemistry, 40, 1829–1834.
Cahu, T., Santos, S., Mendes, A., Cordula, C., Chavante, S., Carvalho, L., ... Bezerra, R.
(2012). Recovery of protein, chitin, carotenoids and glycosaminoglycans from Pacific
white shrimp (Litopenaeus vannamei) processing waste. Process Biochemistry, 47,
570–577.
Chikwati, E., Sahlmann, C., Holm, H., Penn, H., Krogdahl, A., & Bakke, A. (2013).
Alterations in digestive enzyme activities during the development of diet-induced
enteritis in Atlantic salmon, Salmo salar L. Aquaculture, 402, 28–37.
Chikwati, E., Venold, F., Penn, M., Rohloff, J., Refstie, S., Guttvik, A., ... Krogdahl, A.
(2012). Interaction of soyasaponins with plant ingredients in diets for Atlantic
salmon, Salmo salar L. British Journal of Nutrition, 107, 1570–1590.
Couto, A., Kortner, T. M., Penn, M., Bakke, A. M., Krogdahl, Å., & Oliva-Teles, A. (2014).
Effects of dietary phytosterols and soy saponins on growth, feed utilization efficiency
and intestinal integrity of gilthead sea bream (Sparus aurata) juveniles. Aquaculture,
432, 295–303.
Datta, S., Christena, L. R., & Rajaram, Y. R. S. (2013). Enzyme immobilization: An
overview on techniques and support materials: 3. Biotech, 3, 1–9.
Esposito, T., Marcuschi, M., Amaral, I., Carvalho, L., & Bezerra, R. (2010). Trypsin from
the processing waste of the lane snapper (Lutjanus synagris) and its compatibility
with oxidants, surfactants and commercial detergents. Journal of Agricultural and Food
Chemistry, 58, 6433–6439.
Ferraro, V., Cruz, I., Jorge, R., Malcata, F., Pintado, M., & Castro, P. (2010). Valorisation
of natural extracts from marine source focused on marine by-products: A review. Food
Research International, 43, 2221–2233.
Freitas, A., Costa, H., Icimoto, M., Hirata, I., Marcondes, M., Carvalho, L., Oliveira, V., &
Bezerra, R. (2012). Giant Amazonian fish pirarucu (Arapaima gigas): Its viscera as a
source of thermostable trypsin. Food Chemistry, 133, 1596–1602.
Guillamon, E., Pedrosa, M., Burbano, C., Cuadrado, C., Sanchez, M., & Muzquiz, M.
(2008). The trypsin inhibitors present in seed of different grain legume species and
cultivar. Food Chemistry, 107, 68–74.
Hashemi, P., Bagheri, H., Afkhami, A., Amidi, S., & Madrakian, T. (2018). Graphene
nanoribbon/FePt bimetallic nanoparticles/uric acid as a novel magnetic sensing layer
of screen printed electrode for sensitive determination of ampyra. Talanta, 176,
350–359.
Hedrera, M., Galdames, J., Jimenez-Reyes, M., Reyes, A., Avendano-Herrera, R., Romero,
J., & Feijoo, C. (2013). Soybean meal induces intestinal inflammation in zebrafish
larvae. Plos One, 8.
Jayakumar, R., Prabaharan, M., Kumar, P., Nair, S., & Tamura, H. (2011). Biomaterials
based on chitin and chitosan in wound dressing applications. Biotechnology Advances,
29, 322–337.
Jin, Y., Tian, L., Zeng, S., Xie, S., Yang, H., Liang, G., & Liu, Y. (2013). Dietary lipid
requirement on non-specific immune responses in juvenile grass carp
(Ctenopharyngodon idella). Fish & Shellfish Immunology, 34, 1202–1208.
Khangembam, B. K., & Chakrabarti, R. (2015). Trypsin from the digestive system of carp
Cirrhinus mrigala: Purification, characterization and its potential application. Food
Chemistry, 175, 386–394.
Kishimura, H., Klomklao, S., Benjakul, S., & Chun, B.-S. (2008). Characteristics of trypsin
from the pyloric ceca of walleye pollock (Theragra chalcogramma). Food Chemistry,
106, 194–199.
Knudsen, D., Jutfelt, F., Sundh, H., Sundell, K., Koppe, W., & Frokiaer, H. (2008). Dietary
soya saponins increase gut permeability and play a key role in the onset of soyabean-
induced enteritis in Atlantic salmon (Salmo salar L.). British Journal of Nutrition, 100,
120–129.
Kondyurin, A., Nosworthy, N. J., & Bilek, M. M. (2011). Effect of low molecular weight
additives on immobilization strength, activity, and conformation of protein im-
mobilized on PVC and UHMWPE. Langmuir, 27, 6138–6148.
Kortner, T., Skugor, S., Penn, M., Mydland, L., Djordjevic, B., Hillestad, M., Krasnov, A., &
Krogdahl, A. (2012). Dietary soyasaponin supplementation to pea protein
R.D.S. Azevedo et al.
concentrate reveals nutrigenomic interactions underlying enteropathy in Atlantic
salmon (Salmo salar). BMC Veterinary Research, 8.
Krajewska, B. (2004). Application of chitin- and chitosan-based materials for enzyme
immobilizations: A review. Enzyme and Microbial Technology, 35, 126–139.
Krogdahl, Å., Penn, M., Thorsen, J., Refstie, S., & Bakke, A. M. (2010). Important anti-
nutrients in plant feedstuffs for aquaculture: An update on recent findings regarding
responses in salmonids. Aquaculture Research, 41, 333–344.
Kumirska, J., Weinhold, M., Thoming, J., & Stepnowski, P. (2011). Biomedical activity of
chitin/chitosan based materials-influence of physicochemical properties apart from
molecular weight and degree of N-acetylation. Polymers, 3, 1875–1901.
Liu, K., Li, X., Li, X., He, B., & Zhao, G. (2010). Lowering the cationic demand caused by
PGA in papermaking by solute adsorption and immobilized pectinase on chitosan
beads. Carbohydrate Polymers, 82, 648–652.
Maciel, J., Andrad, P., Neri, D., Carvalho, L., Cardoso, C., Calazans, G., ... Silva, M.
(2012). Preparation and characterization of magnetic levan particles as matrix for
trypsin immobilization. Journal of Magnetism and Magnetic Materials, 324,
1312–1316.
Martínez-Alvarez, O., Chamorro, S., & Brenes, A. (2015). Protein hydrolysates from an-
imal processing by-products as a source of bioactive molecules with interest in animal
feeding: A review. Food Research International, 73, 204–212.
Mohamad, N. R., Marzuki, N. H. C., Buang, N. A., Huyop, F., & Wahab, R. A. (2015). An
overview of technologies for immobilization of enzymes and surface analysis tech-
niques for immobilized enzymes: Biotechnology. Biotechnological Equipment, 29,
205–220.
Moyano López, F. J., Martı́nez Dı́az, I., Dı́az López, M., & Alarcón López, F. J. (1999).
Inhibition of digestive proteases by vegetable meals in three fish species; seabream
(Sparus aurata), tilapia (Oreochromis niloticus) and African sole (Solea senegalensis).
Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology,
122, 327–332.
Outzen, H., Berglund, G. I., Smalås, A. O., & Willassen, N. P. (1996). Temperature and pH
sensitivity of trypsins from atlantic salmon (Salmo salar) in comparison with bovine
and porcine trypsin. Comparative Biochemistry and Physiology Part B: Biochemistry and
Molecular Biology, 115, 33–45.
Perera, E., & Yúfera, M. (2017). Effects of soybean meal on digestive enzymes activity,
expression of inflammation-related genes, and chromatin modifications in marine
fish (Sparus aurata L.) larvae. Fish Physiology and Biochemistry, 43, 563–578.
309
Food Chemistry 257 (2018) 302–309
Plaipetch, P. (2013). Effect of replacing soybean meal with yeast-fermented canola meal on
growth and nutrient retention of Nile tilapia, Oreochromis niloticus (Linnaeus 1758) (pp.
1–10). .
Refstie, S., Korsoen, O., Storebakken, T., Baeverfjord, G., Lein, I., & Roem, A. (2000).
Differing nutritional responses to dietary soybean meal in rainbow trout
(Oncorhynchus mykiss) and Atlantic salmon (Salmo salar). Aquaculture, 190, 49–63.
Saele, O., Nordgreen, A., Olsvik, P., Hjelle, J., Harboe, T., & Hamre, K. (2013). Toxic
effects of dietary hydrolysed lipids: An in vivo study on fish larvae. British Journal of
Nutrition, 109, 1071–1081.
Seo, D., Jang, Y., Park, R., & Jung, W. (2012). Immobilization of chitinases from
Streptomyces griseus and Paenibacillus illinoisensis on chitosan beads. Carbohydrate
Polymers, 88, 391–394.
Silva, J., Esposito, T., Marcuschi, M., Ribeiro, K., Cavalli, R., Oliveira, V., & Bezerra, R.
(2011). Purification and partial characterisation of a trypsin from the processing
waste of the silver mojarra (Diapterus rhombeus). Food Chemistry, 129, 777–782.
Simons, J., Mosisch, M., Torda, A., & Hilterhaus, L. (2013). Site directed immobilization
of glucose-6-phosphate dehydrogenase via thiol-disulfide interchange: Influence on
catalytic activity of cysteines introduced at different positions. Journal of
Biotechnology, 167, 1–7.
Singh, A., Singh, S., Suthar, N., & Dubey, V. (2011). Glutaraldehyde-activated chitosan
matrix for immobilization of a novel cysteine protease. Procerain B: Journal of
Agricultural and Food Chemistry, 59, 6256–6262.
Vieira, D. C., Lima, L. N., Mendes, A. A., Adriano, W. S., Giordano, R. C., Giordano, R. L.
C., & Tardioli, P. W. (2013). Hydrolysis of lactose in whole milk catalyzed by beta-
galactosidase from Kluyveromyces fragilis immobilized on chitosan-based matrix.
Biochemical Engineering Journal, 81, 54–64.
Weska, R., Moura, J., Batista, L., Rizzi, J., & Pinto, L. (2007). Optimization of deacety-
lation in the production of chitosan from shrimp wastes: Use of response surface
methodology. Journal of Food Engineering, 80, 749–753.
Xu, X., Deng, C., Yang, P., & Zhang, X. (2007). Immobilization of trypsin on super-
paramagnetic nanoparticles for rapid and effective proteolysis. Journal of Proteome
Research, 6, 3849–3855.
Yamamoto, T., Murashita, K., Matsunari, H., Sugita, T., Furuita, H., Iwashita, Y., ...
Suzuki, N. (2012). Influence of dietary soy protein and peptide products on bile acid
status and distal intestinal morphology of rainbow trout Oncorhynchus mykiss.
Fisheries Science, 78, 1273–1283.