Injection of Natural Protease Inhibitors and

Evaluation of Their Impact on Cooked Pacific

Whiting (Merluccius productus) Fillets
Silvana Dinaintang Harikedua and Christina A. Mireles DeWitt

Abstract: Brine injection of natural proteases, dried egg white (EW), and dried potato extract (PE), was investigated as a
means to prevent softening during cooking of Pacific whiting fillets. Treatments included fillets injected with Water (W),
3% sodium chloride and 3% sodium tripolyphosphate brine (B), B with 0.1% xanthan gum (BXG ), B with 1%, 2%, or 3%
EW (B1-3EW ), B with 1%, 2%, or 3% PE (B1-3PE ) or non-injected (NI). Non-injected Pacific cod (Cod) was also utilized as
a reference for texture analysis. Fillets were subsequently cooked using cooking protocol 1 (90 °C for 20 min) or cooking
protocol 2 (60 °C for 30 min then 90 °C for 20 min). Cooked fillet pH, moisture, total protein, total extractable protein
(TEP), total non-extractable protein (TNEP), texture profile analysis (TPA), and electrophoretic pattern (SDS-PAGE)

Food Chemistry
were measured. Cooking protocol 2 significantly reduced moisture, increased total protein, TEP and TNEP. Loss of
moisture suggested cooking protocol 2 promoted proteolysis. Electrophoretic evaluations of the fillets treated with the
cooking protocol 2 confirmed proteolysis was enhanced by cooking protocol 2 and myosin band integrity was protected
in fillets containing natural protease inhibitors. For TPA, there was no significant difference between protease inhibitor
types although means for BEW did trend higher than BPE . However, PCA analysis clearly demonstrated B2EW and B3EW
fillets were most similar to Cod. Results demonstrated brine injection with protease inhibitors prevented fillet softening
during cooking.
Keywords: brine injection, cooked fillets, Pacific whiting, protease inhibitor, texture profile analysis

Practical Application: Texture softening in fish has typically been addressed by diverting product to lower-value minced
applications where ingredients can be blended-in to counteract softening caused by enzymes. Investigations demonstrate
feasibility of brine injection of protease inhibitor ingredient, dried egg white, to preserve protein integrity during cooking.
Dried potato extract required a suspension aide, xanthan gum, in order to become sufficiently suspended in the brine
for injection. Dried potato extract was able to inhibit protease activity, however, the suspension aide negatively impacted
protein–protein interactions during cooking. An alternative suspension aide is therefore required for dried potato extract.

Introduction
Texture softening of fillets is a significant problem for many
different species of fish. Both Pacific whiting (Merluccius productus)
and arrowtooth flounder (Atheresthes stomia) fillets have limited
acceptance in the retail market. This is due to the development
of a mushy texture during cooking which is caused by a heat-
induced cysteine protease. As a result, fillets from these species are
low value and often are converted into a minced product to make
surimi. Their conversion to surimi is made possible by the incor-
poration of protease inhibitors (An, Peters, & Seymour, 1996; An,
Weerasinghe, Seymour, & Morrissey, 1994; Morrissey, Hartley, &
An, 1996; Porter, Koury, & Kudo, 1993; Rawdkuen, Benjakul,
JFDS-2017-1345 Submitted 8/15/2017, Accepted 2/23/2018. Author Harike-
dua is with Dept. Fish Processing Technology, Faculty of Fisheries and Marine Science,
Sam Ratulangi Univ., Manado, North Sulawesi, Indonesia. Author Mireles DeWitt is
with OSU Seafood Research & Education Center Experiment Station, Dept. of Food
Science and Technology, Oregon State Univ., Astoria, OR, U.S.A. Direct inquiries to
author Harikedua Dept. Fish Processing Technology, Faculty of Fisheries and Marine
Science, Sam Ratulangi Univ., Manado, North Sulawesi 95115, Indonesia. (E-mail:
silvana.harikedua@unsrat.ac.id).
Previous address: Author Harikedua was a graduate student at OSU Seafood
Research & Education Center Experiment Station, Dept. of Food Science and
Technology, Oregon State Univ., Astoria, OR 97103, USA.
Visessanguan, & Lanier, 2007; Reppond & Babbit, 1993, 1997;
Wasson, Reppond, Babbitt, & French, 1993). Ingredients such as
1% beef plasma protein, 2% dried egg white and 3% dried potato
extract have all been identified as effective protease inhibitors for
surimi products (Morrissey, Wu, Lin, & An, 1993).
Only two studies have investigated the injection or immersion
of protease inhibitors into fillets to prevent their softening upon
cooking. Both studies were conducted using arrowtooth flounder.
Babbitt (1990) investigated brine treatments that included beef ex-
tract, albumin proteins, egg white proteins, “various phosphates,”
and EDTA as potential inhibitors. Kang and Lanier (2005) in-
vestigated a purified recombinant cystatin produced by Escherichia
coli. These authors reported inhibition and prevention of soften-
ing upon cooking using injected or immersed inhibitor ingre-
dients. However, brine composition utilized was either obscured
for proprietary reasons (Babbitt, 1990) or not feasible because the
inhibitor utilized is not commercially available (Kang & Lanier,
2005). Preliminary experiments were conducted to determine the
maximum level natural occurring protease inhibitors, such as dried
egg white (BEW ) or dried potato extract (BPE ), could be suspended
in a brine. Dried egg white could be sufficiently suspended in a
3% salt (NaCl) and 3% sodium tripolyphosphate (STPP) brine
without a suspension aide up to 3% w/w. Dried potato ex-
tract required a suspension aide. Xanthan gum was selected due
to its rapid solubility in a cold salt/phosphate brine and it was

C 2018 Institute of Food Technologists

 R

doi: 10.1111/1750-3841.14126 Vol. 0, Iss. 0, 2018 r Journal of Food Science 1

Further reproduction without permission is prohibited
 Injection of natural protease inhibitors . . .
determined a minimum of 0.1% was required to achieve sus-
pension of BPE up to 3% w/w. Initial studies were subsequently
designed to determine whether 3% BEW or 3% BPE could be suf-
ficiently distributed by injection into a fillet to inhibit protease ac-
tivity (Harikedua & Mireles DeWitt, 2017). Results demonstrated
that both BPE and BEW inhibited protease activity in the raw fillet.
However, a preliminary sensory evaluation of the cooked fillet re-
sulted in either a distinct eggy or slight starch taste that masked fish
flavor. The objective of this study, therefore, was to determine the
minimal level of dried egg white and dried potato extract required
to prevent softening of cooked Pacific whiting fillets.
Materials and Methods
Raw materials
Pacific whiting (Merluccius productus) (±300 fish) <24 hr of
Food Chemistry
postharvest, were obtained once from a local processor. Fish used
for this study were 27 to 30 cm in length and 235 ± 40 g in weight.
Fish were transported to the OSU Seafood Laboratory in an in-
sulated container in ice (transportation time <15 min). Fish were
eviscerated, butterflied and immediately frozen at –18 °C. Vacuum
refined granulated salt (Morton Salt Inc., Chicago, IL, U.S.A.),
sodium tripolyphosphate (STPP) (Nutrifos-088, Integra Chem-
ical, Kent, WA), dried egg white (DEW) (P-110, Henningsen
Food, Omaha, NE), dried potato extract (DPE) (NP3, Pacific
Blends Ltd., Port Coquitlam, BC, Canada), and xanthan gum
(XG) (Pacific Blends Ltd.) were used for preparation of brines.
Fish injection
This study utilized previously frozen fillets as this would be the
most likely product form available for further processing. Prior
to injection, butterflied fillets were thawed for 10 to 14 hr in
a cold room (2.0 ± 1.0 °C). The final internal temperature of
the butterflied fillet ranged from 0 to 1 °C. All injections were
conducted in a cold room 3.0 ± 1.0 °C.
The level of NaCl and STPP selected for brines in this study
was based on previous studies (Almli & Hersleth, 2013; Bigelow &
Lee, 2007; Harikedua & Mireles DeWitt, 2017) and recommenda-
tions from the phosphate supplier. Ten treatments were evaluated
(Table 1): (1) no injection (NI), (2) injection of water (W), (3)
injection of base brine containing 3% salt and 3% STPP (B), (4)
injection of base brine coupled with 1% EW (B1EW ), (5) injec-
tion of base brine plus 2% EW (B2EW ), (6) injection of base brine
and 3% EW (B3EW ), (7) injection of base brine combined with
the addition of 0.1% XG as a suspension aid (BXG ), (8) injection
of base brine, 0.1% XG, and 1% PE (B1PE ), (9) injection of base
brine, 0.1% XG, and 2% PE (B2PE ), and (10) injection of base
brine, 0.1% XG, and 3% PE (B3PE ).
Brine target uptake was 10% (w/w) of initial weight. But-
terfly fillets (n = 1/treatment/replication) were injected using
a multichannel pipette fitted with 22-gauge hypodermic needles
(Eppendorf, Hauppauge, N.Y., U.S.A.). In total, 4 (n = 4) fillets for
each treatment were used in this study. Butterfly fillet weight was
obtained immediately prior to injection (w0 ). In addition, weight
was remeasured after an equilibration period of 30 min follow-
ing injection (w1 ). Brine uptake was calculated (w1 – w0 )/w0 ×
100%. Following brine equilibration, butterfly fillets were halved
and each half was placed individually in a non-air permeable bag
(Oxygen transmission rate 24 hr at 23 °C is 63 cc/sq.m, Summit
Packaging, Auburn, WA, U.S.A.), vacuumed to 0.9 bar (setting
at 7), sealed for 15 s (setting at 7) (Reiser, Canton, MA, U.S.A.).
2 Journal of Food Science r Vol. 0, Iss. 0, 2018
Vacuum packaged fillets were stored on ice for ±30 min prior to
being randomly assigned to a cooking protocol.
Cooking protocols
Cooking protocols were adapted from Kang and Lanier (2005).
Internal temperature was measured using a t-type thermocouple
and measurements were taken every 5 s. A target internal tem-
perature of 62.8 °C (145 °F) was selected for cooking protocol
1 as it is the safe minimum internal temperature as suggested
by Food Safety and Inspection Service, United States Dept. of
Agriculture (USDA, 2012). Briefly, one packaged fillet (–2 ±
1 °C) was fully submerged in approximately 20 L at 90 °C using
a recirculating water bath (WB1120 Lindberg Blue M, Waltham,
MA, U.S.A.) with lid cover for 20 min. For cooking protocol
2, fillets were similarly submerged in 60 °C water and held for
30 min, and then transferred to 90 °C and held for 20 min. Fol-
lowing heat treatment, fillets were submerged in an ice slurry
for 15 min (average temperature of ice slurry was –1.0 °C) then
immediately frozen (–18 °C) and stored at –18 °C until further
analysis.
pH, moisture, and protein content measurement
Duplicate pH measurements were obtained of the brine used for
injection and in the cooked fillets (Accumet
R
Excel XL15, Fisher
Scientific, Pittsburg, PA, U.S.A.). One gram of cooked sample
was homogenized with 10 mL of deionized water (Sheard, Nute,
Richardson, Perry, & Taylor, 1999). The moisture content was
determined by AOAC methods 950.46 (AOAC, 2005). Samples
were weighed 3.0 ± 0.05 g then dried at 105 °C overnight (12
to 18 hr) until a constant weight was obtained. The total crude
protein content was determined by Kjeldahl method according to
AOAC method 981.10 (AOAC, 2005). Duplicate measurement
per treatment per replication were recorded for both moisture and
protein content.
Determination of total extractable protein (TEP) and total
non-extractable protein (TNEP) content
Duplicate samples were weighed, 0.6 ± 0.02 g, into cen-
trifuge tubes. Samples were extracted with 7.5 mL buffer con-
tain 8 M Urea (Alfa Aesar, Haverhill, MA, U.S.A.), 2% SDS
(sodium dodecyl sulphate) (Bio-Rad, Hercules, CA, U.S.A.), 2%
β-mercaptoethanol (Sigma-Aldrich, St. Louis, MO, U.S.A.), and
20 mM Tris-HCl (Lonza, Alpharetta, GA, U.S.A.) as outlined by
Pongviratchai (2002). Protein content was measured using Quick
StartTM Bradford Protein Assay Kit 1 #5000201 (Bio-Rad) using
bovine serum albumin (BSA) as a standard (Quick Start Bovine
Serum Albumin Standard #5000206 Bio-Rad) and reported as an
average of the duplicate. For TEP, supernatant from the sample
extraction described above was diluted with deionized water (dd
H2 O) at a ratio of 0.1:3.4 mL and then tested for protein content.
The supernatant was diluted using dd H2 O to lessen the concen-
tration of interfering compounds from the extraction buffers, that
is, Urea, SDS, β-mercaptoethanol, and Tris-HCl that would inter-
fere with protein determination. The absorbance was measured at
595 nm using spectrophotometer UV-VIS 2401 PC (Shimadzu,
Kyoto, Japan). TNEP content was calculated by subtracting the
total extractable protein content from crude protein and reported
as an average of the duplicate.
Electrophoretic pattern
The supernatant from TEP was further analyzed with
Sodium Dodecyl Sulphate-Polyacrylamide Gel Electrophoresis
Injection of natural protease inhibitors . . .

Table 1–Treatments designation, characteristics of solution, and brine uptake (%) of Injected Pacific whiting (Merluccius productus)
fillet.

Abbreviation Brine composition and concentration pH in solution Brine uptake (%)∗

NI No injection – –
W Water 7.00 9.3 ± 1.2
B 3% salt (S) + 3% sodium tripolyphosphate (STPP) 8.50 9.4 ± 0.6
B1EW 3% S + 3% STPP + 1% Dried Egg White (EW) 8.39 9.3 ± 1.6
B2EW 3% S + 3% STPP + 2% Dried Egg White (EW) 8.37 9.8 ± 0.8
B3EW 3% S + 3% STPP + 3% Dried Egg White (EW) 8.34 9.7 ± 1.2
BXG 3% S + 3% STPP +0.1% Xanthan Gum (XG) 8.42 9.0 ± 2.3
B1PE 3% S + 3% STPP + 0.1% XG + 1% Dried Potato Extract (PE) 8.28 10.6 ± 1.0
B2PE 3% S + 3% STPP + 0.1% XG + 2% Dried Potato Extract (PE) 8.23 11.4 ± 1.4
B3PE 3% S + 3% STPP + 0.1% XG + 3% Dried Potato Extract (PE) 8.16 10.8 ± 1.3
∗
(Mean ± standard deviation).

calibrated to the load cell 1 kg and the height of the probe was ad-
justed to 15 cm. The test conditions were two consecutive cycles

Figure 1–Sample portion for texture analysis.∗
∗ The structure of fish muscle adapted from Dunajski(1980).
(SDS-PAGE) as outlined by Laemmli (1970) using 10% TGX
criterion precast gel (Bio-Rad). One sample per treatment per
replication was evaluated by SDS-PAGE. The protein content of
samples was adjusted to 4 mg/mL using 5% SDS (Bio-Rad) and
loaded 15 μL into the well. Kaleidoscope precision plus proteinTM
all blue standard #1610732 (Bio-Rad) was used as a protein stan-
dard. Completed electrophoresis gels were washed with distilled
water for 5 min three times. Gels were placed in 50 mL of 0.125%
(w/v) Coomassie dye (Brilliant Blue Coomassie G-250) (Bio-
Rad) and gently shaken on an orbital shaker heated to 52 °C for
1 hr. The gels were destained using 100 mL of 50% (v/v) methanol
(BDH R
VWR Chemicals, Radnor, PA, U.S.A.) and 10% (v/v)
glacial acetic acid (Sigma-Aldrich) for 1 hr while being heated to
52 °C on an orbital shaker (3528-5 Orbit Enviro-Shaker, Lab-Line
Instrument Inc., Melrose Park, IL, U.S.A.). After 1 hr, gels were
placed into 100 mL of 10% (v/v) methanol and 10% (v/v) glacial
acetic acid for another hour, after which the gels were placed in
100 mL of distilled water for 1 hr. Molecular weights were de-
termined by comparing the relative band mobility of a molecular
standard with Gel Doc XR System (Bio-Rad).
Textural measurement
Cooked fillets were partially thawed in a cold room (3 ± 1 °C)
for about 2 hr (average temperature around –1.1 °C). From each
fillet approximately 3 cubes (2.5 cm × 2.5 cm; L × W) where ob-
tained from the thickest parts of the fillet (Figure 1). Thickness of
the cubes roughly varied from 1.5 to 2.5 cm due to natural varia-
tions in fillet structure and brine treatment. Skin was subsequently
removed from cubes. Cubes (n = 3/fillet) were equilibrated in a
cold room (3 ± 1 °C) before measurement for about 12 hr (Kang
& Lanier, 2005). A 2-cycle compression test was conducted using
a cylindrical stainless steel probe of 3-inch diameter. Texture Pro-
file Analysis (TPA) was applied using TA-XT2 Texture Analyzer
(Stable Microsystems, Godalming, Surrey, U.K.). The analyzer was
Food Chemistry
of 30% compression with 5 s between cycles. The crosshead moved
at the constant speed of 1 mm/s (Skipnes, Van der Plancken, Van
Loey, & Hendrickx, 2008). The values for hardness, chewiness,
cohesiveness, springiness, adhesiveness, and resilience were deter-
mined from samples. There are no baseline value on TPA of fish
fillets available in literature that indicate mushy, unacceptable, and
acceptable. For that reason, Pacific cod fillets (Gadus macrocephalus)
were utilized as a positive control since the fillet does not suffer
from softening during cooking and is commonly sold for retail and
food service consumption. Non-injected Pacific cod fillets were
procured frozen from a local retail store.
Statistical analysis
A two-factorial complete randomized block design was repli-
cated four times. Independent variables were brine treatment
(n = 10) and cooking protocol (n = 2). Blocking was both the
day of the injection and a replication. Each replication involved
injection of 10 butterfly fillets (1 per fish) with 10 different brine
treatment conditions. In total, this study utilized forty (40) butter-
fly fillets. Following injection, butterfly fillets were split along the
backbone to create two fillets. Fillets were randomly assigned to
a cooking protocol. Overall, there were 80 fillets for texture and
chemical measurement.
Two-way analysis of variance (ANOVA) was conducted and a
multiple comparisons of the mean according to Tukey’s Honestly
Significance Difference Test were analyzed using JMP R
13 (Cary,
NC, 2016). Treatments were significantly different if the P value
was less than 0.05. In addition, PCA of TPA attributes and Pear-
son’s correlation of chemical measures and TPA attributes were
also performed.
Results and Discussions
Fish injection
The injected fillets were targeted for a 10% increase over ini-
tial weight. No differences existed between treatments (Table 1,
P = 0.225). The highest solution uptake was B2PE (11.4 ± 1.4%)
while the lowest was BXG (9.0 ± 2.3%) (Table 1). No difference in
treatments uptake provide a good basis for treatment comparison
for statistical analysis (Lowder, Goad, Lou, Morgan, & DeWitt,
2011).
pH measurement
The pH of treatment brines is reported in Table 1. In general,
increasing the level of protease inhibiting ingredients slightly

Vol. 0, Iss. 0, 2018 r Journal of Food Science 3

 Injection of natural protease inhibitors . . .
Table 2–Correlation between chemical properties (moisture, TEP, and TNEP) and TPA attributes of injected Pacific whiting
(Merluccius productus) fillets, cooked at 90 °C, where bolded values are significant at P < 0.05.
Moisture TEP TNEP Hardness
Moisture 1.00
TEP −0.44 1.00
TNEP 0.17 −0.71 1.00
Hardness −0.49 0.06 0.09
Springiness −0.15 −0.21 0.25
Cohesiveness −0.41 −0.04 0.11
Resilience −0.31 −0.07 0.12
Chewiness −0.47 −0.02 0.14
Adhesiveness −0.08 −0.09 0.04
TEP: total extractable protein; TNEP: total non-extractable protein.
decreased brine (B) pH. For the cooked fillets, pH was not
impacted by brine treatment, cooking protocol or their interac-
Food Chemistry
tion (brine treatment × cooking protocol; Figure 2). A multiple
comparison of the means suggested that the average pH value
of fillets treated by cooking protocol 2 (60/90 °C) were lower
than fillets treated with cooking protocol 1 (90 °C) (Figure 2).
This finding seem to be consistent with Poulter, Ledward,
Godber, Hall, and Rowlands (1985) result which reported that as
temperature increased pH increased. They attributed the increase
in pH to an overall decrease in available acidic groups in the
muscle due to their involvement in inter- and intramuscular
links that occur as the protein denatures. However, in this study
proteolysis was encouraged by using the 60 °C cooking protocol
and seems to have interfered with the ability of acidic groups to
be involved in inter- and intramuscular links during denaturation.
Moisture content
There was no brine treatment × cooking protocol interaction
for moisture content of the cooked fillet (P = 0.90). However, the
main effects of brine treatment and cooking protocol were signif-
icant (P < 0.001). There was no significant differences between
any of the brine treated fillets. Fillets treated by cooking protocol
2 had a significantly lower moisture content compared to those
treated by cooking protocol 1 (Figure 2). A multiple comparison
of means (Figure 2) demonstrated that NI treated with cooking
protocol 2 had significantly lower moisture content than all in-
jected fillets except with W which was also treated with cooking
protocol 2.
It was expected that fillets treated with cooking protocol 2
would have lower moisture. Cooking protocol 2 was a longer heat
treatment which should result in higher water loss due to denat-
uration of myosin, shrinkage or oxidization of myofibril protein
(Liu, Xiong, & Chen, 2009; Ofstad, Kidman, Myklebust, & Her-
mansson, 1993). In addition, cooking protocol 2 was designed
to promote cathepsin L activity which should result in degrada-
tion of protein (An et al., 1994) and thereby reduce the muscle’s
capacity to retain water. Furthermore, moisture content was nega-
tively correlated with TPA measures of hardness, cohesiveness and
chewiness (P < 0.05; Table 2). Fillets with lower moisture values
would therefore display higher TPA measures.
Protein content
There was no brine treatment × cooking protocol interaction
for protein content of the cooked fillet (P = 0.27). However,
the main effects of brine treatment and cooking protocol were
significant (P < 0.05). Protein content of NI and W were higher
than B (Figure 2). Also, fillets treated with cooking protocol 2 had
a higher protein content compare to fillets cooked with cooking
4 Journal of Food Science r Vol. 0, Iss. 0, 2018
Springiness Cohesiveness Resilience Chewiness Adhesiveness
1.00
0.39 1.00
0.64 0.74 1.00
0.73 0.64 0.84 1.00
0.95 0.61 0.79 0.81 1.00
0.12 0.28 0.31 0.55 0.16 1.00
protocol 1 (Figure 2). Further evaluation of data using multiple
comparison of means demonstrated that protein content of NI
treated with cooking protocol 2 was significantly higher than B
treated with cooking protocol 1 (Figure 2). Protein content is
a reflection of two parameters, brine uptake and moisture loss
during cooking. Samples that experienced lower moisture content
had higher protein content and not surprisingly all brine treated
samples were able to retain the highest amount of moisture.
Total extractable protein (TEP) and total non-
extractable protein (TNEP) content
There was no brine treatment × cooking protocol interaction
and neither of the main effects, brine treatment and cooking pro-
tocol, were significant for either TEP or TNEP (P > 0.05). The
highest reported mean values for TEP were NI treated fillets. In
addition, there is a trend of samples treated with cooking proto-
col 2 having higher TEP, except for BXG, B2PE , and B3PE treated
fillets (Figure 2). TEP and TNEP analysis of cooked fillets was
performed to understand whether brine treatment impacted the
extractability of the myofibrillar protein after cooking. TEP is
an indicator of protein solubility and a high TEP value suggests
higher proteolytic activity. TNEP is an indirect measurement of
intermolecular networking and is negatively correlated with TEP.
These results, in conjunction with moisture and protein content
data support the idea that cooking protocol 2 promoted proteo-
lysis in fillets and suggest proteolysis occurred more in NI treated
samples.
Electrophoretic pattern
Figure 3 compares the SDS-PAGE pattern of TEP extracts from
all treatments. The intensity of the myosin heavy chain (MHC)
was greater from fillets treated by cooking protocol 1 with the
possible exception of B3EW . This finding, in conjunction with
moisture, protein and TEP data, confirm proteolysis was enhanced
by cooking protocol 2. In addition, the MHC band from cooked
fillets treated with protease inhibitor ingredients (BPE and BEW )
and cooking protocol 2 were more intense than in fillets with-
out the protease inhibitor ingredients (B, W, and BXG ; Figure 3).
Data suggests that protease inhibitor ingredients prevented myosin
degradation during cooking protocol 2. In the case of B3EW the
MHC band intensity appears similar regardless of cooking proto-
col treatment. This suggests the B3EW treatment performed best
in preventing myosin degradation when challenged with condi-
tions that promoted proteolysis in the fillet. The results from our
study provide evidence that a natural protease inhibitor ingredient
(BEW and BPE ) can be incorporated in a brine at sufficiently high
levels to interfere with proteolysis in the whole fillet. Our result
are in agreement with Kang and Lanier’s (2005) findings which
Injection of natural protease inhibitors . . .

Food Chemistry
Figure 2–Chemical characteristics of Injected Pacific whiting Cooked Fillets on pH (A), Moisture content (B), Total protein content (C), Total extractable
protein (TEP) (D), and Total non-extractable protein (TNEP) (E). a, b, c, d: Means with different letter are significantly different (P < 0.05). Fillet Brine
Treatment: NI = Non-Injected; W = Injected with water; B = Injected with Brine; B1EW = Injected with Brine + 1%EW; B2EW = Injected with Brine +
2%EW; B3EW = Injected with Brine + 3%EW; BXG = Injected with Brine+ 0.1% xanthan gum; B1PE = Injected with BXG + 1%PE; B2PE = Injected
with BXG + 2%PE; B3PE = Injected with BXG + 3%PE.

demonstrated that a purified recombinant cystatin protease could Figure 4). Multiple comparison of means (Figure 4A to F) shows
be incorporated through injection at sufficient levels to interfere highest reported hardness, chewiness and cohesiveness values from
with proteolysis in a whole fillet. NI treated fillets. Taken out of context, it would be easy to con-
clude that since those textural measures for NI were higher than
Texture profile analysis (TPA) brine treated fillets, the NI treated fillets were superior. However,
these same texture attributes were found to be negatively corre-
There was no brine treatment × cooking protocol interac-
lated with moisture (Table 3) which was also reported lowest for
tion for all TPA attributes of the cooked fillet (Figure 4). The
NI (Figure 2). This highlights the difficulty in making direct com-
main effect of brine treatment was significant for all TPA at-
parisons between a non-injected and an injected product because
tributes (P < 0.05) except for adhesiveness (P = 0.745). Moreover,
their composition is so clearly different. Even the addition of a
the main effect of cooking protocol was significant (P < 0.001;

Vol. 0, Iss. 0, 2018 r Journal of Food Science 5

 Injection of natural protease inhibitors . . .

Figure 3–Electrophoretic pattern of Pacific whiting

injected cooked fillet. R = Precision plius proteinTM
all blue standard #1610732 (Bio Rad, Hercules,
CA). Fillet Brine Treatment: NI = Non-Injected;
W = Injected with water; B = Injected with Brine;
B1EW = Injected with Brine + 1%EW;
B2EW = Injected with Brine + 2%EW;
B3EW = Injected with Brine + 3%EW;
BXG = Injected with Brine + 0.1% xanthan gum;
B1PE = Injected with BXG + 1%PE; B2PE = Injected
with BXG + 2%PE; B3PE = Injected with BXG +
3%PE. Abbreviation without aportrophe for fillets
cooked at 90 °C for 20 min. Abbrevitation with
apostrophe ( ) for fillets cooked at 60 °C for 30 min
+ 90 °C for 20 min. MHC = Myosin Heavy Chain.
Food Chemistry

Table 3–Principal component analysis (PCA) loading matrix of Since it is difficult to make conclusions based on any 1 textural
texture profile analysis (TPA) attributes of Injected Pacific whit- measure, Principal Component Analysis (PCA) was conducted to
ing (Merluccius productus) Cooked Fillet
determine potential relationship of TPA measured attributes to
TPA attributes PC1 PC2 samples. The variability of first two principle components (PC)
Hardness 0.84 −0.39 was 79.7% (Figure 5). PC1 (61.8% of the total variance) was highly
Springiness 0.64 0.25 explained by hardness, chewiness, cohesiveness, springiness, and
Cohesiveness 0.90 −0.10 resilience while PC2 (17.9% of total variance) was only explained
Resilience 0.87 0.25 by adhesiveness. Cod, B3EW , B2EW , and NI were positively cor-
Chewiness 0.93 −0.32
Adhesiveness 0.41 0.83
related with PC1 whereas B, BXG , B1PE, B2PE , and B3PE were
negatively correlated. Moreover, B3PE and W were highly related
with PC2 but NI was negatively related. This suggested that Cod
texture attributes, well-defined by hardness, chewiness, cohesive-
ness, springiness, and resilience, were more similar to B3EW and
“water injection” is difficult to interpret as a control treatment. B2EW in comparison to NI, which was not only characterized by
For all texture measures (except adhesiveness) W treated fillets hardness, chewiness, cohesiveness, springiness, and resilience but
appear to outperform B. However, when taken in context with also adhesiveness. This confirms that NI was impacted by prote-
electrophoretic evaluations, it is apparent that myosin integrity was olytic activities.
degraded more in W. The PCA analysis also provides evidence that fillets containing
This further supports the role of cooking protocol 2 in promot- xanthan gum (XG) interfered with protein-protein interaction.
ing texture degradation of fillets. However, a multiple compari- Ponte, Roozen, and Pilnik (1985) reported the use of 0.5% XG in
son of means (Figure 4) was not able to demonstrate there was whiting mince as a cryoprotectant resulted in pastes with very soft
significant differences between fillets treated by the same brine texture, excellent water holding ability and good stability during
but cooked using different protocols. Figure 4, on whole, does frozen storage. Similarly, Ramırez, Barrera, Morales, and Vázquez
seem to suggest that when compared to their respective brine (2002) reported that the addition of 1% XG decreased shear stress
only counterparts (B for BEW and BXG for BPE ) improvements of silver carp surimi gels. This study effectively only incorporated
in texture measurements occurred when a protease inhibitor was 0.01% XG directly into the fish muscle. It was hoped that at very
incorporated. low levels of incorporation the previous negative effects of XG

6 Journal of Food Science r Vol. 0, Iss. 0, 2018

Injection of natural protease inhibitors . . .

Food Chemistry
Figure 4–Texture profile analysis on hardness (A), Cohesiveness (B), Springiness (C), Resilience (D), Chewiness (E), and Adhesiveness (F). Means with
different letter are significantly different (P < 0.05). Fillet Brine Treatment: NI = Non-injected; W = Injected with water; B = Injected with Brine;
B1EW = Injected with Brine + 1%EW; B2EW = Injected with Brine + 2%EW; B3EW = Injected with Brine + 3%EW; BXG = Injected with Brine+ 0.1%
xanthan gum; B1PE = Injected with BXG + 1%PE; B2PE = Injected with BXG + 2%PE; B3PE = Injected with BXG + 3%PE.

observed with fish mince or surimi would not occur. However,
this study clearly demonstrated that even at very low levels of
XG, all XG treated samples were negatively correlated with PC1
attributes. This finding corroborates the ideas of Foegeding and
Ramsey (1987), who suggested that XG surface properties or an
unknown role in the protein aggregation process were responsible
for its interference with protein–protein interactions.
Conclusions
It is possible through the use of brine injection to incorporate
sufficient natural protease inhibitor ingredients into Pacific whit-
ing fillets to reduce the subsequent texture softening that occurs
when they are cooked. At all levels of incorporation, protease
inhibitor ingredients were able to protect integrity of myofibril
structure. Fillets injected with a brine containing 2% and 3% egg
white were the most similar to non-injected Cod in terms of tex-
ture. Although SDS-PAGE and moisture retention demonstrated
that dried potato extract did inhibit protease degradation of fish
myofibril, it was not enough to counteract the detrimental effects
of the suspension aide, xanthan gum, on texture. It is possible that
with the right suspension aide, dried potato extract may be able
to perform similar to dried egg white. Therefore, further study
should be conducted to find an alternative replacement of xan-
than gum as suspension agent in the brine with potato extract. In
addition, sensory evaluations are needed to confirm the consumer
acceptability of fillets treated with protease inhibitor ingredients.

Vol. 0, Iss. 0, 2018 r Journal of Food Science 7

 Injection of natural protease inhibitors . . .
Food Chemistry
Figure 5–Principal component biplot og brine treatment and TPA attribute,
fillets cooked at 90 °C for 20 min (cooking protocol 1). Fillet Brine Treat-
ment: NI = Non-Injected; W = Injected with water; B = Injected with Brine;
B1EW = Injected with Brine + 1%EW; B2EW = Injected with Brine +
2%EW; B3EW = Injected with Brine + 3%EW; BXG = Injected with Brine+
0.1% xanthan gum; B1PE = Injected with BXG + 1%PE; B2PE = Injected
with BXG + 2%PE; B3PE = Injected with BXG + 3%PE.
Acknowledgments
The research was supported by the Hatfield Marine Science
Center through William & Francis McNeil Award, Coastal Ore-
gon Experiment Station, Astoria, OR, and USAID-Fulbright
Agricultural Scholarship for S.D. Harikedua. Authors acknowl-
edge the contribution of Dried Potato Extract by Pacific Blends
Ltd. and Pacific whiting by Jessie’s Ilwaco Fish and Pacific Seafood.
Also, authors would like to thank Ryan Smith BS., Clara Hinter-
meister MS., Carole Le Mestre MS., and Megan Ooi for their
assistance in sample preparation and data collection.
Author Contributions
Silvana D. Harikedua designed the experiment, prepared sam-
ple, collected data, analyzed data, interpreted results, drafted and
edited the manuscript. Christina A. Mireles DeWitt reviewed ex-
perimental design, assisted data analysis and data interpretation,
and edited manuscript. There was no conflict of interest in this
study.
References
Almli, V. L., & Hersleth, M. (2013). Salt replacement and injection salting in smoked salmon
evaluated from descriptive and hedonic sensory perspectives. Aquaculture International, 21(5),
1091–1108.
8 Journal of Food Science r Vol. 0, Iss. 0, 2018
An, H., Peters, M. Y., & Seymour, T. A. (1996). Roles of endogenous enzymes in surimi
gelation. Trends in Food Science & Technology, 7(10), 321–327.
An, H., Weerasinghe, V., Seymour, T. A., & Morrissey, M. T. (1994). Cathepsin degradation of
Pacific whiting surimi proteins. Journal of Food Science, 59(5), 1013–1017.
Latimer JW & Horwitz W (2005). Official methods of analysis (18th ed.). Gaithersburg, Md.:
AOAC Int.
Babbitt, J. 1990. Arrowtooth flounder protease inhibitor injection project. Prepared for Alaska
fisheries foundation Inc., by All Alaskan Seafood and McFarland’s Foods.
Bigelow, W., & Lee, C. M. (2007). Evaluation of various infused cryoprotective ingredients
for their freeze–thaw stabilizing and texture improving properties in frozen red hake muscle.
Journal of Food Science, 72(1), C57–C64.
Dunajski, E. (1980). Texture of fish muscle. Journal of Texture Studies, 10(4), 301–318.
Foegeding, E. A., & Ramsey, S. R. (1987). Rheological and water-holding properties of gelled
meat batters containing iota carrageenan, kappa carrageenan or xanthan gum. Journal of Food
Science, 52(3), 549–553.
Harikedua, S. D., & Mireles DeWitt, C. A. (2017). Preventing soft texture fish fillets
through brine injection technology. Journal of Food Quality, 2017, Article ID 187587,
https://doi.org/10.1155/2017/1875872.
Kang, I., & Lanier, T. C. (2005). Inhibition of protease in intact fish fillets by soaking in
or injection of recombinant soy cystatin or bovine plasma. Journal of Agricultural and Food
Chemistry, 53(25), 9795–9799.
Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of
bacteriophage T4. Nature, 227(5259), 680–685.
Liu, Z., Xiong, Y. L., & Chen, J. (2009). Identification of restricting factors that inhibit swelling
of oxidized myofibrils during brine irrigation. Journal of Agricultural and Food Chemistry, 57(22),
10999–11007.
Lowder, A. C., Goad, C. L., Lou, X., Morgan, J. B., & DeWitt, C. A. M. (2011). Evaluation
of a dehydrated beef protein to replace sodium-based phosphates in injected beef strip loins.
Meat Science, 89(4), 491–499.
Morrissey, M. T., Hartley, P. S., & An, H. (1996). Proteolytic activity in Pacific whiting and
effects of surimi processing. Journal of Aquatic Food Product Technology, 4(4), 5–18.
Morrissey, M. T., Wu, J. W., Lin, D., & An, H. (1993). Protease inhibitor effects on torsion
measurements and autolysis of Pacific whiting surimi. Journal of Food Science, 58(5), 1050–
1054.
Ofstad, R., Kidman, S., Myklebust, R., & Hermansson, A. M. (1993). Liquid holding capacity
and structural changes during heating of fish muscle: Cod (Gadus morhua L.) and salmon
(Salmo salar). Food Structure, 12(2), 163–174.
Pongviratchai, P. (2002). Gelation properties of Alaska pollock surimi with functional ingredients
under ohmic heating (Doctoral dissertation).
Ponte, D. J. B. Da, Roozen, J. P., & Pilnik, W. (1985). Effects of additions on the stability of
frozen stored minced fillets of whiting. II: Various anionic and neutral hydrocolloids. Journal
of Food Quality, 8(2–3), 183–198.
Porter, R. W., Koury, B., & Kudo, G. (1993). Inhibition of protease activity in muscle extracts
and surimi from Pacific whiting, Merluccius productus, and Arrowtooth flounder, Atheresthes
stomias. Marine Fisheries Review, 55(3), 10–15.
Poulter, R. G., Ledward, D. A., Godber, S., Hall, G., & Rowlands, B. (1985). Heat stability of
fish muscle protein. Journal of Food Technology, 20, 213–217.
Ramırez, J. A., Barrera, M., Morales, O. G., & Vázquez, M. (2002). Effect of xanthan and
locust bean gums on the gelling properties of myofibrillar protein. Food Hydrocolloids, 16(1),
11–16.
Rawdkuen, S., Benjakul, S., Visessanguan, W., & Lanier, T. C. (2007). Effect of cysteine pro-
teinase inhibitor containing fraction from chicken plasma on autolysis and gelation of Pacific
whiting surimi. Food Hydrocolloids, 21(7), 1209–1216.
Reppond, K. D., & Babbitt, J. K. (1997). Gel properties of surimi from various fish species as
affected by moisture content. Journal of Food Science, 62(1), 33–36.
Reppond, K. D., & Babbitt, J. K. (1993). Protease inhibitors affect physical proper-
ties of arrowtooth flounder and walleye pollock surimi. Journal of Food Science, 58(1),
96–98.
Sheard, P. R., Nute, G. R., Richardson, R. I., Perry, A., & Taylor, A. A. (1999). Injection of
water and polyphosphate into pork to improve juiciness and tenderness after cooking. Meat
Science, 51(4), 371–376.
Skipnes, D., Van der Plancken, I., Van Loey, A., & Hendrickx, M. E. (2008). Kinetics of heat
denaturation of proteins from farmed Atlantic cod (Gadus morhua). Journal of Food Engineering,
85(1), 51–58.
Wasson, D. H., Reppond, K. D., Babbitt, J. K., & French, J. S. (1993). Effects of additives of
proteolytic and functional properties of arrowtooth flounder surimi. Journal of Aquatic Food
Product Technology, 1(3–4), 147–165.