Citation
The Use of acid and alkali-aided protein solubilization and precipitation methods to produce functional protein ingredients from tilapia

Material Information

Title:
The Use of acid and alkali-aided protein solubilization and precipitation methods to produce functional protein ingredients from tilapia
Creator:
Ingadottir, Bergros ( Dissertant )
Kristinsson, Hordur G. ( Thesis advisor )
Gregory, Jesse F. ( Reviewer )
Johnson, Dwain ( Reviewer )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Language:
English
Physical Description:
xii, 93 p.

Subjects

Subjects / Keywords:
Alkalies ( jstor )
Centrifugation ( jstor )
Gels ( jstor )
Heating ( jstor )
Muscle proteins ( jstor )
pH ( jstor )
Solubility ( jstor )
Surimi ( jstor )
Tilapia ( jstor )
Viscosity ( jstor )
Food Science and Human Nutrition thesis, M.S ( local )
Dissertations, Academic -- UF -- Food Science and Human Nutrition ( local )

Notes

Abstract:
There is great interest in increasing the use of fish muscle proteins, as food source and food ingredient, due to their functional and nutritional properties. Over the years several methods have been developed in order to isolate proteins from fish muscle, many of which cause loss of their functional properties. Therefore, processing methods that focus on retaining valuable functional properties of proteins are preferred. A good example of such a method is production of surimi. Surimi is a washed fish muscle used for manufacturing of imitation seafood products. The demand for surimi products is increasing, whereas the traditional resources for surimi making are limited. Other resources, preferably less expensive, have to be utilized, such as dark muscle species or byproducts. The use of such raw material has not been successful in conventional surimi processing until the development of the acid and alkali-aided processes which enables isolation of functional proteins. The processes are based on solubilization of muscle proteins at low and high pH, separation of the soluble proteins via centrifugation and then precipitation of soluble proteins at their isoelectric pH. The overall objective of this study was to investigate the use of acid and alkali-aided processing to recover functional proteins from tilapia white muscle. The effects of different low and high solubilization and precipitation pHs on solubility, viscosity and protein recovery were determined. In addition, quality of gels prepared from acid and alkali treated proteins was compared to gels prepared using washed tilapia muscle. The use of the acid and alkali aided processes on tilapia muscle proteins in regard to isolation parameters showed that the type of proteins recovered using low and high pH was significantly different whereas the quantity of total protein recovered was not. The ability of the protein isolates to form gels upon heating was compared to lab scale conventional surimi processing. Gels were prepared from acid and alkali treated protein isolates with and without the addition of 2% NaCl (w/w) at neutral pH and compared to gels prepared from washed muscle. Gel quality was determined using torsion, folding, oscillatory testing and water holding capacity. Hardness (shear stress) and elasticity of gels (shear strain) was improved using 2% NaCl (w/w) compared to treatments without salt. Overall, the acid treated proteins exhibited poorer gelling ability compared to alkali treated proteins. Total content of SH groups was measured before and after gelation and S-S bonding did not explain the difference in gel forming ability of different treatments. The results indicate that the alkali aided process can be used to produce high quality protein gels usable for manufacturing of imitation seafood products
Subject:
gelation, protein, rheology, solubility, surimi, tilapia
General Note:
Title from title page of source document.
General Note:
Document formatted into pages; contains 105 pages.
General Note:
Includes vita.
Thesis:
Thesis (M.S.)--University of Florida, 2004.
Bibliography:
Includes bibliographical references.
General Note:
Text (Electronic thesis) in PDF format.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Ingadottir, Bergros. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
12/18/2004
Resource Identifier:
57734984 ( OCLC )

Downloads

This item has the following downloads:


Full Text












THE USE OF ACID AND ALKALI-AIDED PROTEIN SOLUBILIZATION AND
PRECIPITATION METHODS TO PRODUCE FUNCTIONAL PROTEIN
INGREDIENTS FROM TILAPIA
















By

BERGROS INGADOTTIR


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2004

































Copyright 2004


by

Bergros Ingadottir
































This thesis is dedicated to my loving grandparents.















ACKNOWLEDGMENTS

My sincere gratitude goes to my major advisor Dr. Hordur G. Kristinsson for his

guidance, support and encouragement for the past 2 years. His drive and enthusiasm

for the field has been an inspiration. I would also like to thank my committee members

Dr. Jesse F. Gregory III and Dr. Dwain Johnson for their valuable suggestions and

guidance with this research project. It has been an honor to have the opportunity to work

with such talented people.

I would like to thank all my lab mates for making my 2 years here in Florida an

experience I will never forget. I would like to thank my parents for always believing in

me and supporting me through the years, making it possible for me to reach my goals.

Finally I would like to thank my fiance for his endless love and support over the years.
















TABLE OF CONTENTS


page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES .......................................... viii

LIST OF FIGURES ....................................... .. ......... ............................ ix

ABSTRACT .................................................. ................. xi

CHAPTER

1 INTRODUCTION ........................... .. .......... .............................1

2 LITERATURE REVIEW ............................................ ............................4...

Preparation of Fish Protein Ingredients ..................................................4...
R aw M material ............................................................................................ . 4
C conventional Surim i Processing ...................................................... ...............5...
The A cid and A lkali-A ided Processes.............................................. ...............8...
Functional Properties of Muscle proteins............................................................. 12
S o lu b ility ............................................................................................................. 12
V isc o sity .............................................................................................................. 14
G e latio n ........................................................................................................... . 1 5
W after H holding Capacity ................. ......................................................... 18
Protein D enaturation ...... .. ................................................................................ 19
Research Objectives.............................. .......... ........ ............... 20

3 MATERIALS AND METHODS .........................................................................21

R aw M material .................. ..................................................................................... 2 1
Preparation of M uscle H om ogenate ...................................................... ................ 21
P rotein S o lu b ility ........................................................................................................ 2 1
S olu b ility C u rv e ............................................................ .................. ............. 2 1
Solubility before Precipitation ....................................................... ................ 22
Solubility after Isoelectric Precipitation ................ ...................................22
P rotein M easurem ents ......................................... ........................ ................ 23
C alculation of Solubility ....................................... ....................... ................ 23
V isc o sity ..................................................................................................................... 2 4


v












30000

25000

20000

15000

10000


0 with 2% NaCI U without 2% NaCI


I


5000



2.5 2.9 11 11.2
Treatment



Figure 7-4. Storage modulus (G') of protein pastes at 50C after gelation. The data
represents the first replicate. The final G' for gels was obtained by heating the
protein samples from 5 to 800C followed by cooling from 80 to 50C at a rate
of 20C/min using a small strain oscillatory procedure. Results are mean + SD.









8 CON CLU SION S ..................................... ..... 86

L IST O F R EFE R E N C E S .... ........................................................................ ................ 87

B IO G R APH ICAL SK ETCH ..................................................................... ................ 93















LIST OF TABLES


Table page

3-1. Grading system based on five-point system .......................................................31

7-1. Quality of protein gels as assessed by the folding test........................................63















LIST OF FIGURES


Figure page

2-1. C conventional surim i processing ............................................................ ...............6...

2-2. Schem atic of the acid and alkali processes ........................................... ...............9...

3-1. Soluble protein before and after centrifugation................................... ................ 25

3-2. Schematic showing protein isolation from tilapia white muscle...........................27

3-3. The modified Brookfield DV-II viscometer and the milling equipment...............31

4-1. Protein solubility (%) of tilapia homogenate at pH 1.5 to pH 12.0.......................35

4-2. Viscosity (Pa*s) of tilapia homogenate at pH 1.5 to pH 12.0...............................37

4-3. Solubility and viscosity curves demonstrated on the same graph.........................38

5-1. Effects of different homogenization times on protein solubility (%).....................40

5-2. Viscosity (Pa*s) of tilapia homogenate at low solubilization pH .........................41

5-3. Viscosity (Pa*s) for tilapia homogenate at high solubilization pH.......................42

5-4. Theoretical and actual % protein recovery after 1 st centrifugation ......................45

6-1. Soluble protein (%) in supernatant after 2nd centrifugation...............................47

6-2. Viscosity, (Pa*s) of tilapia proteins at different precipitation pH.........................49

6-3. Shows the same as Figure 6-2, but in a different arrangement. ..............................50

6-4. Theoretical and actual protein recovery (%) after 2nd centrifugation ..................51

6-5. Total theoretical and actual protein recovery (%) in final protein isolate ............53

6-6. Electrophoresis for w ashed tilapia m uscle ......................................... ................ 54

6-7. Electrophoresis for precipitated proteins without solubilization...........................54

6-8. Electrophoresis for pH 2.5 treated proteins......................................... ................ 55









6-9. Electrophoresis for pH 2.9 treated proteins......................................... ................ 55

6-10. Electrophoresis for pH 11.0 treated proteins....................................... ................ 56

6-11. Electrophoresis for pH 11.2 treated proteins....................................... ................ 56

7-1. Shear stress values (kPa) of tilapia gels using torsion ........................ ................ 58

7-2. Shear strain values of tilapia gels using torsion. ................................. ................ 60

7-3. Summary of storage modulus (G') of protein pastes at 50C before gelation...........64

7-4. Summary of storage modulus (G') of protein pastes at 50C after gelation..............66

7-5. Summary of storage modulus (G') of protein pastes at 5C after gelation..............67

7-6. Typical rheogram for washed tilapia muscle with 2% NaCl (w/w). .....................69

7-7. Typical rheograms for proteins treated at pH 2.5 with 2% NaCl (w/w) ...............70

7-8. Typical rheograms for proteins treated at pH 2.9 with 2% NaCl (w/w). ...............71

7-9. Typical rheograms for proteins treated at pH 11.0 with 2% NaCl (w/w)..............72

7-10. Typical rheograms for proteins treated at pH 11.2 with 2% NaCl (w/w) .............73

7-11. Typical rheogram for washed tilapia muscle without 2% NaCl (w/w)................. 75

7-12. Typical rheograms for proteins treated at pH 2.5 without 2% NaCl (w/w). ............76

7-13. Typical rheograms for proteins treated at pH 2.9 without 2% NaCl (w/w) .............77

7-14. Typical rheograms for proteins treated at pH 11.0 without 2% NaCl (w/w) ...........78

7-15. Typical rheograms for proteins treated at pH 11.2 without 2% NaCl (w/w) ...........79

7-16. Water loss (%) of tilapia muscle protein pastes on cooking ...............................81

7-17. Water loss (%) of tilapia muscle protein gels under pressure ..............................82

7-18. Total SH group content in tilapia muscle protein paste before cooking ..................84

7-19. Total SH group content in tilapia muscle protein gel after cooking ........................ 84















Abstract of Thesis Presented to the Graduate School of the University of Florida
in Partial Fulfillment of the Requirements for the
Degree of Master of Science

THE USE OF ACID AND ALKALI-AIDED PROTEIN SOLUBILIZATION AND
PRECIPITATION METHODS TO PRODUCE FUNCTIONAL PROTEIN
INGREDIENTS FROM TILAPIA

By

Bergros Ingadottir

December 2004

Chair: Hordur G. Kristinsson
Major Department: Food Science and Human Nutrition

There is great interest in increasing the use of fish muscle proteins, as food source

and food ingredient, due to their functional and nutritional properties. Over the years

several methods have been developed in order to isolate proteins from fish muscle, many

of which cause loss of their functional properties. Therefore, processing methods that

focus on retaining valuable functional properties of proteins are preferred. A good

example of such a method is production of surimi. Surimi is a washed fish muscle used

for manufacturing of imitation seafood products. The demand for surimi products is

increasing, whereas the traditional resources for surimi making are limited. Other

resources, preferably less expensive, have to be utilized, such as dark muscle species or

byproducts. The use of such raw material has not been successful in conventional surimi

processing until the development of the acid and alkali-aided processes which enables

isolation of functional proteins. The processes are based on solubilization of muscle






xii


proteins at low and high pH, separation of the soluble proteins via centrifugation and then

precipitation of soluble proteins at their isoelectric pH.

The overall objective of this study was to investigate the use of acid and alkali-

aided processing to recover functional proteins from tilapia white muscle. The effects of

different low and high solubilization and precipitation pHs on solubility, viscosity and

protein recovery were determined. In addition, quality of gels prepared from acid and

alkali treated proteins was compared to gels prepared using washed tilapia muscle.

The use of the acid and alkali aided processes on tilapia muscle proteins in regard

to isolation parameters showed that the type of proteins recovered using low and high pH

was significantly different whereas the quantity of total protein recovered was not.

The ability of the protein isolates to form gels upon heating was compared to lab

scale conventional surimi processing. Gels were prepared from acid and alkali treated

protein isolates with and without the addition of 2% NaCl (w/w) at neutral pH and

compared to gels prepared from washed muscle. Gel quality was determined using

torsion, folding, oscillatory testing and water holding capacity. Hardness (shear stress)

and elasticity of gels (shear strain) was improved using 2% NaCl (w/w) compared to

treatments without salt. Overall, the acid treated proteins exhibited poorer gelling ability

compared to alkali treated proteins. Total content of SH groups was measured before and

after gelation and S-S bonding did not explain the difference in gel forming ability of

different treatments. The results indicate that the alkali aided process can be used to

produce high quality protein gels usable for manufacturing of imitation seafood products














CHAPTER 1
INTRODUCTION

There is a great interest in increasing the use of muscle proteins, as a food source

and food ingredient, due to their functional and nutritional properties (1). Muscle proteins

from fish are nutritive, easily digested and exhibit good functionality (2;3) which makes

them desirable for various food applications. However, the use of fish proteins as a food

or food ingredient, has been limited due to several reasons, such as, rapid bacterial

spoilage, lipid oxidation, protein oxidation, low stability compared to mammalian and

vegetable proteins and loss of functionality during processing (4). Over the years several

methods have been developed in order to isolate proteins from fish muscle, many of

which cause a loss of their functional properties. Many of them have been relatively

harsh where conditions such as combination of very low or high pH in the presence of

organic solvents and high heat have been used. These methods were highly unsuccessful

since functionality and nutritional quality of the products were negatively affected (5).

Processing methods that focus on retaining valuable functional properties of the proteins

are preferred and have received increased attention. A good example of such a method is

surimi processing which involves washing fish muscle and adding cryoprotectants prior

to freezing to stabilize the proteins. Surimi is then used for manufacturing of imitation

seafood products by heating. The popularity of surimi based products is gradually

increasing both in the US and Europe. The surimi industry is largely based on the

utilization of Alaska Pollack, which is now under pressure from over fishing and

therefore other species have to be found (4).









Processing methods which could employ inexpensive raw materials to make a

quality surimi would be highly desirable both to reduce the pressure of over fishing of

currently used species and to reduce the cost of production. Potential raw materials could

be; industrial fisheries presently exploited, seafood processing by-products and by-catch

or unexploited/under-exploited stocks. Employing conventional surimi processing on

these raw materials has been met with numerous problems (6; 7). To address the problem

of utilization of raw materials unusable for surimi processing, two novel processes were

developed originally at the University of Massachusetts. These processes, which both

work by the same principle, involve acid or alkaline solubilization and isoelectric

precipitation of muscle proteins to give a highly functional and stable protein isolate from

low value underutilized species and by-products (1; 7). The new process has been shown

to work well for various cold water species such as cod (8), herring (9), and Pacific

whiting (10) but currently there is little data available for the potential of using these

processes to produce functional proteins from warm water species such as tilapia.

Tilapia aquaculture is rapidly growing worldwide (11), generating large amounts of

by-products (and primary products) which could be utilized for its protein content,

provided the proper process is used. In addition, consumption of tilapia is increasing

both in the US and globally. Conventional surimi processing from tilapia has been

somewhat successful although the yields are fairly low. To the best of our knowledge the

acid and alkali-aided processes have not been applied on tilapia muscle materials.

Therefore, tilapia is a species of great interest to investigate. To reach the goal of full

utilization of tilapia it is essential to investigate the use of the newly developed acid and

alkali-aided processes on whole muscle before by-products can be utilized. The results






3


from this research are expected to give important information on the production of a high

quality protein isolate from aquacultured tilapia.














CHAPTER 2
LITERATURE REVIEW

Preparation of Fish Protein Ingredients

Raw Material

Quality and characteristics of a protein ingredient or a finished product are highly

dependent on the source of muscle protein and the processing procedures applied (5;12).

Species with white flesh and low fat content are considered most suitable for surimi

manufacturing, for example Alaska pollock, Pacific whiting, hoki, southern blue whiting,

northern blue whiting and yellow croaker (12). The use of fish species with higher

content of dark/red muscle and fat has met some complications such as low grade protein

gels, color problems and lipid oxidation (12). The reason for poorer gelling ability of

species with darker muscle has been related to its characteristics; higher proteolytic

activity, lower muscle pH which can lead to more rapid protein denaturation, higher

concentration of sarcoplasmic proteins, higher lipid content and high concentration of

heme proteins in the muscle. All these factors have been reported as problems when

producing high quality surimi from materials containing large amounts of dark muscle

(6). Surimi manufacturers have resorted to several methods to alleviate this problem (e.g.,

by removing the dark muscle before surimi processing). The disadvantage of doing this is

decreased protein recovery. The use of the acid and alkali aided processes has made it

possible to produce good quality surimi from dark muscle species.

A major problem facing any protein extraction and recovery process is proteolysis

by endogenous proteases. Postmortem fish muscle is very prone to proteolysis and the









problem varies greatly with species and season. The effect of proteolytic activity on

muscle protein gels has detrimental effect on their quality due to rapid degradation of

myofibrillar proteins, in particular myosin. Proteolytic activity and types of proteases

vary among species. Yongsawatdigul et al. (13) reported that serine type protease(s)

were most probably involved in proteolysis of tilapia surimi resulting in gel weakening.

Gel weakening due to proteolysis has also been observed in Arrowtooth flounder,

threadfin bream, Atlantic menhaden and lizard fish (12).

Proteolysis can be a significant problem during the acid and alkali-aided processes.

The former process is thought to be more problematic since low pH's can activate

contaminating gut enzymes (pepsin) and also certain lysosomal muscle enzymes (14).

Undeland et al. (9) found that proteolysis occurred when herring proteins were held at

pH 2.7, while no proteolysis occurred at pH 10.8. Choi and Park (10) also reported

proteolytic degradation of muscle proteins during acid-aided processing of Pacific

whiting.

Conventional Surimi Processing

Surimi was originated in Japan where it has been a traditional food source for

centuries. It is a minced fish muscle washed with water and used as an ingredient for

imitation seafood products, primarily crab substitutes. For many years the industry was

dependent on supply and availability of fresh fish. The discovery of adding

cryoprotectants to surimi in order to prevent protein denaturation during freezing,

revolutionized the industry (12) which was no longer dependent on fluctuations in supply

of fresh fish.











Fish



Mince fish meat



Mechanically
debone


Wash with water



Refine



Dewater



Add cryoprotectants, package,
freeze = SURIMI J



Figure 2-1. Conventional surimi processing. The fish (skinless fillets, fillets with
the skin on, or headed and gutted fish) is minced, mechanically
deboned, washed with water, reined, dewatered, cryoprotectants added ,the
product packaged, and finally frozen in blocks until used. This product is
termed "surimi" and is used in finished products, characterized as gels,
formed by using heat.

In conventional surimi processing as shown in Figure 2-1, the raw material is

minced and then mechanically deboned (bones, skins and scales are removed by

pressing soft tissue through small holes in a screen). Washing with water concentrates the

myofibrillar proteins, removes components that can have negative effects on gelation and

compounds that can cause flavor, odor, stability and color problems. The amount of

water and number of washes depends on the raw material, the desired final product and

water availability. The temperature should be maintained low enough to prevent protein









denaturation which varies according to species (15;16). In the refining step, remaining

bone pieces, skin, scales and connective tissue are removed. Dewatering is necessary

because during the process water is absorbed (approximately 100% increase) due to

repulsion of negatively charged proteins in the washed mince (pH ~ 6.4 to 7.0). The

water minimizes the repulsion by separating the proteins and the addition of salt (0.1 to

0.3%), reduces the repulsive forces by shielding negative charges which allows the

proteins to approach one another. Thus, expelling water and reducing the tendency of the

tissue to absorb water. Finally cryoprotectants, usually sucrose (4%), sorbitol (4%) and

sodium tripolyphosphate (0.2 to 0.3%) are added in order to protect the proteins from

denaturation and aggregation during freezing which would result in reduced gelation

ability of the proteins (12). Prior to freezing, proteolytic enzyme inhibitors are

sometimes added along with cryoprotectants to prevent proteolytic degradation of

proteins during heating. For example is the use of Pacific whiting for surimi

manufacturing based on addition of enzyme inhibitors and application of fast heating rate

to minimize proteolytic degradation of muscle proteins (12;17).

Generally, high quality surimi is produced from white fleshed fish such as Alaska

pollock but to meet the increasing demand new resources have to be found. Utilization of

dark fleshed underutilized species has often led to products with poor gelation properties.

Recent studies show however that the use of warm water species like tilapia have resulted

in better gel quality compared to more common surimi species. For example, Klesk and

coworkers (17) compared the gel quality of tropical tilapia surimi to Alaska pollock and

Pacific whiting and found that it was comparable to Alaska pollock and better than

Pacific whiting (without enzyme inhibition) at when heated at 900C for 15 min.









The Acid and Alkali-Aided Processes

The acid and alkali processes, originally developed by Hultin and coworkers are

able to overcome some of the problems that are involved using dark muscle, small

species or by-products for surimi production. The processes involve using acid or

alkaline solubilization of muscle proteins, followed by isoelectric precipitation of soluble

proteins to give highly functional and stable protein isolate used to produce quality surimi

(]; 7).

As outlined in Figure 2-2, the production is carried out by subjecting a diluted slurry

of homogenized muscle tissue to either a very acidic (2.0 to 3.5) or alkaline (10.5 to 11.5)

pH to solubilize the majority of muscle proteins via electrostatic repulsion. As a

consequence of disruption of the muscle cell and solubilization of the proteins the

viscosity of the protein solution is drastically lowered. The lowering of viscosity allows

for separation of insoluble material, such as, bones, skin, connective tissue, cellular

membranes and neutral storage lipids from the soluble proteins by centrifugation. The

soluble proteins are collected after centrifugation and recovered by isoelectric

precipitation (by adjusting the pH to 5.2 to 6.0) and then collected by centrifugation. The

sediment (protein isolate) is kept and the supernatant discarded. Cryoprotectants are

added to the protein isolate prior to freezing to protect the proteins from denaturation.

Protein gels made from protein isolates recovered with the new process from

several species have shown to have equal and sometimes significantly better gelation

properties than those produced using conventional surimi processing techniques

(6;9;10;18). The process has also shown to improve other functional properties. The

process has given excellent results for some cold water species as well as temperate and

warm water species and is now in the route of commercialization for North Atlantic and









Pacific species as well as species in the US Southeast. A study done by Kristinsson and

Hultin (8) on cod muscle proteins showed that the alkali treatment improved functional

properties emulsificationn and gelation) of cod myosin and myofibrillar proteins. This

improvement was found to be directly linked to a unique partially unfolded structure of

cod myosin after alkali-treatment.


Figure 2-2. Schematic diagram of the acid and alkaline process used in the production of
functional protein isolates.

A different response to the acid and alkali process can be expected for warm water

species compared to cold water species in part because their proteins are more heat stable

due to their living environment. A study performed on threadfin bream actomyosin

indicated that aggregation underwent at higher temperatures than was seen for cold water









species such as cod and pacific whiting (19). A number of studies at the University of

Florida have shown that several warm water species have a great potential to be utilized

via these two processes. Kristinsson and Demir (18) compared the acid and alkali-aided

processes to surimi on channel catfish, Spanish mackerel, croaker and mullet and

demonstrated that the two processes led to significantly higher protein recoveries and

lipid reduction than lab scale conventional surimi processing but also that the alkali

process resulted in significantly improved gelling ability, color and oxidative stability

(due to removal of heme proteins) compared to the acid treatment and surimi process.

Theodore and Kristinsson (20) showed that using the alkali-aided process on catfish led

to protein gels of greater strength when compared to using acid-aided process and

conventional surimi process over a wide range of pH (pH 6-8) and ionic strengths (0-600

mM NaC1). Results by Davenport and Kristinsson (21) indicate that there are major

changes in myosin (from channel catfish) that contribute to the improved gel strength

after alkali processing. In addition Kristinsson and Crynen (22) demonstrated that both

myofibrillar and sarcoplasmic proteins of muscle (from channel catfish) are positively

affected in terms of gelation ability after alkali-treatment, but negatively affected after

acid-treatment. The molecular phenomena responsible for these differences are under

investigation.

Studies on tilapia using conventional processing have shown promising results

(17;23-26) whereas few studies have been conducted with tilapia muscle proteins using

the acid and alkali processes to the best of the researchers' knowledge. Preliminary

significantly greater gel forming ability of the proteins vs. untreated muscle proteins.

Treating the tilapia muscle proteins to a low pH (2.5) followed by readjustment to pH 7









resulted in significantly poorer gel forming abilities compared to untreated proteins. The

reasons behind these major differences are unknown and are important to determine to be

able to successfully produce protein isolates from tilapia muscle and its by-products.

The acid and alkali-aided processes have several advantages over the conventional

surimi process in regard to isolating functional and quality protein from fish muscle

* In the acid and alkali-aided processes protein recoveries are normally significantly
higher due to recovery of both myofibrillar and a sizable amount of sarcoplasmic
proteins. Recoveries using conventional techniques are usually lower due to loss of
sarcoplasmic proteins (- 30% of total protein) during washing steps.

* The acid and alkali-aided processes are simple and require less labor than surimi
processes. For example, liquefaction of the material of interest makes it easier to
move it around a processing plant compared to a more solid material such as in
surimi processing.

* Whole fish with skin and bones and fatty fish can be utilized in the acid and alkali-
aided processes because proteins are selectively separated and recovered from
undesirable muscle components. This is not feasible using typical surimi
processing without negatively affecting the recoveries and quality (5).

* Lipids and cellular membranes can be effectively removed in the acid and alkali-
aided processes. This significantly increases the color and oxidative stability of the
final product compared to proteins produced using surimi. Microorganisms are
also effectively precipitated out during centrifugation.

* Removal of heme proteins are more effectively obtained with the alkali treatment
compared to surimi processing resulting in a product that is whiter and more stable
to lipid oxidation. Heme proteins are also protected from denaturation and
autoxidation during high pH treatment. Acid treatment leads to heme protein
denaturation and co-precipitation with muscle proteins, leading to color and
oxidative problems.

* Functional properties are either retained, decreased (in few cases for the acid
process) and often significantly improved (alkali process) using the process. The
effect on functionality is highly dependent on species and type of functional
property.

* Problems related to foaming and emulsification of lipids have been encountered in
commercialization of the acid and alkali-process and have led to reduced protein
recoveries.









* The use of high pH in combination of high heat can lead to formation of
lysinoalanine. Acid hydrolysis can cause razemization of amino acids. This is
normally not a problem in the acid and alkali-processing due to rapid process time
and very low process temperature.

Functional Properties of Muscle proteins

Quality and stability of a final product are affected by functional properties of

proteins (27). The most important functional properties when producing a functional

muscle protein isolate for use in food are solubility, viscosity, water holding capacity and

gelation (1). These properties are discussed in more detail below.

Solubility

Solubility of myofibrillar proteins is believed to play an important role in gelation

and water holding capacity of muscle proteins (28). Therefore, solubility of muscle

proteins has been the subject of much research.

A change in solubility can be obtained in various ways, for example, by varying

ionic strength, ion types, pH, and/or temperature and thus affecting the hydrophobic

and/or ionic nature of the proteins. For a long time it was a general belief that

solubilization of myofibrillar proteins in high salt concentration (0.3 to 0.6 M) was required

to form good fish gels. Studies have shown that this is not necessarily the case,

Stefansson and Hultin (29) showed, that cod muscle myofibrillar proteins were soluble if

the ionic strength was sufficiently reduced (< 0.3 mM) at both neutral and acidic pH.

They suggested that the negative charge of myofibrillar proteins at neutral pH and/or in

water or solutions of very low ionic strength, repulsive forces from negatively charged

side chains are enough to drive the individual protein molecules apart when sufficient

water is available (29).









Changes in protein solubility can also be obtained by varying the pH of the

solution, as is done in the acid and alkali-aided processes. By changing the pH, the

protein acquires a net negative or net positive charge where hydration of the charged

residues and electrostatic repulsion results in an increase in solubility (30). Conditions

like lowering the pH near to the isoelectric point reduces the repulsive forces and allows

the proteins to associate. Therefore, many proteins exhibit minimum solubility at the

isoelectric point (pI) where the lack of electrostatic repulsion promotes aggregation

between protein molecules. Due to aggregation of proteins at these conditions they can

be removed from solution by appropriate centrifugal force.

In order to find the most appropriate pH to solubilize and recover proteins from a

protein solution a solubility curve can be constructed (protein concentration vs. pH). For

example, a study done by Choi and Park (10) showed that solubility of Pacific whiting

proteins was lowest at pH 5.0 which indicated a suitable pH to precipitate the proteins.

Maximum solubility was observed at pH 1.5 on the acidic side and at pH 11.0 on the

alkali side. The solubility curve of channel catfish showed that most solubility was

observed at pH 2.5 and 11 and least at pH 5.5, suggesting that the former two pH values

would be suitable to solubilize the proteins and the latter pH to precipitate them from

solution (18).

The ionic strength of a solution has a dramatic impact on the pH-dependent

solubility profile of muscle proteins. For example, Dagher et al. (31) showed that

solubility of a washed cod muscle mince increased dramatically between 8.9 and 9.2 at an

ionic strength of 0.001 M. However the effects of pH at high salt concentration were not

as dependent as when the salt concentration was low. A recent study by DeWitt et al.









(32) on using the acid-aided process to solubilize and recover proteins from beef heart

showed that increasing the ionic strength at low pH led to a decrease in solubility and

thus their extractability. This reduction in solubility stems from screening of electrostatic

repulsive forces between the proteins (33).

Viscosity

The viscosity of muscle protein homogenates at low and high pH in the acid and

alkali process is important since low viscosity is necessary to separate insoluble material

from soluble proteins via centrifugation. Viscosity of a protein solution is believed to be

affected by factors like protein concentration, pH, salt, and raw material processing

which in turn can affect size, shape, flexibility and hydration of the proteins (33).

Viscosity of protein solutions usually increases exponentially with protein concentration.

This is due to increased interaction between the hydrated protein molecules. When two

hydrated molecules are in close proximity, short range repulsive interaction forms where

the strength depends on the degree of hydration (more hydration => stronger repulsion

and longer range) (34). An increase in ionic strength usually decreases viscosity by

affecting the hydration capacity of the proteins (33).

The two most important factors affecting viscosity of a protein solution according

to Damodaran (33) are hydrodynamic size and shape of the protein molecules. Partial

denaturation and/or heat induced polymerization, increases hydrodynamic size of proteins

and thus increases viscosity. Most macromolecular solutions do not exhibit Newtonian

behavior instead viscosity is decreased with increasing shear rate. This is called

pseudoplastic behavior or shear thinning. This is due to the fact that the proteins align

themselves in the direction of flow and weakly bound dimers and oligomers are

dissociated into monomers (30).









Gelation

One of the most important functional properties of muscle proteins is their ability to

form gels upon heating. A gel is an intermediate stage between a solid and a liquid where

polymers (e.g., proteins) form a three-dimensional network that is capable of holding water

and other low molecular compounds (30). Characteristics of protein gels are determined

by the type and number of protein-protein interactions, aggregation and arrangement of

unfolded proteins which are in turn affected by pH, ionic strength, protein concentration,

and heating and cooling rates (35).

The pH and ionic strength are by far the most important determinants of strength

and quality of muscle protein gels. Optimum pH for the gelation of muscle proteins has

often been reported to be between 5.5 and 7.0, and depended on animal species, protein

concentration, salt concentration, and instrument used to analyze gel strength. The use of

salt (0.3 to 0.6 M) has been regarded a prerequisite in order for myofibrillar proteins to form

good gels. Recently, this general belief was challenged as muscle proteins have been

found to form excellent gels in the absence of salt provided that electrostatic repulsion is

sufficiently high to create a strong osmotic pressure in the gel matrix (18;36).

The formation and characteristics of muscle gels are also highly dependent on the

heating procedures. Different species respond very differently to the same heat

treatment, which makes it important to determine the optimal heating and cooling scheme

for good gel formation. As an example, gel strength can be improved for some gels by

holding the protein paste at a temperature below 500C before the final heat treatment at

ca. 80 to 900C, a process called setting. However, for some species this results in loss of

gel quality if the temperature is held at 50 to 600C for too long due to proteolytic

degradation of muscle proteins (5). In a study performed by Klesk and coworkers (37),









tilapia formed the best gel when no setting was applied when compared to pollock and

pacific whiting. Whereas, using temperature of 600C during setting was found to reduce

strain and stress values of the gels and indicated increased protease activity leading to the

degradation of myosin heavy chain.

According to Lanier et al. (38) the setting of fish muscle proteins at temperatures

below that at which rapid aggregation occurs (ca. 400C) may be viewed as a process

where partially denatured proteins begin to interact non-covalently to form a fine elastic

network. Setting below 400 C would allow for slow ordering of the molecules and give

gels with greater firmness and cohesiveness. Hermansson (39) also reported that

denaturation of proteins prior to aggregation results in a finer gel structure, exhibiting

greater elasticity than if random aggregation occurs prior to denaturation. Park et al.

concluded that optimum setting temperature is highly related to habitat temperatures (40).

The proteins that have been found to be primarily responsible for gel formation and

gel strength are the myofibrillar proteins most notably myosin and to some extent the

complex of actin and myosin (actomyosin) (41). Grabowska and Sikorski (42) did a gel

study on myofibrillar, sarcoplasmic and stroma fraction and the sarcoplasmic fraction

showed no gelling ability. It has been a general believe that sarcoplasmic proteins have a

negative effect on gelation. However recent studies point to that that sarcoplasmic

proteins may actually improve gelation in some cases. Sarcoplasmic proteins are

recovered in the acid and alkali-aided processes but not in surimi processing. Recently

Kristinsson and Crynen (22) reported that a very complex interaction occurs between

sarcoplasmic and myofibrillar proteins when mixed in different ratios under different pH

and ionic strength conditions. In many cases the sarcoplasmic proteins had a positive









impact on the gel forming ability of myofibrillar proteins, especially when both had been

treated at a high pH (11.0), as done in the alkali-aided process.

Gels are complex and their properties and the mechanism of their formation has

been the subject of many studies. There are several ways available to study the properties

of a gel and the mechanism of their formation. Rheology deals with flow and

deformation of matter which is induced by applied force (43). There are mainly two

quantities that rheology is concerned with; stress, which is the force applied divided by

an area of matter that it is applied to and strain which is the deformation of a matter

induced by the stress. One type of deformation is shear, which is when matter changes

shape without changing volume (44). An important property of a gel is deformability,

which is how a gel responds to strain without breaking. There are two main theological

studies available in order to obtain information on properties of gels and the mechanism

of formation, small strain testing and large strain testing. The combination of these

measurements can give valuable information on acceptability of a gel.

Small strain testing is a study where a sample is deformed without breaking the

structure. Heating and cooling of a gel, using low frequency and small strain oscillatory

experiments is one of the best suited methods to follow changes in physical properties

relating molecular properties of a gel (44;45). The primary parameters of interest in

small strain testing are; the storage modulus (G') which describes the elastic component

of a protein gel, the loss modulus (G") a measurement of the viscous attribute of the gel

and the phase angle (6), where for a perfectly elastic material stress and strain are in

phase or 6 = 0 and for a perfectly viscous material 6 = 90.









Large strain testing is when a sample is deformed until the structure is permanently

broken. Large strain testing estimates fundamental properties of a gel and have shown to

correlate with sensory texture. An example of large strain testing is puncture test and

torsion (twisting) which was used in the experiments performed for this thesis (44). The

torsion test is a typical method to test the hardness (shear stress) and elasticity (shear

strain) of surimi gels. The advantages using the torsion over a compression test e.g. is that

pure shear is applied, the volume is unchanged even if the sample is compressable and

the shape is maintained during testing. To maintain accuracy of the measurements the

cross-sectional diameter needs to have a constant length. It was observed that shear

stress was extremely sensitive (40 to 70 kPa) to changes in gel diameter (0.90 to 1.10 cm)

(40).

Water Holding Capacity

Water holding capacity (WHC) is an important factor in muscle protein gels as it

not only affects the economics of their production but also their quality. Water holding

capacity can be defined as the ability of a matrix (e.g., a protein gel to retain water against

gravitational force) (30). The level of water retained in a gel is affected by much the same

factors that affect the formation of a good gel matrix; pH and ionic strength (i.e., salt).

Feng and Hultin (36) reported that gels with an evenly distributed gel structure showed

improved WHC. A poor gel matrix on the other hand can lead to synerisis (i.e., a

discharge of water from the gel). Gels prepared in the pH range of 6.4 to 7.4 had

increasingly higher WHC as pH increased (46). An increase in WHC was shown to

correlate well with increased negative charge on the muscle proteins. A comparison

between the WHC of protein isolates prepared from the acid and alkali-aided processes









and those prepared using surimi processing has not been conducted, to the best of the

writers knowledge.

Protein Denaturation

It is a commonly held view that denaturing fish muscle proteins has a detrimental

impact on their functional properties. Denaturation often results in negative changes in

protein functionality such as enzyme activity or loss of functional properties. In other

cases, denaturation of the proteins can result in improvement of functional properties

such as foaming and emulsification of egg albumin (47) or improved gelation. Loss of

protein functionality has been correlated to loss in ATPase activity, a common indicator

of muscle protein denaturation (2;48). Interestingly, ATPase activity is essentially lost in

the acid and alkali processing where the proteins are partially denatured at low and high

pH and then only partially refolded when the pH is readjusted (49). The fact that acid

and alkali-treated proteins often have significantly improved functionality goes against

common believe. Kristinsson and Hultin (49) have shown that the unique structure the

proteins possess after pH-treatment is responsible for improved functional properties such

as gelation, emulsification and solubility. For example, the partially unfolded/folded

structure is more flexible and is able to form better protein networks on heating (gelation)

and is able to adsorb more readily to interfaces and lower interfacial tension

emulsificationn). Results indicate that acid treatment has a different effect on the

structure of the muscle proteins compared to the alkali treatment, also in a species

dependent manner (21;49). It is of importance to understand what specific changes occur

with the structure of the proteins during the process to optimize the functionality of these

proteins.






20


Research Objectives

The overall objective of this study was to investigate use of acid and alkali-aided

processing to recover functional proteins from tilapia white muscle. The effects of low

and high solubilization pH and precipitation pH on solubility, viscosity and protein

recovery was determined. In addition, quality of gels prepared from acid (pH 2.5 and

2.9) and alkali (pH 11.0 and 11.2) treated proteins was evaluated and compared to gels

prepared using washed tilapia muscle.














CHAPTER 3
MATERIALS AND METHODS

Raw Material

Tilapia fillets were obtained fresh from a local supplier (Rain Forest Aquaculture,

Sunrise, FL), transported on ice to the laboratory and stored in a cold room at 40C until

processed (within 24 hours). Red muscle was manually excised from the white muscle

and discarded. The remaining white muscle was ground using a Scoville grinder

(Hamilton Beach, Washington, NC) with 6 mm holes. All preparation of raw material

and samples was performed in a cold room at 40C or on ice when applicable to maintain

temperature below 50C.

Preparation of Muscle Homogenate

Ground muscle was mixed with cold (40C) DI water (1:9, w:w) and homogenized

for 60 sec (2x30 sec) with a Waring blender (Waring Products Division, New Hartford,

CT) at 40% electrical output. The homogenate was carefully poured into a plastic beaker

on ice and was adjusted to the appropriate pH by adding 2 M HCI or 2 M NaOH followed

by stirring with a plastic spatula.

Protein Solubility

Solubility Curve

A protein solubility curve was constructed from pH 1.5 to 12.0 in 0.5 intervals.

Two sets of homogenates were prepared from the same raw material. The first set was

adjusted from the native muscle pH (6.5-6.9) to pH 1.5 with 2 M HCI and the second set

was adjusted from the native muscle pH to pH 12.0 using 2 M NaOH. At each pH









approximately 30 g of homogenate was accurately weighed into centrifuge tubes (50 mL)

and centrifuged at 10,000 G for 20 min in a Sorvall RC-5B using a SS-34 rotor (29),

separating insoluble material from the soluble proteins.

Total protein was determined by taking a 2 g sample at each pH before

centrifugation and diluting it 10 times with cold DI water at pH 11.0 to aid in

solubilization of the proteins. The samples were homogenized with a hand held Tissue

Tearor (Biospec Products, Inc, Bartlesville, OK) on speed 7 for 20 sec and then analyzed

for protein content using the Biuret Method (see below). Soluble protein was determined

by taking a 1 mL sample from the supernatant after centrifugation and diluting it 5-fold

with cold DI pH 11.0 when applicable (pH 5 to 6 were undiluted).

Solubility Before Precipitation

From the previously constructed solubility curve, protein solubility at low pH (2.3 to

2.9) and high pH (10.8 to 11.4) was studied in more detail. Effects of different

homogenization times (60, 90, and 120 sec) using these low and high pHs were

investigated. Total protein before centrifugation and soluble protein after centrifugation

were determined as previously described.

Solubility After Isoelectric Precipitation

Protein solubility after isoelectric precipitation, indicating loss of protein after 2nd

centrifugation was determined using four solubilization pHs 2.5, 2.9, 11.0, and 11.2; and

four precipitation pHs (5.1, 5.3, 5.5, or 5.7). After solubilization of the proteins, the

homogenate was centrifuged at 1000 Og for 20 min using a GS-3 rotor. The supernatant,

containing the soluble proteins, was recovered by pouring the content of the centrifuge bottle

through a strainer covered with a double layer of cheesecloth, thereby separating the top layer

and sediment from the supernatant. The supernatant was divided in two parts and the proteins









precipitated using pH 5.1 and 5.3, then centrifuged in Sorvall RC-5B at 10,000 G for

20 min using a SS-34 rotor. This was repeated for precipitation of pH 5.5 and 5.7. Soluble

protein after the 2nd centrifugation was determined by taking a 1 mL sample from the

resulting supernatant and analyzed for protein content using the Biuret Method (see

below).

Protein Measurements

Protein content was determined using the Biuret Method (50) with the addition of

10% deoxycholic acid to reduce cloudiness due to lipids. Absorbance was read at 540 nm

using an Agilent 8453 UV-VIS spectrophotometer (Agilent Technologies, Palo Alto,

CA) and protein concentration estimated using a standard curve prepared with bovine

serum albumin (BSA) in the range of 1-10 mg/mL.

Calculation of Solubility

Theoretical Protein Solubility: Defined as the fraction of the total protein soluble

after centrifugation, assuming that no protein is lost in the top layer or sediment at

centrifugation.

% Soluble Protein = mg/mL soluble protein
mg/mL total protein

Actual Protein Solubility: Defined as the fraction of the total protein soluble after

centrifugation based on the weight of the homogenate before centrifugation, the weight of

the supernatant recovered after centrifugation and its protein concentration.

% Soluble Protein = mg/mL soluble protein mL supernatant *100
mg/mL total protein mL homogenate









Viscosity

Preparation of Samples

Preparation of viscosity samples was identical to the solubility samples, except

centrifugation was skipped for construction of the viscosity curve for pH values between

1.5 and 12 and for viscosity of the solubilization pHs (2.3 to 2.9 and 10.8 to 11.4). The

second centrifugation was skipped for preparation of the homogenates solubilized at

pHs 2.5, 2.9, 11.0, and 11.2; and precipitated at pHs 5.1 to 5.7

Viscosity Measurements

The viscosity was determined using single gap cylinder geometry in an AR2000

Advanced Rheometer (TA Instruments, New Castle, DE). Homogenate sample size was

approximately 15 mL. Measurements were performed at 5C using an oscillatory time

sweep program with frequency set at 0.1 Hz, oscillatory stress at 0.1809 Pa, temperature

at 50C and a run time of 2 min (18). Graphs were constructed using the final readings of

viscosity (Pa*s).

Recovery of Proteins

For each solubilization pH (2.5, 2.9, 11.0 and 11.2) 3000 g of homogenate was

prepared. The homogenate was divided in two parts, one was used for precipitation

pHs 5.1 and 5.3 and the other for pHs 5.5 and 5.7. Waiting time for the second part of

the homogenate before pH solubilization adjustment was approximately 30 min. After

solubilization of the proteins the homogenate was centrifuged at 10,000 G for 20 min in a

Sorvall RC-5B using a GS-3 rotor. The supernatant was recovered and divided in two

parts before precipitation. Precipitated proteins were recovered by centrifugation as

described for the solubilization step. Recovery of proteins was determined and reported









as theoretical and actual recovery: a) after 1st centrifugation b) after 2nd centrifugation

and c) through the whole process.



1" centrifugation 2nd centrifugation
I--- F-- B -
'A I B: I
2.5, 2.9, 11.0 or 11.2 5.5

Figure 3-1. A: Soluble protein in initial homogenate (total protein), B: Soluble protein in
supernatant after 1st centrifugation, C: Soluble protein in supernatant after
2nd centrifugation.

Protein Recovered after 1st centrifugation:
B
% protein = B*100
A
Protein recovered after 2nd centrifugation:

% protein= 100- *100
B
Protein recovered through the whole process:

% protein = B- *100
A
Equations for actual recovery are identical, except the weight of the supernatant and

homogenate are multiplied with the protein concentration.

Electrophoresis

Proteins in the initial homogenate; top layer, supernatant and sediment after 1st

centrifugation; supernatant and sediment after the 2nd centrifugation were separated

according to the method described by Laemmli (51). Samples were prepared in small

plastic vials by adding 170 pL of diluted protein samples to 330 [tL of Laemmli sample

buffer (BioRad Laboratories, Inc., CA), mixed well and 25 [tL P-mercaptoethanol added.

The vials were placed in boiling water for 5 min, cooled on ice and frozen at -300C.

The samples (-3 mg/mL) were applied to precast gels 4 to 15% or 4 to 20% (BioRad

Laboratories Inc., Hercules, CA) and run in a Mini PROTEAN 3 system (BioRad,









Laboratories Inc., Hercules, CA) using a constant current of 200mA (-45 min). A wide

range (6.5 to 205 kDa) SigmaMarkerTM molecular weight standard (Sigma Chemical Co.,

St. Louis, MO) was run for each gel. The wide marker contains thirteen proteins;

mysosin (205 kDa), P-galactosidase (116 kDa), phosphorylase b (97 kDa), fructose-6-

phosphate kinase (84 kDa), albumin (66 kDa), glutamic dehydrogenase (55 kDa),

ovalbumin (45 kDa), glyceradlehyde-3-phosphate dehydrogenase (36 kDa), carbonic

anhydrase (29 kDa), trypsinogen (24 kDa), trypsin inhibitor (20 kDa), oa-lactalbumin

(14.2 kDa), aprotinin (6.5 kDa). The running buffer (pH 8.3) was prepared (500 mL) for

each run by diluting a stock solution (30.3 g of Tris base, 144.0 g glycine and 10.0 g of

SDS to 1000 mL) ten fold with cold DI water followed by thorough mixing. The protein

bands were fixed in 12% TCA for 1 hour and stained overnight using an EZBlueTM Gel

Staining Reagent (Sigma Chemical Co., St. Louis, MO).

Preparation of Protein Isolates

Ground muscle was mixed with cold DI water (1:9, w/w), homogenized for 60 sec

(2x30 sec) with a Waring blender (Waring Products Division, New Hartford, CT) at 40%

electrical output and carefully poured (to avoid foaming) into a plastic beaker on ice. The

homogenates were adjusted to pHs 2.5, 2.9, 11.0 or 11.2 (-10 min) to solubilize the

proteins, using 2 M HCI or 2 M NaOH. Then the homogenate was transferred to

centrifuge bottles and centrifuged at 10,000 G for 20 min in a Sorvall RC-5B using a GS-3

rotor. Centrifugation resulted in the formation of 3 layers; the top layer containing

mostly neutral lipids, the supernatant containing the soluble proteins and the sediment,

containing insoluble material. The supernatant was separated from the top layer and the

sediment by pouring the contents of the centrifuge bottles through a strainer covered with











two layers of cheesecloth. The top layer and sediment were discarded. The collected

supernatant was subjected to isoelectric precipitation by adjusting the pH to 5.5,

following centrifugation at 10,000 G for 20 min. The resulting supernatant was discarded

but the sediment (protein isolate) was placed in a zip-lock bag and stored on ice in a cold

room at 40C overnight.


I- I-
Top layer \_ Supernatant I_
Soluble Protein \
------- Protein Isolate
Insoluble Material
Isolate

'A ID
centrifugation 2nd centrifugation
Solubilization Precipitation
of muscle of soluble Protein I
proteins I protein I Isolte


A. Tilapia white muscle was homogenized for 60 sec using a Waring blender, transferred to a plastic
beaker on ice and the muscle proteins solubilized by lowering the pH to 2.5 or 2.9, or increasing the
pH to 11.0 or 11.2.
B. The pH adjusted homogenate was centrifuged for 20 min at 10,000 G in a Sorvall RC-5B using a
GS-3 rotor.
C. After 1st centrifugation, three layers are formed; top layer, sediment and supernatant. The
supernatant was recovered by pouring the contents through a strainer covered with two layers of
cheesecloth. The top layer and sediment are discarded.
D. The soluble proteins are precipitated by adjusting the pH to 5.5
E. The precipitated proteins are centrifuged again for 20 min at 1,0000 G in a RC-5B using a GS-3 rotor.
F. Two layers are formed after 2nd centrifugation, supernatant (contains soluble proteins) which is
discarded and the protein isolate that is recovered and used for subsequent experiments.
G. The recovered protein isolate is stored in a zip-lock bags on ice at 40C.



Figure 3-2. A schematic showing protein isolation from tilapia white muscle.

Preparation of Washed Muscle

Ground muscle was mixed with cold DI water (1:3, w/w) and allowed to sit for

15 min, stirring every few minutes with a plastic spatula. The slurry was poured into a

strainer covered with a double layer of cheesecloth and the water manually squeezed out.

This was repeated 2 times, with the last wash water containing 0.2% NaCl to aid in the









dewatering of the tilapia white muscle proteins. The washed muscle was placed in a zip-

lock bag and stored overnight on ice in the coldroom at 40C before gel preparation.

Rheology

Preparation of Samples

To determine moisture content of the isolates and washed muscle a ~5 g sample

was placed in a Cenco moisture balance (CSC Scientific Company, Inc., Fairfax, VA).

Moisture content was adjusted to 90% by adding cold DI water based on formula (3).

Tests however revealed that the samples usually required slightly less water than the

formula indicated.

g X g sampe (Xmosture of sample Xwanted moisture)
g I 20 = ----------
Xwanted moisture I

Protein isolate (25 or 30 g) was weighed into a 100 mL plastic beaker,

homogenized with a hand held Tissue Tearor (Biospec Products, Inc, Bartlesville, OK).

After homogenizing for 1 min on speed 6, 25 mM of sodium phosphate dibasic

(pulverized with a mortar and pestle to reduce particle size) was added and the paste

homogenized again for 2 min. Finally 2% NaCl was added (when applicable) and mixed

well with a stainless steel spatula. The pH of the paste was adjusted to 7.1 to 7.2 with 2 M

NaOH (-4-5 120 [tL was needed of 2M NaOH to samples with salt while samples

without salt needed -3 120 [tL) followed by mixing with a stainless steel spatula. After

the pH adjustment the beaker was covered with parafilm and the paste allowed to sit on

ice for 30 min before theological measurements. Total sample preparation including the

30 min setting time was approximately 50 to 60 min.









Measurements

Viscoelastic changes on heating and cooling were determined using single gap

geometry in an AR2000 Advanced Rheometer (TA Instruments, New Castle, DE).

Approximately 20 g of sample were placed in the sample chamber at 50C and the head

set to a specified gap (4950 [tm). After the head reaches the gap, excess sample was

removed with a stainless steel spatula and a layer of mineral oil was placed on top of the

sample to prevent evaporation on heating. The opening was covered with a metal

moisture trap also to prevent evaporation. The samples were heated from 5 to 800 C at a

rate of 2 C/min and cooled from 80 to 50 C at the same rate and measurement conducted

using a oscillatory mode with constant frequency set at 0.1 Hz and maximum strain at

0.01(8).

Gel Preparation

Protein isolates were adjusted to 83% moisture content by squeezing the water

manually out of the isolates. The washed muscle was centrifuged at 10,000 G for 20 min

in a RC-5B centrifuge using a GS-3 rotor to reach the right moisture content. Moisture

was determined as described before. Approximately 130 g of isolate/washed muscle

were accurately weighed and placed in a Mini Chopper (Sunbeam Products Inc., Boca

Raton, FL) with 25 mM Sodium Phosphate Dibasic buffer added after blending for 20 sec

and 2% NaCl added after ~1 min (when applicable). The pH was adjusted to 7.1 to 7.2

by adding 2M NaOH dropwise. The paste was mixed for a total of 4 min with all steps

performed in a cold room at 40C. The paste was then manually stuffed into steel tubes

(diameter 19 mm) and the ends sealed with a rubber cap and fastened with a hose clamp.

The paste was cooked for 30 min at 800C in a Precision water bath (Precision Scientific,

Winchester, VA) and cooled in ice water for 15 min. Gels were removed from the









stainless steel tubes and placed in zip-lock bags and stored in a cold room at 40C for

48 hours prior to testing.

Gel Quality Analysis

Torsion Testing

After storing the gels at 4 C for 48 hours they were cut into 28.7 mm long samples

using a cutting motive. After reaching room temperature (-40 min) the gels were milled

into a dumbbell shape with a minimum center diameter of 1.0 cm. Gels were tested using

a modified Brookfield DV-II viscometer (Brookfield Engineering Laboratories, Inc.,

Stoughton, MA) or (Gel Consultants, Raleigh, NC) and the samples twisted at 2.5 rpm

until structure failed. Shear stress (resistance to breakage) and shear strain (distance until

breakage) of the gels were obtained using computer software linked to the viscometer.




















C D


Figure 3-3. Modified Brookfield DV-II viscometer. A) Brookfield Engineering
Laboratories, Inc., Stoughton, MA; or Gel Consultants, Raleigh, NC.
B to D) Milling equipment.

Folding

Folding test was performed within 60 hours of storage at 40C, according to the

method of Kudo et al. (52). Approximately 3 mm slices were cut and folded by hand at

room temperature. The ability of the gels to fold was graded using a five point system

(Table 3-1).

Table 3-1. Grading system based on five-point system.
Grade Description

5 No crack occurs even if folded in four

4 No crack when folded in two but forms a crack when folded in four

3 No crack when folded in two but splits when folded in four

2 Cracks when folded in two

1 Splits when folded in two


Water Holding Capacity (WHC)

Water holding capacity on cooking was determined by analyzing the moisture

content of the paste before cooking and moisture content of the gels after cooking.









Moisture content was determined based on weight before and after drying the samples

overnight at 1060C in an oven.

WHC of the gels on pressing was determined using the method of Feng (36) where

pressing loss is defined as the water loss of a 3 mm thick slice under 3000 G pressure

(using a 3L beaker full of water) for 1 min. The sample was sandwiched between five

layers of Whatman filter paper (Whatman Inc., Clifton, NJ) which absorbed the

expressible water. Weight before and after pressing was recorded and moisture content

was determined using an oven as described before.

Expressible Water ( ) = Pre pressed weight (g) After pressed weight (g) X 100
Pre pressed Weight (g)


WHC (%1) Expressible Water Content (g) 100
Total Moisture of Pre Pressed Sample (g)

Sulfhydryl Content

Total sulfhydryl (SH) content was determined on the paste before cooking and on

the gels after cooking by using the method of Choi and Park (10) with slight

modifications. The pastes and gels were diluted 100 times to give a protein concentration

between 1 and 2 mg/mL. A 0.25 mL sample of the protein solution was added to 2.5 mL of

8 M urea, 2% sodium dodecylsulfate (SDS) and 10 mM EDTA in 0.2 M Tri-HCI buffer at

pH 7.1. To this solution 50 uL of 10 mM Ellman's reagent (10 mM 5,5'-dithiobis -

(2-nitrobenzoic acid) was added, mixed and heated in a water bath at 400C for 15 min.

After the reaction the absorbance of the solution was measured at 420 nm using an

Agilent 8453 UV-VIS spectrophotometer (Agilent Technologies, Palo Alto, CA) and total

SH content was calculated using a molar extinction coefficient of 13600 mol/cm.









Statistics

Results are expressed as means + SD. Analysis of variance (ANOVA) was used to

determine significant differences (p < 0.05) between treatments. Transformation using

arcsin was used for precipitation solubility. All results were analyzed by using the

Statistical Analysis System (SAS).














CHAPTER 4
SOLUBILIZATION OF MUSCLE PROTEINS OVER WIDE PH RANGE

Solubility

The basis for utilizing the acid and alkali processes on tilapia white muscle is to

solubilize the muscle proteins at low and high pH. In order to achieve high protein

recovery it is important to obtain high solubility at certain pH values to separate the

soluble proteins from undesirable constituents of muscle. Low solubility at certain pH

values is also important to precipitate (i.e., recover) the solubilized proteins. Therefore,

to understand how the muscle proteins react to different pH values a solubility curve was

constructed ranging from pH 1.5 to 12.0 and changes in percent soluble protein were

observed (Figure 4-1). Minimum protein solubility was observed between pH 5.0 and 6.0,

which is in the range of the isoelectric point for the majority of proteins in muscle (2).

Lowering the pH away from the isoelectric point resulted in a dramatic increase in

protein solubility up to pH 4.0 where the proteins were -94% soluble. Maximum

solubility at low pH was at pH 2.5 or -96% but decreased down to 90% at pH 1.5. This

decrease could possibly be due to anion induced aggregation, since more HCl would

increase the ionic strength of the solution and may reduce some of the electrostatic

repulsion between the proteins (53;54). An increase in pH away from the isoelectric

point resulted in a slight increase in protein solubility until pH 9.0 to 10.0 where it rapidly

increased between pH 10.0 and 11.0; and reached a maximum -99% at pH 12.0. This

solubility curve is similar to what Undeland et al. (55) observed for herring and Kim et

al. (56) observed for Pacific whiting.













100 -


80 -


S60 -


40 -



0

0 2 4 6 8 10 12 14
pH

Figure 4-1. Shows protein solubility (%) of tilapia light muscle homogenate at pH 1.5
to 12.0. Protein content was determined using the Biuret method with the
addition of 10% deoxycholic acid using a BSA standard curve. Protein
solubility was defined as the fraction of the total protein soluble after
centrifugation, assuming no protein was lost in the top layer or sediment.
Results are mean + SD.

As Figure 4-1 shows, approximately 20% of the proteins were soluble at

physiological pH (6.5 to 7.0) at the 10 fold dilution commonly used in acid and alkali

processes. Sarcoplasmic proteins which are soluble in water or solutions of low ionic

strength usually constitute about 20 to 30% of fish muscle (57). The proteins which were

soluble in the isoelectric point range were mostly sarcoplasmic proteins along with some

dissociated actin (Figure 6-7). At the isoelectric pH (5.0 to 6.0) where minimum solubility

was observed, lack of electrostatic repulsion between charged residues results in closer

contact of the proteins which promotes hydrophobic interactions (protein-protein

interaction) and aggregation. However, at pH below or above isoelctric pH, the proteins

become positively or negatively charged which results in electrostatic repulsion between









the molecules and hydration of charged residues which in turn promotes solubility of the

proteins (30). Mechanism of solubilization of myofibrillar proteins is believed to take

place in two steps; first depolymerization of the thick filaments and then dissociation of

actin from myosin (58). Addition of acid or base which raises ionic strength is not

believed to have significant effects on protein solubility as suggested by Undeland et al.

(9). The rapid increase in solubility at low pH compared to a more steady increase at

higher pH might be attributed to more ionizable groups with pKa values between 2.5 and 7.0

than between 7.0 and 11.0 (6;9).

Viscosity

Low viscosity of the tilapia muscle homogenate is important in the acid and alkali

process since it facilitates separation of soluble protein from insoluble material via

centrifugation. As an example, Undeland et al. (9) showed that total lipid removal was

reduced with increasing viscosity at low pH for homogenate prepared from herring light

muscle. Low viscosity was also shown to be important to separate membrane

phospholipids from muscle proteins, which could lead to increased oxidative stability of

the final protein isolate (6;9).

A viscosity curve of homogenate prepared from tilapia white muscle was

constructed in the pH range of 1.5 to 12.0 (Figure 4-2) in accordance to the solubility curve.

The viscosity of the solubilization pHs of interest for the acid and alkali processes were

low, or < 140 mPa*s. Two viscosity peaks were observed, at pH 5.0 and pH 9.5,

respectively. At low pH, viscosity started to increase at pH 4.5, reaching a maximum

~2 Pa*s at pH 5.0. At high pH, the viscosity started to increase around pH 8.0 and reached

a maximum ~ 10 Pa*s at pH 9.5. Homogenate prepared from catfish showed similar










results (59) whereas Undeland and coworkers (55) observed similar viscosity peaks at

low and high pH for herring light muscle homogenate.



12000

10000

8000 o

6 000

4000

2000

0000
00 20 40 60 80 100 120 140
pH
Figure 4-2. Shows viscosity (Pa*s) of tilapia light muscle homogenate at pH 1.5 to pH
12.0 in 0.5 increments. Viscosity was measured using oscillatory time sweep
in an AR2000 Rheometer. Results are mean + SD.

The large peak observed at pH 9.5 could be attributed to greater water binding

capacities of proteins at pH 9.0 to 10.0 than at any other pH as a result of ionization of

sulfhydryl and tyrosine residues (30). Interactions between hydrated residues are short

range repulsive interactions that become stronger with increasingly hydrated molecules

(34). This might contribute to increased hydrodynamic size of the proteins which in turn

leads to an increase in viscosity.

Decreased water binding above pH 10 could be due partly to loss of positively

charged g-amino groups of lysyl residues and increased solubility of the proteins (30).

Lower viscosity could also be contributed to large myofibrillar assemblies, which have a

high hydrodynamic volume, but when they break up and the effective hydrodynamic

volume significantly decreases. At pH 5.0 which is close to the isoelectric pH of the

proteins a smaller peak was observed. Moving away from the isoelectric point increases












net charge and repulsive forces resulting in facilitation of the proteins to bind water and


swell. Even just below to the isoelectric pH, proteins carry a net positive charge which


results in repulsion and hydration of residues that could increase the hydrodynamic size


of the proteins and cause an increase in viscosity (30). The drop in viscosity after pH 5.0


could be related to solubilization of the proteins which would result in lowering of


viscosity. Figure 4-3 shows that at low pH, a peak in viscosity was observed just before a


dramatic increase in solubility was observed. At high pH solubility was more gradually


increasing with increasing pH and viscosity could not be directly related to increase in


viscosity.




100

90 .-U- Solubility 12 0
-8- Viscosity
80
100
70 -

60 8 0

50
60
40

30 .40

20 -
S20
10

0 00
0 2 4 6 8 10 12 14
pH


Figure 4-3. Shows solubility and viscosity curves previously demonstrated on the same
graph. Results are mean + SD.














CHAPTER 5
SOLUBILIZATION OF MUSCLE PROTEINS AT LOW AND HIGH PH BEFORE 1ST
CENTRIFUGATION

Effect of Homogenization Times

It has been reported that different homogenization times may affect muscle protein

solubility (29) which could in turn lead to variations in protein recovery. For this reason,

effects of different homogenization times (60, 90, 120 sec) on protein solubility (%) were

investigated at solubilization pHs of interest to the acid (pH 2.3 to 2.9) (Figure 5-1 A) and

alkaline process (pH 10.8 to 11.4) (Figure 5-1 B). Overall, increasing the homogenization

time did not have considerable effects on protein solubility. Longer homogenization

times did not seem to result in lower solubility as Stefansson and Hultin (29) observed

when homogenizing cod muscle for 120 sec. At low pH, using 90 sec at pH 2.3 led to

significantly lower solubility (p < 0.01) compared to other pH values and homogenization

times. For high pH, using 120 sec at pH 11.2 led to significantly higher protein solubility

than was seen for other treatments (p < 0.01). The value obtained was above 100%,

which could be in part be reflected by experimental error or deviation. Based on these

results, further studies (solubility, viscosity and recovery) were performed by

homogenizing the tilapia white muscle for 60 sec because of good stability in solubility

and minimum foaming. To minimize the extent of protein denaturation due to shear

force or temperature increase, which in turn leads to foaming (and less protein recovery),

the tilapia muscle was homogenized in two steps of 30 sec at 40C.

















I


A 60 sec
[ 90 sec
-- 120 sec


2.2 2.3 2.4 2.5 2.6 2.7

pH


120 -

100 -

80-

60-

40-

20

0-


2.8 2.9 3.0


-A-- 60 sec
-R-90 sec
-A-120 sec


10.7 10.8 10.9 11.0 11.1 11.2 11.3 11.4 11.5

pH


Figure 5-1. Effect of different homogenization times (60, 90 and 120 sec) on protein
solubility (%) at low (A) and high (B) pH. The tilapia white muscle was
homogenized at 40 C for 30 sec at a time with a 30 sec rest time in between.
Results are mean + SD.

Viscosity

Examination of viscosity for low and high solubilization pHs showed that at low


pH viscosity was significantly higher (p < 0.05) for pH 2.9 compared to pH 2.3 and 2.5


(Figure 5-2). At high solubilization pH there was no significant difference (p > 0.05) in










viscosity (Figure 5-3). Lower viscosity was observed for certain high pH values over the

range tested compared to viscosity at low pH. There was also more variation in viscosity

at low pH compared to high pH. Adjustment to low pH (2.3 to 2.9) after storage of

homogenate over night resulted in a dramatic lowering of viscosity, down to similar

values as were seen for alkali solubilized protein (data not shown). This decrease could

possibly have been attributed to limited hydrolysis which was observed at low pH

(Figures 6-8 and 6-9, lanes 3 to 4).


2.0 -
1.8 -
1.6 -
1.4 -
1.2 -
S1.0-
0.8 -
0.6 -
0.4 -
0.2-
0.0 ...ii
-0.2 2 2.4 -- 2.6 2.8 3.0
pH

Figure 5-2. Shows viscosity (Pa*s) of tilapia homogenate at low solubilization pH (2.3
to 2.9) determined with oscillatory time sweep with controlled stress in an
AR2000 Rheometer. Results are mean SD.











2.0 -
1.8
1.6
1.4
s 1.2
0- 1.0
0.8
0.6 -
0.4 -
0.2
0.0 -.
-0.21 .6 10.8 11.0 11.2 11.4 11.6
pH

Figure 5-3. Shows viscosity (Pa*s) for tilapia homogenate at high solubilization pH
(10.8-11.4) determined with oscillatory time sweep with controlled stress in
an AR2000 Rheometer. Results are mean + SD.

At extreme low and high pH the proteins are partially denatured. Partial

denaturation of proteins can increase their hydrodynamic size and thus result in an

increase in viscosity (33). It has been suggested that proteins at extreme low pH are more

extensively denatured than proteins at extreme high pH (49) and work on herring light

muscle homogenate (55) and cod myosin (49) supported that suggestion. Kristinsson and

Hultin's work with myosin suggested that at pH 2.5 the myosin heavy chains were

dissociated whereas at pH 11.0 the heavy chains were still associated, which led to a

higher viscosity of myosin at low pH compared to high pH. Increased hydrodynamic

volume of the myosin molecule at low pH could contribute to the high viscosity observed

at low pH.

When a protein unfolds, hydrophobic groups buried in the interior of a protein are

exposed to the water surrounding them. The water tries to minimize contact to the non-

polar groups resulting in strong attractive forces between non-polar groups (60). The









proteins could form "aggregates" or other elongated structural assemblies and resulting in

an increase in their hydrodynamic volume (34) resulting in higher viscosity at pH 2.9

compared to pH 2.3. With increasing charge these forces are reduced and the viscosity is

lowered as was seen for pH 2.3.

The large variation in viscosity at low pH was interesting to note. When proteins

are denatured at low or high pH they can take on many different structural states,

depending pH and other solution conditions (61). These different states may thus have

different viscosities. It has also been hypothesized that at low pH proteins may take on

more numerous structural states than at high pH, and there may be rapid conversions

between one state to another (61). This could be one explanation between the great

variations seen in viscosity at low pH in contrast to high pH. Time might also contribute

to the variability seen at low pH compared to high pH. Undeland and coworkers (55)

showed that viscosity was reduced when acidified herring homogenate was stored on ice

for up to 25 min.

More volume was needed of 2 M HCI than 2 M NaOH to adjust the pH. The

chloride ion (1.81A) has twice the radius of the sodium ion (0.98A) and this size

difference as well as larger volume could possibly contribute to increased volume of the

proteins and therefore viscosity (60).

The trend towards decrease in viscosity at low pH with decreasing pH could be due

to dilution of the solution when 2 M HCI is added. Diluting the homogenate may result

in decreased interactions between hydrated protein molecules and their ability to absorb

water and swell which in turn results in a decrease in viscosity (33).









Recovery of Soluble Proteins

The amount of proteins recovered at low and high pH was studied by comparing

two acid solubilization pHs (2.5 and 2.9) and two alkali solubilization pHs (11.0 and

11.2). The basis for the selection of these solubilization pHs was as follows: pH 10.8 was

very difficult to work with due to instability, pH 11.4 was considered too high due to

amount of base needed to adjust the pH and possible deamination of the proteins (e.g.

indicated by the smell). For the low pH solubilization pHs, pH 2.5 and 2.9 were selected

because one of the most widely used pH in the acid aided process is pH 2.5 (10;18) and

pH 2.9 has been used with good results in other labs (62), pH 2.3 was excluded due to

very large quantities of acid needed to lower the pH (55;56).

In Figure 5-4, results are summarized and expressed as theoretical and actual protein

recovery (%) after 1st centrifugation based on total protein in the homogenate. The

difference between theoretical and actual recovery lies in difference in calculation and

representation of the results. Theoretical protein recoveries are calculated based on the

concentration of protein present in the supernatant after centrifugation compared to

before centrifugation. On the other hand, actual recovery takes into account proteins that

are lost in the separation process of soluble proteins from insoluble material, e.g. proteins

lost in the fat layer and in the cheesecloth that covers the strainer used for separation.

The bottom layer containing the insoluble material can be reprocessed to recover some of

the proteins that are lost during separation especially when using low solubilization pH

(63).

After the 1st centrifugation, actual recovery for alkali treated proteins was

significantly higher (p < 0.01). Theoretical recovery of proteins was higher than actual

protein recovery and ranged from 89% for pH 11.0 to 95% for pH 11.2. Actual recovery










ranged from 62% for pH 2.5 to 73% for pH 11.2. In regard to actual recovery, no

significant difference was observed within the acid or alkali treated proteins. There was

no significant difference (p > 0.05) in regard to theoretical recovery between low and high

solubilization pH.

A difference between the acid and alkali treated proteins was observed using

electrophoretic separation of the proteins. Hydrolysis of the myosin heavy chain (-205 kD)

was observed of supernatants at pH 2.5 (Figure 5-8, lanes 3 to 4) and 2.9 (Figure 5-9,

lanes 3 to 4) after 1st centrifugation. This was not seen for the supernatants recovered after

1st centrifugation for solubilization pHs 11.0 (Figure 5-10, lanes 3 to 4) and 11.2

(Figure 5-11, lanes 3 to 4). Apart from that similar polypeptides were recovered after 1st

centrifugation using different solubilization pHs.




0 Theoretical Recovery U Actual Recovery
94.5
100 89.8 91.8 88.7


a 61.9 a 63.8 a a
60

40

20

0
2.5 2.9 11 11.2
Solubilization pH

Figure 5-4. Theoretical and actual % protein recovery after 1st centrifugation for four
different solubilization pHs (2.5, 2.9, 11.0, 11.2). Results are mean SD.
Different letters indicate significant difference (p < 0.05) between treatments.














CHAPTER 6
PRECIPITATION OF MUSCLE PROTEINS

Solubility

Soluble proteins which are separated from neutral fat and insoluble material in 1st

centrifugation are subjected to isoelectric precipitation in the acid and alkali processes to

recover the proteins (1). Therefore, it is of great interest to investigate how different

solubilization pHs respond to various precipitation pHs and to determine the combination

which gives the highest recovery. Based on results from the solubility curve (Figure 3-1)

where minimum solubility was observed at pH 5.0-6.0, four pHs (pH 5.1, 5.3, 5.5, and

5.7) were selected to precipitate the proteins.

Precipitation of proteins solubilized at low pH values resulted in significantly

(p < 0.01) lower amount of soluble protein after 2nd centrifugation (Figure 6-1) compared

to high solubilizing pH values. This is in agreement with results for catfish reported by

Kristinsson and Demir (18). At high solubilization pH (11.0 and 11.2), the use of

precipitation pH 5.5 and 5.7 was not significantly different (p > 0.05). Proteins

precipitated at pH 5.1 contained more soluble protein after 2nd centrifugation for all

treatments except for pH 11.0 probably because pH 5.1 is further away from the

isoelectric point.

The large difference seen between low and high solubilization pHs in the

precipitation is likely to be partly explained by different degree of denaturation of the

proteins at low and high pH and then subsequently different degree of refolding as pH is

readjusted to pH 5.1 to 5.7. A more extensively denatured protein has more hydrophobic







47


areas exposed and is better able to form more and stronger protein-protein interactions.

Work with trout hemoglobin by Kristinsson and Hultin (64) demonstrated that the protein

was more extensively denatured at low pH compared to high pH, and was also less able

to refold on pH readjustment to 5.5 and 7, leaving more exposed hydrophobic groups.

This led to substantially more protein aggregation for hemoglobin refolded from low pH

compared to high pH, and thus less hemoglobin remained in solution after centrifugation,

in accordance with that seen here.


ElpH2.5 pH2.9 -pH11.0 0 pH11.2
18 -
15.7 16.0
16 --
14.3 14.3
14 -13.4 13.4 13.4 13.4
1 -----i- i--r--

12 -

10 -
_8.2
O 7.7
S8 7.1 67 6.9 6.6 6.4

6- a fg b h f c j d j j

4

2-

0-
5.1 5.3 5.5 5.7
Precipitation pH


Figure 6-1. Soluble protein (%) in supernatant after the 2nd centrifugation, using four
different precipitation pHs. Protein solubility was defined as the fraction of
the total protein soluble after centrifugation, assuming no protein was lost.
Results are mean + SD. Different letters indicate significant difference
(p < 0.05).

Viscosity

Differences in viscosity between low and high pH using various precipitation pHs

before 2nd centrifugation were not as evident as was observed in solubility after

precipitation (Figure 6-1). This indicated that the differences seen in solubility after









precipitation were not necessarily due to differences in viscosity. However, pH 2.9 did

seem to give significantly higher viscosity at precipitation pHs 5.1 and 5.3 (p < 0.05).

This difference did not reflect the results seen for precipitation solubility. Viscosity of

precipitated proteins solubilized at pH 2.9 formed larger units than when solubilized at

the other pH values, presumably due to different protein-protein interactions which

resulted from unfolding and refolding for that pH treatment. The viscosity values after

readjusting the proteins to pH 5.1 to 5.7 resulted in an enormous increase in viscosity values

compared to native protein at similar pH values (Figure 3-2). The reason for this

dramatic increase is probably due to the fact that the proteins have not refolded to their

original state leading to increased interaction and aggregation, leading to an increase in

hydrodynamic volume and therefore increase in viscosity. These changes were also

observed in a study performed on acidified and alkalized sarcoplasmic proteins from

herring white muscle did not show an increase in viscosity indicating that the myofibrillar

proteins are probably responsible for this changes in viscosity (55). Then it can be

concluded that the use of different solubilization pH (low vs. high) has more impact on

protein conformation than the use of different precipitation pHs.






49



300 -
D pH2.5 pH2.9 r pH11.0 0 pH11.2

250 -

S200 -

S150 -
0
S100 -
a s a a ma a a *b b as aa
50



5.1 5.3 5.5 5.7

Precipitation pH


Figure 6-2. Viscosity, (Pa*s) of tilapia muscle proteins at different precipitation pH
before the 2nd centrifugation. Tilapia white muscle proteins were solubilized
using four different pHs (2.5, 2.9, 11.0, and 11.2) and precipitated using four
different pHs (5.1, 5.3, 5.5, and 5.7). Results are mean + SD. Different letters
WITHIN each precipitation pH indicate significant difference (p < 0.05).










300 -
D pH5.1 pH5.3 D pH5.5 D pH5.7

250 -

200 -

150 -

a sa ib

a. a a a .b ab a .b ab
50 -

0
2.5 2.9 11 11.2
Precipitation pH


Figure 6-3. Shows the same as Figure 6-2, but in a different arrangement, categorized
based on solubilization pH not precipitation pH. Results are mean + SD.
Different letters WITHIN each solubilization pH indicate significant
difference (p < 0.05).

Protein Recovery Between 1st and 2nd Centrifugation

Protein recovery after 2nd centrifugation using different precipitation pHs was

expressed as proteins recovered after 2nd centrifugation as a fraction of total soluble

protein after 1st centrifugation. Theoretical and actual protein recovery (%) were

calculated (Figure 6-4). Opposite to what was seen after 1st centrifugation, recoveries

were higher for the acid treated proteins, or 92 to 94% whereas recovery for the alkali

treated proteins ranged from 84 to 89%. These data are in direct agreement with the data on

the amount of soluble proteins after precipitation. Where the proteins treated at low pH

resulted in lower amount of soluble protein in the supernatant after 2nd centrifugation

compared to proteins treated at high pH, indicating higher amount of protein recovered in

the protein isolate. The reason for the higher level of precipitation can be explained as







51



before that the proteins adjusted to low pH were more extensively denatured and less


refolded and thus had a higher level of aggregation. The use of different precipitation pH


therefore does not seem to have impact on protein recoveries since similar percentages


are observed for all precipitation pHs. This finding indicates that selecting a specific


precipitation pH is not critical for tilapia muscle when using the acid and alkali processes.


SActual Recovey Theoretical Recovery


D Actual Recovery Theoretical Recovery
94 1 93 0 94 2 93 1 94 3 93 2


51 53 55 57
Precipitation pH


53 55
Precipitation pH


A. B.



100 OActualRecovery Theoretical Recovery 100 OActualRecovery Theoretical Recovery

95 95

90 869 882 876 877 881 870 84 891







51 53 55 57 51 53 55 57
Precipitation pH Precipitation pH
858 869 867g85o_____6












Figure 6-4. White muscle tilapia proteins were solubilized at pH 2.5 (A), 2.9 (B), 11.0
(C) and 11.2 (D) and precipitated at four different pHs (5.1, 5.3, 5.5, and 5.7).
Theoretical and actual protein recovery (%) after 2nd centrifugation were
expressed as the percentage of total soluble protein recovered after 1st
centrifugation. Different capital letters indicate significant difference
(p < 0.05) for actual recovery. Different small letters indicate a significant
difference for theoretical recovery.









Total Protein Recovery

Protein recoveries through the whole process were determined for all treatments

including recoveries from washed muscle which was used as a control. The washed

muscle mimics the current industry standard of making surimi. Figure 6-55 summarizes

the results and showed that actual recoveries were lower than theoretical recoveries, as

would be expected. No one treatment gave statistically significantly (>0.05) better results

than other treatments. Theoretical recoveries ranged from 83% to 88% for low

solubilization pH but more variability was observed at high solubilization pH where it

ranged from 75 to 85%. This was also observed for herring light muscle (9) as well as

catfish, mackerel, mullet and croaker (18). Lower recoveries (theoretical) using the

acidic treated proteins could be explained by larger amount of proteins lost in the 1st

centrifugation. Kristinsson and Hultin (8) found that emulsification ability of pH-treated

treated proteins was significantly improved compared to untreated proteins due to higher

hydrophobicity. The acid treated proteins may thus have emulsified with lipids in the

tilapia and therefore less was recovered on the first centrifugation.

Actual recoveries for high solubilization pH were on the other hand slightly higher

when compared to acidified proteins, or ranging from 61 to 68% compared to 56 to 61%,

respectively. There were however no statistically significant (p > 0.05) differences

between low and high solubilization pHs. That could be explained by the fact that after

1st centrifugation more proteins are lost during the acidic treatment than alkali but after

2nd centrifugation more proteins are lost in the alkali treatment which evens out the

difference. Protein recoveries from washed fish muscle were 65.1% which was similar to

the results obtained for high solubilization pHs. Preparation of washed muscle involves







53



washing most of the sarcoplasmic proteins (2) possibly along with some of the


myofibrillar proteins (65).


Actual Recovery O Theoretical Recovery
83 0 837 836


* Actual Recovery D Theoretical Recovery
87 9


83 7


Precipitation pH Precipitation pH


M Actual Recovery D Theoretical Recovery


74 9 75 0


51 53 55 57
Precipitation pH


51 53 55 57
Precipitation pH


C. D.

Figure 6-5. White muscle tilapia proteins were solubilized at pH 2.5 (A), 2.9 (B), 11.0
(C) and 11.2 (D) and precipitated at 4 different pHs 5.1, 5.3, 5.5 and 5.7.
Theoretical and actual protein recovery (%) in final protein isolate was based
on total protein in initial homogenate. Results are mean + SD. Different
capital letters indicate significant difference (p < 0.05) for actual recovery.
Different small letters indicate a significant difference (p < 0.05) for theoretical
recovery.

An SDS-PAGE analysis was performed on samples collected during washing of the


muscle (Figure 6-6). The final washed muscle (lane 2) showed that the washed muscle


mainly contained protein bands tentatively identified as myosin heavy chain or MHC


(-205 kD), actin (-43 kD) and possibly some tropomyosin (35.5 kDa). During washing







54


of the muscle a large amount of actin was washed out a long with proteins with tentative

weights; -100 kDa, -69 kDa, -56 kDa (desmin), -41 kDa troponinn T), -38 kDa

(tropomyosin-beta) and -17 kDa (unidentified).


*"""""""""*



mom




1 2 3 4 5


low"


6 7


Figure 6-6. SDS page for washed tilapia muscle. Lane 1: sigma wide marker, lane 2:
washed muscle after 3 washes, lane 3: 1st wash, lane 4: 2nd wash, lane 5: 3rd
wash (0.2% NaC1), lane 6: skip, lane 7: initial sample.





mom- --W-W













1 2 3 4 5 6 7 8 9 10




Figure 6-7. Precipitation of proteins at pH 5.1-5.7 without solubilization. Lane 1: Sigma
wide marker, 2: Initial homogenate, lanes 3-6: Protein isolate precipitated at
pHs 5.1, 5.3, 5.5, 5.7, lanes 5 to 7: supernatants after centrifugation at pHs 5.1,
5.3, 5.5, and 5.7.




























1 2 3 4 5 6 7 8 4 4


Figure 6-8. Lane 1: Wide Sigma Marker, lane 2: Initial homogenate at pH 2.5, lane 3:
supernatant at pH 2.5 after 1st centrifugation, lane 4: supernatant at pH 2.5
after 1st centrifugation, lanes 6-8: isolates precipitated at pH 5.1, 5.3, 5.5 and
5.7, respectively. Lanes 9-12: Supematants after precipitation and 2nd
centrifugation at pH 5.1, 5.3, 5.5, 5.7.






i --
'I-








12 3 4 5 6 7 8



Figure 6-9. Lane 1: Wide Sigma Marker, lane 2: Initial homogenate at pH 2.9, lane 3:
supernatant at pH 2.9 after 1st centrifugation (used for precipitation pH 5.1 &
5.3), lane 4: supernatant at pH 2.9 after 1st centrifugation (used for
precipitation pH 5.5 & 5.7), lanes 6-8: isolates precipitated at pH 5.1, 5.3, 5.5
and 5.7, respectively. Lanes 9-12: Supematants after precipitation and 2nd
centrifugation at pH 5.1, 5.3, 5.5, 5.7.






56







-







1 2 3 4 5 6 7 8 9 10 11 12

Figure 6-10. Lane 1: Wide Sigma Marker, lane 2: Initial homogenate at pH 11.0, lane 3:
supernatant at pH 11.0 after 1st centrifugation (used for precipitation pH 5.1
& 5.3), lane 4: supernatant at pH 11.0 after 1st centrifugation (used for
precipitation pH 5.5 and 5.7), lanes 6-8: isolates precipitated at pH 5.1, 5.3, 5.5,
and 5.7, respectively. Lanes 9-12: Supernatants after precipitation and 2nd
centrifugation at pH 5.1, 5.3, 5.5, and 5.7.














1 2 3 4 5 6 7 8 9 10 11 12'


Figure 6-11. Lane 1: Wide Sigma Marker, lane 2: Initial homogenate at pH 11.2, lane 3:
supernatant at pH 11.2 after 1st centrifugation (used for precipitation pH 5.1 and
5.3), lane 4: supernatant at pH 11.2 after 1st centrifugation (used for
precipitation pH 5.5 and 5.7), lanes 6-8: isolates precipitated at pH 5.1, 5.3, 5.5,
and 5.7, respectively. Lanes 9-12: Supernatants after precipitation and 2nd
centrifugation at pH 5.1, 5.3, 5.5, and 5.7.














CHAPTER 7
GELATION

Gel forming ability of tilapia white muscle protein isolates prepared with the acid

and alkali processes were determined and compared to washed tilapia white muscle

(analog to conventional surimi processing). Gel characteristics were determined using

torsion, rheology viscoelasticc changes), folding and water holding capacity. Gels were

prepared with and without addition of 2% NaCl (w/w) at pH 7.1 to 7.2. These are common

gelation conditions for fish muscle gels. Adjusting the pH of the protein isolates from pH

5.5 to slightly above neutrality leads to increases electrostatic repulsion between proteins,

thus giving a more even distribution of proteins in the gel matrix resulting in good gel

quality (36;66). The use of 2.0% NaCl (w/w) is commonly used in gelation since it

partially solubilizes the muscle proteins. Previous work with washed tilapia muscle has

shown higher strength and deformation of gels prepared with 2% NaCl (w/w) or higher

compared to gels with 0.5 to 1.5% NaCl added (24). Recently, salt free muscle protein

gels at pHs above neutrality have also been found to have good gelation properties (8),

which made it of interest to study these conditions for the tilapia muscle proteins.

Torsion

In Figure 7-1, it can be seen that the shear stress (i.e., resistance to breakage) of gels

prepared using 2% NaCl (w/w) was significantly higher (p < 0.01) than that seen for gels

without added NaCl, except for gels prepared from washed muscle. For the gels with 2%

added salt, the lowest shear stress, thus least resistance to breakage, was obtained for

washed muscle gels (32.1 + 4.3 kPa). Gels from isolates made using the pH 2.5, 11.0, and










11.2 treatments gave the highest shear stress values, ranging from 80 to 84 kPa. There was

however no significant difference between these gels (p > 0.05). For the gels without

added salt, the highest stress value (69.1 + 12.0) was obtained for the isolates made using

pH 11.0 treatment while the lowest stress value was obtained for the isolates made using

pH 2.9 treatment and the washing procedure (30.4 7.0 and 26.0 2.0, respectively.


100.0 0 with 2% NaCl U Without 2% salt

90.0 83.6 80.2
69.1
80.0
70.0
58.7
2.5 47.5
60.0 a T a a
45.2
S50.0
30.4 32.1
S 40.0

30.0
20.0 C

10.0
0.0
2.5 2.9 11.0 11.2 Washed muscle
Treatment



Figure 7-1. Shows shear stress values (kPa) of gels produced from white muscle proteins
of tilapia. The use of low and high solubilization pH treatment was compared
to a three cycle washing treatment (control). Gels were cooked in steel tubes
at 800C for 30 min. The gels were stored in a cold room at 40 C for 48 hours
prior to testing with a Torsion Gelometer. Results are mean + SD. Different
capital letters indicate significant difference (p < 0.01) for treatments without
2% NaC1. Different small letters indicate a significant difference for
treatments with 2% NaC1. For each treatment gels with 2% NaCl (w/w) had a
significantly higher stress value, except for washed muscle.

Resistance of the gels to deformation or shear strain showed that the addition of 2%

NaCl (w/w) resulted in significantly higher (p < 0.01) strain values for all treatments

compared to gels without added NaCl (Figure 7-2). For samples without added salt,

isolates made using pH 11.0 and 11.2 treatments gave significantly (p < 0.05) higher strain









values (1.6 + 0.1 and 1.5 0.2, respectively) compared acid treated proteins and washed

muscle. On the other hand, isolates made using pH 2.9 and the washed muscle had the

lowest shear strain values (1.1 + 0.1 and 1.2 + 0.1, respectively), which is in line with the

results seen for shear stress. If isolates made using low pH treatment are compared it is

clear that pH 2.5 treatment with and without 2% NaCl (w/w) gave significantly higher

strain values than pH 2.9 treatment, thus being significantly more elastic.

The strain values obtained for the treatments were lower or borderline for the ideal

shear strain values for surimi from fish muscle, which are expected to be between 2 and 3.

However, the reason for lower values could be explained by the absence of

cryoprotectants and setting which are used in surimi production. Bakir et al. (67)

observed that gels prepared from Atlantic mackerel and Bluefish without using

cryoprotectants resulted in significantly lower strain values possibly due to the preventive

action towards the proteins during heating which could result in improved gelation

ability.

Shear stress and strain values were significantly higher for samples with 2% added

NaCl (w/w). Salt is believed to improve gelation ability and water retention by

solubilizing myofibrillar proteins (2). Addition of salt above 300 mM (-1%) solubilizes

the myofibrillar proteins by breaking up the interactions between myosin in the thick

filament and actin in the thin filaments, along with other cytoskeletal proteins (68). In a

concentrated gel paste however, the osmotic pressure is probably too high to obtain

completed solubilization of the myofibrillar proteins. Uniform dispersion of partially

solubilized proteins is likely to be more important in a gel paste. Addition of salt is also










believed to contribute to a more elastic gel by dispersing the proteins more evenly, which

is a consequence of partial or full solubilization.


2.5
O with 2% NaCl U Without 2% salt

19 2.0
2.0 1.9 1.9 1.7

1.6 1.6 1.5

1.5 -1.2
a b 1.1 a a b

1.0



0.5 -


0.0
2.5 2.9 11.0 11.2 Washed muscle
Treatment



Figure 7-2. Shows shear strain values of gels produced from white muscle of tilapia.
The use of low and high solubilization pH treatment was compared to a three
cycle washing treatment (control). Gels were cooked in steel tubes at 80C
for 30min. The gels were stored in a cold room at 40C for 48 hours prior to
testing with a Torsion Gelometer. Results are mean + SD. Different capital
letters indicate significant difference (p < 0.01) for treatments without 2%
NaCl. Different small letters indicate a significant difference for treatments
with 2% NaCl. For each treatment gels with 2% NaCl (w/w) had a
significantly higher strain value when compared to gels without 2% NaCl
(w/w).

It was interesting to note the difference between the two low-pH treatments. A

difference in gel performance between different low pH treatments has been seen with

other species. A study performed by Kim et al. (56) on Pacific whiting showed that

deformation (mm) of gels (contained 1.5% beef plasma protein) using a puncture test

were higher for proteins treated at pH 2.0 compared to pH 3.0. This was in part









explained by the authors by increased hydrophobicity of the proteins at low pH.

Davenport and coworkers (69) also found a significant difference in gel forming ability

among three different low pH treatments which could on the other hand not be explained

by hydrophobicity changes or changes in protein conformation using tryptophan as

a structural probe.

The pH 2.9 treatment in particular resulted in the formation of weak gels. During

isoelectric precipitation of proteins solubilized at pH 2.9 it was observed that they seemed

to form larger aggregates compared to pH 2.5 and pH 11.0. Larger aggregates do not

form as ordered three dimensional structures and thus form a weaker gel. To many

protein-protein interactions might result in an hard and inelastic gel whereas to many

protein-water interactions might result in a soft and fragile gel (63). Hydrolysis of the

myosin heavy chain (Figures 6-8 and 6-9) was observed for low pH, which could in part

explain the reduced gel forming ability of the protein isolates made with acid treatment.

This is in agreement with results by Undeland et al. (9) who observed significantly higher

stress values for gels prepared from alkali processed isolates (with cryoprotectants and

2% NaCl (w/w)) from herring light muscle compared to acid produced isolates. The

authors suggested that reduced amount of myosin heavy chain (possibly due to

proteolysis) contributed to lower stress values at acidic conditions. It is interesting to

note however that the pH 2.5 treatment with tilapia proteins was better than the pH 2.9

treatment. It is possible the proteases in question were more active at pH 2.9 than pH 2.5.

Differences in protein conformation between the two treatments may also be causing the

difference, as discussed above.









It is also worth noting that the washed muscle, acid isolates and alkali isolates had

different protein composition, which may contribute to the different gel forming ability of

these treatments (Figures 6-6 to 6-11). Interestingly the alkali process and the washing

process removed similar proteins but had significantly different gel forming abilities.

The very low stress values obtained for the washed muscle compared to the acid and

alkali treated proteins indicates that conformational differences contribute to these

differences.

Folding

The fold test is a common test used in the surimi industry for a quick evaluation of

gel quality. The fold test was performed on 3 mm slices of gels which were subjected to

the strain and stress tests mentioned before. All treatments with 2% added NaCl

exhibited excellent folding ability and received the highest score available, or 5

(Table 7-1). Gels prepared using alkali treated proteins without 2% NaCl (w/w) exhibited

excellent folding ability and were double folded without breaking. On the other hand, the

salt free gels prepared using acid treated proteins and washed muscle performed very

poorly. The lowest folding score was obtained for gels prepared from isolates made with

pH 2.9 treatment which split in two during the first fold (score 1). Gels prepared from

washed muscle and pH 2.5 cracked without splitting during the first fold and received a

score of 2.









Table 7-1. Quality of protein gels as assessed by the fold test. Gel quality was estimated
by folding approximately 3 mm thick gel slices by hand at room temperature
and grading the quality using a five point system., 5: No crack occurs even if
folded in four, 4: No crack when folded in two but forms a crack when folded
in four, 3: No crack when folded in two but splits when folded in four, 2:
Cracks when folded in two, 1: Splits when folded in two. Results are mean +
SD. Different letters within each column indicate significant difference
(p < 0.05). Different numbers within each row indicate significant difference
(p < 0.05).

Added NaCI 2.5 2.9 11 11.2 Washed Muscle
2% 5.0 0.0 a' 5.0 0.0 a'1 5.0 0.0 a'1 5.0 +0.0 a, 5.0 0.0 a'1
0% 2.0 1.5 b1 1.0 0.0 b2 5.0 0.0 a'3 5.0 0.0 a,3 2.0 1.2 b 1


Rheology

Small strain oscillatory rheological testing was used to follow changes in

viscoelastic properties of gels during heating and cooling. Initial and final storage

modulus (G') during gelation was determined for all treatments in two separate

experiments (replicate 1 and 2). The results for initial G' for both replicates are

summarized in Figure 7-3. Due to large variation in data obtained for final G' between

the replicates, the results are represented in two separate bar graphs (Figure 7-4 and

Figure 7-5).

All treatments exhibited higher initial G' (Pa) in the absence of NaCl (Figure 7-3)

compared to samples with 2% NaCl (w/w) added. A higher G' translates to a more rigid

system. The protein paste was at pH 7.1 to 7.2 which would give the muscle proteins a

substantial negative charge (70). This negative charge creates strong repulsion forces

between the proteins, creating more space for water to enter, and a more expanded and

rigid system. When 2% NaCl (w/w) is added the ions screen some of these repulsive

forces bringing the proteins closer together and decreasing the space available for water

to enter (46) and therefore decrease hydration and make the system less expanded. This









explains the higher initial G' for samples without added salt. There were some

differences in initial G' between treatments. For example, the washed muscle showed

highest value among the samples that contained salt. This is likely due to more structure

in the washed muscle, since some of the myofibrils would still be intact for that system.





8000 E with 2% NaCl U without 2% NaCl
7000 -
6000 -
5000 -

4000 -
3000 -
2000 -
1000 -


2.5 2.9 11 11.2 Washed
Muscle
Treatment



Figure 7-3. Storage modulus (G') of protein pastes at 50C before gelation. Results are
mean + SD.

The final G' represents samples that have undergone thermally induced gelation

along with setting on cooling (Figure 7-4 and Figure 7-5). The results for final G'

showed that the acid treated samples exhibited better gelling ability in replicate 2 than in

replicate 1, indicating significant variations, which could be attributed to the procedures

used, the isolate or the raw material. There is however evidence that the acid process

may lead to larger variations in gelation when using small strain oscillatory testing









compared to the alkali process (20). The reason for this is unknown, but is hypothesized

to be due to unstable structural protein conformations that can form at low pH, thus

leading to different refolded structures at pH 5.5, where the proteins are precipitated.

Although variations were large, the results indicate that the proteins from the alkali

process have the ability to form stronger gels in both replicates compared to the proteins

from the acid process, which partly agrees with the torsion results.

In contrast to the torsion results where samples containing 2% NaCl (w/w)

exhibited better gelling ability, samples without salt seemed to form stronger gels,

especially in replicate 2, except for the isolates made with the pH 2.9 treatment. The

small scale oscillartory testing and the torsion testing are not necessarily expected to go

hand in hand, since they are different tests done at different protein concentrations. The

small scale oscillatory testing gives more insight into the gel forming mechanism and

protein-protein interaction potential of the muscle proteins at lower concentrations, while

the torsion test measures gel strength and quality at high protein concentration. It is

interesting to note the higher G' for the isolates from the alkaline process, suggesting it

has substantially more protein-protein interactions and higher gel forming potential at

lower protein concentrations than higher concentrations compared to the isolates from

acid process which performed poorer at lower concentrations compared to higher

concentrations. The rheology results also indicated that the washed muscle performs

better relative to the alkali isolates, at low compared to high protein concentrations.












30000

25000

20000

15000

10000


0 with 2% NaCI U without 2% NaCI


I


5000



2.5 2.9 11 11.2
Treatment



Figure 7-4. Storage modulus (G') of protein pastes at 50C after gelation. Data
represent the first replicate. The final G' for gels was obtained by heating the
protein samples from 5 to 80 C followed by cooling from 80 to 50C at a rate
of 20C/min using a small strain oscillatory procedure. Results are mean + SD.













30000 -
O with 2% NaCl U without 2% NaCl
25000 -

20000 -

S15000 -

10000 -

5000 -


2.5 2.9 11 11.2 Washed
Treatment Muscle



Figure 7-5. Storage modulus (G') of protein pastes at 50C after gelation. Data
represent the second replicate. The final G' for gels was obtained by heating the
protein samples from 5 to 80 C followed by cooling from 80 to 5 C at a rate
of 20C/min using a small strain oscillatory procedure. Results are mean + SD

It was of interest to observe the changes in G' during heating and cooling for the

acid and alkali treated proteins and washed tilapia white muscle to gain an insight into

possible differences in gel forming mechanisms between the samples (Figures 7-6 to

7-15).

Washed tilapia muscle with 2% (w/w) added NaCl (Figure 7-6) had a stable G',

from 5 to 380C, where it started to increase until peaking at approximately 440C and

then declining down to a minimum at 530C. The G' then rose steadily until 800C was

reached. A similar curve for heating was observed by Klesk and coworkers (37) who

studied the effect of state of rigor on the gel-forming ability of washed tilapia muscle.

The increase up to 440C has been attributed to cross-linking of myosin and the drop in









G' after that to denaturation of light meromyosin, leading to increased fluidity. The

second increase in G' has been attributed to the formation of permanent cross-linked

myosin filaments. A rapid increase in G' was observed during cooling of the gel

indicating formation of a firm gel structure

Myofibrillar proteins, especially myosin and actin are believed to be largely

responsible for gelation. Therefore, gelation of myosin, actin and their complex

actomyosin has been the subject of many studies. For myosin, the heavy chain is

believed to be the main subunit involved in the gelation. The role of the light chain

seems to be dependendent on ionic strength where at high ionic strength they are not very

important. However, at low ionic strength, removal of the light chains the rigidity of

myosin gels was substantially lowered (2). Results from these studies have been used to

try to elucidate gelation mechanism of complex mixture of myofibrillar proteins (2). The

increase in G' that was observed for the washed muscle on heating is attributed to

dissociation of light chains from the heavy chains resulting in conformational changes to

the molecule. Cross-linking of myosin is also believed to be responsible for the increase

in G' since it forms a strong elastic network. The subsequent drop is postulated to relate

to dissociation of the actin-myosin complex and unfolding of the alpha-helix portion of

the myosin rod (35). This transformation leads to a large increase in fluidity and might

disrupt some protein networks that have already formed (35). All these structural

changes lead to rapid aggregation and formation of gel networks resulting in a steady

increase in G' (2).















20000 -

18000 -
o
16000 o
0
14000 o
0
12000 00

S10000 000 Cooling
0 o0o
8000 o0

6000 Heating oooooooo0oo

4000 -0 o o0 o
OOD Oo moooo mDOOO 0 0 0o
2000 o

0
0 10 20 30 40 50 60 70 80 90
Temperature (0C)





Figure 7-6. Example of a typical rheogram obtained for washed tilapia muscle with
2% NaCl (w/w) that shows storage modulus (G') during heating at 5 to 80C
followed by cooling from 80 to 50C at 20C/min. Samples (-10% protein (w/w))
adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating.

Acid treated proteins with 2% NaCl (w/w) behaved quite differently from the

washed muscle. The pH 2.5 treated proteins showed a continuous increase in G' which

started around 360C and leveled off at approximately 650C (Figure 7-7). A similar

pattern was observed for the heating phase of the pH 2.9 treatment (Figure 7-8). The

onset of gelation was thus at a significantly lower temperature that that seen for washed

tilapia muscle. This could be due to a more unfolded structure of the acid treated proteins,

thus requiring a lower temperature to further unfold and interact on heating. The same







70


was seen for cod muscle proteins by Kristinsson and Hultin (8). The large difference

observed between replicates of the acid treated proteins has been discussed in the

previous section and maybe attributed to unstable protein structures at low pH leading to

different refolded structures which can affect the gelation properties.




16000 -
A Replicate 1
14000 o Replicate 2

12000 0
0
10000 0
0
8000 -

0 Cooling
6000 Heating 0 D

4000 &A Cooling 003[3


2000 A A a
0 0 0 o301:1 n O3hMn n 0 z.AA
A AAAAAA 40AAAAAA AA A AM A A Ata=
0 I I
0 10 20 30 40 50 60 70 80 90
Temperature (C)



Figure 7-7. Examples of typical rheograms from two independent experiments (replicate
1 and 2) obtained for proteins treated at pH 2.5 with 2% NaCl (w/w). The
rheograms show storage modulus (G') during heating at 5 to 800C followed by
cooling from 80 to 50C at 20C/min. Samples (-10% protein (w/w)) adjusted
to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating.














16000 -

14000 b A Replicate 1
03 0 Replicate 2
12000 0

10000 -
0
0
8000 -
Cooling 0 Cooling
6000 A / 00

4000 H eatig AA 000^:

2000 .. n aAA A

0
0 20 40 60 80 100
Temperature (0C)



Figure 7-8. Examples of typical rheograms from two independent experiments (replicate
1 and 2) obtained for proteins treated at pH 2.9 with 2% NaCl (w/w). The
rheograms show storage modulus (G') during heating at 5 to 800C followed by
cooling from 80 to 50C at 20C/min. Samples (-10% protein (w/w)) adjusted
to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating.

The alkali treated proteins showed a similar viscoelastic curve on heating as the

acid treated proteins. The G' started to gradually increase around 400C but as presented

previously had a larger final G' indicating the formation of a firmer gel compared to the

acid treated proteins (Figure 7-9 and Figure 7-10). The slightly higher point of gelation

for the alkali treated proteins compared to acid treated proteins suggested differences in

conformation. The higher onset temperature for the alkali treated proteins may suggest

that they are more refolded than the acid treated proteins. The protein composition










differences between the two isolates might also partly account for the differences in

temperature sensitivity.







22000 -


20000

18000

16000

14000

'" 12000

0 10000

8000

6000

4000

2000

0


A Replicate 1

* Replicate 2


Cooling
0
[]
[]


Cooling


Heating A nA3 OOO
A ^ A 0 D OI
A D D


0 10 20 30 40 50 60 70 80 90

Temperature (0C)


Figure 7-9. Examples of typical rheograms from two independent experiments (replicate
1 and 2) obtained for proteins treated at pH 11.0 with 2% NaCl (w/w). The
rheograms show storage modulus (G') during heating at 5 to 800C followed
by cooling from 80 to 50C at 20C/min. Samples (-10% protein (w/w))
adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating.






73







26000
24000 -
22000 A A Replicate 1
20000 -A
20000 Cooling 0 Replicate 2
18000 A
16000 -
'" 14000 A
00 A3 Cooling
0 12000 0 0
10000 a
8000 0 3 AAA AA A
6000 000 A A
4000 Heating A[ ooH0 13
2000 &1A3 O M0 03LLL0L


0 10 20 30 40 50 60 70 80 90
Temperature (0C)




Figure 7-10. Examples of typical rheograms from two independent experiments
(replicate 1 and 2) obtained for proteins treated at pH 11.2 with 2% NaCl
(w/w). The rheograms show storage modulus (G') during heating at 5 to 80C
followed by cooling from 80 to 50C at 20C/min. Samples (-10% protein (w/w))
were adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before
heating.

The acid and alkali treated proteins with added 2% NaCl (w/w) did not show the

same transitions as were observed for washed muscle possibly due to denaturation of the

muscle proteins during processing. It has been found that major changes occur with

myosin on acid and alkali treatment (8). It was for example found that acid treatment led

to complete dissociation of cod myosin while alkaline treatment only led to the

dissociation of the light chains from the myosin head. As a result the proteins had

different viscoelastic behavior on heating and cooling compared to a washed cod muscle









(8). It was believed that the absence of light chains led to protein aggregation and cross-

linking at lower temperatures due to a more exposed myosin head group.

Yongsawatdigul and Park (71) also observed that acid and alkali treated proteins from

rockfish were denatured during treatment while still retaining ability to form gels. The

lower gelation ability of the acid-treated proteins compared to the alkali treated could be

due to different conformational changes; partly because of loss of myosin heavy chain during

processing, or because of unfavorable conformation of the proteins during acidic

treatments (e.g., too many hydrophobic groups, leading to bigger aggregates and a less

ordered gel). Another explanation for the poor gelling ability of the acid-treated proteins

could be the presence of denatured sarcoplasmic proteins, many which were retained in

the acid process but not in the alkali and washed process. Crynen and Kristinsson

demonstrated with catfish proteins that the acid process not only negatively affect the

myofibrillar proteins but also lead to changes in the sarcoplasmic proteins that, when

mixed with the myofibrillar proteins, have a very detrimental effect on gelation.

Washed muscle without 2% added NaCl (Figure 7-11) exhibited different

viscoelastic behavior on heating compared to the washed muscle containing salt. Initial

G' was higher and decreased with increasing temperature until it reached a minimum at

460C, which was 70C lower than the minimum observed for washed muscle with 2%

NaCl (w/w). This decrease suggests that protein complexes may have dissociated instead

of forming cross-links as in the system with salt. A steady increase, similar to the one

seen for the proteins in salt, was then observed with increasing heating excluding a small

bump at approximately 560C. The increase in G' on cooling was similar for both

washed muscle in the presence and absence of salt.














20000 -
o
18000 -
0
16000 -
o
0
14000 o
o
12000 0 Cooling

10000 0
8000 Heating 0o
8000 o

6000 a +o + ^ S^

4000 -o o

2000 -0

0
0 10 20 30 40 50 60 70 80 90
Temperature (0C)





Figure 7-11. An example of a typical rheogram obtained for washed tilapia muscle
without 2% NaCl. (w/w) that shows storage modulus (G') during heating at
5 to 800C followed by cooling from 80 to 50C at 20C/min. Samples (-10% protein
(w/w)) were adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice
before heating.

When the viscoelastic curves for samples without added NaCl were compared it to

samples with 2% NaCl (w/w) they were quite different and resembled the curves for

washed tilapia muscle. The acid treated proteins showed a slight decrease in G' on

heating until they reached approximately 400C, where the G' dropped down to a

minimum at 47C. The G' then increased again and leveled off at higher temperatures.

Similar to the samples containing salt, the acid treated proteins without added salt showed

large variability and only one curve from both replicates was usable for the isolates made










with the pH 2.5 treatment (Figure 7-12). Protein isolates made with the pH 2.9 treatment

showed a very similar trend, however with a lower final G' (Figure 7-13). The alkali

treated proteins without added salt showed the same trend as the acid treated proteins

except for a higher final G' and less variability (Figure 7-14 and Figure 7-15).


AA A
AA igA
Heating A



AA A A AHtM^


Cooling


A^
^ AA AlM & a


0 10 20 30 40 50

Temperature (C)


60 70 80 90


Figure 7-12. Examples of typical rheograms from obtained for proteins treated at pH 2.5
without 2% NaCl (w/w). The rheograms show storage modulus (G') during
heating at 5 to 800C followed by cooling from 80 to 5C at 20C/min. Samples
(-10% protein (w/w)) adjusted to pH 7.1 to 7.2 were allowed to sit for
30 min on ice before heating.


16000


14000


12000


10000

8000


6000


4000


2000


AA
A
A
A















a Replicate 1
* Replicate 2


Cooling


8000 -

6000- 000 aa

4000 0 a a 0a00a0A

2000 0 0H D10 nn o
Heating


N C nnnm IAAA
00
0 00 0 00003


0 10 20 30 40 50 60 70 80 90


Temperature (0C)



Figure 7-13. Examples of typical rheograms from two independent experiments
(replicate 1 and 2) obtained for proteins treated at pH 2.9 without 2% NaCl
(w/w). The rheograms show storage modulus (G') during heating at 5 to 80C
followed by cooling from 80 to 50C at 20C/min. Samples (-10% protein (w/w))
adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before
heating.


16000

14000

12000

10000
















26000
24000
22000
20000
18000
16000
14000
12000
10000
8000
6000
4000
2000


A Replicate 1
* Replicate 2


AD
AA 0
Vo


Cooling


Heating n A
A0
M AAAAA A AAAA& A
P [D [] [D ] [:][: o Oo


0 10 20 30 40 50 60 70 80 90


Temperature (0C)





Figure 7-14. Examples of typical rheograms from two independent experiments
(replicate 1 and 2) obtained for proteins treated at pH 11.0 without 2% NaCl
(w/w). The rheograms show storage modulus (G') during heating at 5 to 80C
followed by cooling from 80 to 50C at 20C/min. Samples (-10% protein (w/w))
adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before
heating.













26000 -
24000 A Replicate 1
22000
20000 o Replicate 2
18000 [b
16000 -
14000 -
12000 4A A Cooling
10000 o
8000- Heating A A /
6 0 0 0 o\ A A A A n0 1
4000 AD QDD 3 33IMM0 o0'o^ S^
2000 AAA M AM A AA
0
0 10 20 30 40 50 60 70 80 90
Temperature (0C)




Figure 7-15. Examples of typical rheograms from two independent experiments
(replicate 1 and 2) obtained for proteins treated at pH 11.2 without 2% NaCl
(w/w). The rheograms show storage modulus (G') during heating at 5 to 80C
followed by cooling from 80 to 5C at 20C/min. Samples (-10% protein (w/w))
adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before
heating.

Not many gelation studies have been performed with fish proteins in the absence of

salt, in part due to the long held believe that salt was necessary for gelation. The results

here show that gels can form in the absence of salt, and higher final G' values were in

most cases seen in the absence of salt. The absence of salt leads to more electrostatic

repulsion between the proteins at pH 7.1 to 7.2 compared to samples with salt, which may

explain the higher G'. The onset of thermal gelation was also different for samples in the

absence of salt. For example the rheological curve of washed muscle without salt









exhibited a similar pattern reaching a minimum 70C lower than compared with salt.

Another observation was that the rheology curves for acid and alkali treated proteins

without salt exhibited all similar drop in G' around 400C. This drop could be explained

by a breakup of protein aggregates with higher temperatures, and the subsequent increase

in G' due to proteins denaturing and forming permanent cross-links at higher

temperatures. Washed muscle consists mainly of myofibrillar proteins because much of

the sarcoplasmic proteins are removed during the washing step (12). The acid-produced

protein isolate contained a sizable amount of sarcoplasmic proteins, while the alkali

isolate had significantly less of the sarcoplasmic proteins. The types of proteins

recovered therefore does not seem to explain this similar behavior of the acid and alkali

isolates on heating. The above emphasizes that the differences between the two systems

are far from trivial and more investigation is needed to understand the underlying

mechanisms for gelation in salt and no salt.

Water holding Capacity

Water Holding Capacity on Cooking

The same gels which were tested with the Torsion Gelometer were analyzed for

water loss upon cooking (Figure 7-16). All treatments gave gels with good water-holding

properties. Addition of 2% NaCl (w/w) significantly reduced (p < 0.05) water loss for

protein isolates made with the pH 2.9 and pH 11.0 treatment compared to other

treatments. Most water loss or -1% was observed for acidified proteins at pH 2.9

without the addition of 2% NaCl (w/w) which is still quite good. A study performed by

Feng and Hultin (36) on washed chicken muscle protein at pH 7.0 with 0.15M NaCl

resulted in 1.7% loss of water during cooking. Another study done by Kristinsson and







81


Hultin (46) on chicken breast muscle showed that samples without salt lost more water on


cooking compared to samples containing salt.







30
With 2% added salt
20 Wthout 2% salt

10

00
25 29 110 112 Washed muscle
-1 0

-20

-30

-40
Treatments


Figure 7-16. Water loss (%) of tilapia muscle protein pastes on cooking at 800C for
30 min. Results are mean + SD.

Water Holding Capacity on Pressing

The same connection between salt concentration and water holding capacity was


seen for the gels when they were subjected to a press test (i.e., gels without 2% salt lost


more water than those containing salt, except for gels from the washed muscle where no


difference was seen). Again, the gel prepared from proteins treated at pH 2.9 lost most


water on pressing than any other treatment. The gel without added salt from the other


pH-treatments had similar water loss. The gel from the washed muscle without added salt


lost the least amount of water upon pressing. A smaller difference between treatments


was seen for samples containing 2% NaCl (w/w).













20.0

18.0

16.0 w ith 2% NaCI m without 2% NaCI

14.0
j 12.0 -

S10.0
.o 8.0

6.0 -

4.0

2.0 -
0.0
2.5 2.9 11 11.2 Washed
Treatments



Figure 7-17. Water loss (%) of heat induced tilapia muscle protein gel with and without
2% NaCl under pressure (3kg) pressing. Results are mean + SD.

Gel structure is believed to greatly influence the ability of a gel to hold water. In a

well ordered gel, movement of water might be restricted compared to a less ordered

structure (45). This is partly due to strong capillary forces within the ordered structure

and also high gel pressure if water is being held due to strong repulsive forces between

negative charges in the protein matrix (46). Between pH 7.1 and 7.2 there are strong

repulsive forces between muscle proteins and these repulsive forces strengthen even more

in the absence of salt (46). The water-holding results agree with the gelation results, in

that the weaker and more brittle gels had lower water-holding ability. Formation of

localized aggregates is believed to contribute to lower gel forming ability (36) and a

correlation between a gel elasticity and water holding capacity has been seen for chicken

breast gels (46). For example the gels from the pH 2.9 treatment had the poorest gel









forming ability (especially in the absence of salt), thus having a poorer gel structure, and

also had the least ability to hold water than the other treatments. The salt free gels in

general had poorer gel forming ability than the gels with salt, and these did also have

worse water-holding ability. The addition of salt to the proteins thus appears to have led

to a better ordered structure than the absence of salt. The salt is able to partially

solubilize the muscle proteins, and this could have been important for the formation of a

ordered structure since proteins would have been dispersed better prior to gelation.

SH-Content

Changes in free sulfhydryl (SH) groups and disulfide bonds (S-S) are often

associated with gelation of muscle proteins. It is also known that SH groups become

more reactive at the high pH values representative of the alkaline process (71). It was

thus of interest to see if the processes led to a change in SH and S-S groups. Therefore,

total SH content of protein pastes before cooking and gels after cooking was determined

by unfolding the proteins using urea and thus exposing SH groups that might be located

in the interior of the proteins (Figures 3-1 and 3-2). Washed muscle exhibited the highest

total SH content before cooking for samples without added NaC1, whereas washed

muscle with 2% NaCl (w/w) had significantly lower (p < 0.05) total SH content. There

were no significant differences (p > 0.05) between gels with no added salt and added salt

between the other treatments. There were also no significant difference (p > 0.05) in total

SH-groups after cooking between treatments and salt concentrations. It was also

interesting to see that there were no significant (p > 0.05) difference in total SH-groups

before and after gel formation.















O with 2% sal
* without 2% sal


90

80

70

.* 60

50

40

30

20

10

00


25 29 11 11 2 washed muscle

Treatments


Figure 7-18. Shows total SH group content in tilapia muscle protein paste before

cooking. Results are mean + SD.



10 0 -
O With 2% NaCl
Without 2% NaCl
90

80

70

60

50

40


30

20

10


25 29 11 112 washedmuscle
Treatments


Figure 7-19. Shows total SH group

Results are mean SD.


content in tilapia muscle protein gel after cooking.









A reduction in total SH groups indicates the formation of S-S bonds between

proteins, and has been found on many occasions on gelation of muscle proteins (68). A

significant difference between the low and high pH treatments was not found with the

tilapia protein isolates. Other studies with the acid and alkali processes have either

reported differences or no differences in reactive and total SH-groups. Kristinsson and

Hultin (49) showed that alkali treated myosin had more reactive SH groups than acid

treated myosin, and also had better get forming ability. A study by Yongsawatdigul and

Park (71) performed on rockfish muscle showed that washed muscle had lower content of

total SH compared to unwashed muscle. These authors observed a higher amount of total

SH groups in proteins made with the alkaline process compared to acid process, thus

suggesting S-S bonds formed during the acid process. The proteins made with the alkali

process had the largest decrease in SH groups, suggesting that more S-S bonds were

formed during gelation for that treatment compared to other treatments. This was not

seen with the tilapia muscle proteins. Another study performed on Pacific whiting

showed a significantly higher total SH content of isolates made at low pH compared to

high pH, suggesting SH groups became more reactive on alkali treatment. Even though

differences were seen in SH groups with Pacific whiting proteins there were no link

between these and gel forming ability. A study performed on threadfin bream showed

that total sulfhydryl groups were stable up to 300C but decreased from 40 to 800C which

indicates formation of disulfide linkages during heating (19). Again, the data with the

tilapia proteins indicate that no significant S-S bonds were formed on heating and thus

differences in SH groups could not explain the different functionality of the proteins.














CHAPTER 8
CONCLUSIONS

The use of the acid and alkali aided processes on tilapia muscle proteins regarding

isolation parameters showed that the type of proteins recovered using low and high pH

was significantly different whereas the quantity of total protein recovered was not

different.

Gel forming ability of the acid and alkali treated proteins was determined and

compared to washed tilapia muscle. Using large strain testing (torsion) showed that gels

with 2% added NaCl were stronger and more elastic compared to samples without 2%

NaCl. The alkali treated proteins also gave better results when compared to acid treated

proteins and washed muscle. The use of small strain oscillatory testing gave very

interesting results. Samples without 2% NaCl performed in some cases better than

samples with 2% NaCl. Variability at low pH was great possibly due to a more unstable

structure of the proteins. The small and large strain testing are not necessarily expected

to go hand in hand because they are different tests performed at different protein

concentration. Torsion determines fundamental gel quality and has been related to

acceptability of a product texture whereas oscillatory small strain testing gives more

insight in gel forming ability of the proteins at lower protein concentration. More

investigation is needed to understand the underlying mechanism and the molecular

properties involved in gelation of tilapia muscle proteins.

The findings from this research indicate that alkali treated tilapia muscle proteins

can form strong and elastic gels upon heating.















LIST OF REFERENCES


1. Hultin, H. 0.; Kelleher, S. D. Process for Isolating a Protein Composition from a
Muscle Source and Protein Composition. U.S. Patent 6,005,073, December 21,
1999.

2. Xiong, Y. L. Structure-Function Relationship of Muscle Proteins. In Food
Proteins and Their Applications; S. Damodaran and A. Paraf, Eds.; Marcel
Dekker: New York, 1997; pp 341-392.

3. Kristinsson, H. G.; Rasco, B. A. Fish Protein Hydrolysates: Production,
Biochemical, and Functional Properties. Critical Reviews in Food Science and
Nutrition 2000, 40, 43-81.

4. Mackie, I. M. Fish Protein. In New and Developing Sources of Food Proteins; B.
J. F. Hudson, Ed.; Chapman and Hall: New York, 1994; pp 95-143.

5. Hultin, H. 0. Recent Advances in Surimi Technology. In Recent Advances in
Marine Biotechnology; M. Fingerman and R. Nagabhushanam, Eds.; Science
Publisher, Inc.: Enfield, New Hampshire, 2002; pp 241-251.

6. Hultin, H. 0.; Kelleher, S. D. Surimi Processing from Dark Muscle Fish. In
Surimi and Surimi Seafood; J. W. Park, Ed.; Marcel Dekker: New York, 2000; pp
59-77.

7. Hultin, H. 0.; Kelleher, S. D.; Feng, Y. M.; Kristinsson, H. G.; Richards, M. P.;
Undeland, I. A. High Efficiency Alkaline Protein Extraction. U.S. Patent
6,136,959, October 24, 2000.

8. Kristinsson, H. G.; Hultin, H. 0. Effect of Low and High pH Treatment on the
Functional Properties of Cod Muscle Proteins. Journal ofAgricultural and Food
Chemistry 2003, 51, 5103-5110.

9. Undeland, I. A.; Kelleher, S. D.; Hultin, H. 0. Recovery of Functional Proteins
from Herring (Clupea Harengus) Light Muscle by an Acid or Alkaline
Solubilization Process. Journal ofAgricultural and Food Chemistry 2002, 50,
7371-7379.

10. Choi, Y. J.; Park, J. W. Acid-Aided Protein Recovery from Enzyme-Rich Pacific
Whiting. Journal of Food Science 2002, 67, 2962-2967.









11. Fitzsimmons, K. Tilapia Aquaqulture in the 21st Century. 2004; pp power point
presentation.

12. Park, J. W.; Lanier, T. C. Processing of Surimi and Surimi Seafoods. In Marine &
Freshwater Products Handbook; R. E. Martin, Ed.; Technomic Publishing
Company: Lancaster, NH, 2000.

13. Yongsawatdigul, J.; Park, J. W.; Virulhakul, P.; Viratchakul, S. Proteolytic
Degradation of Tropical Tilapia Surimi. Journal of Food Science 2000, 65, 129-
133.

14. Kristinsson, H. G.; Rasco, B. A. Hydrolysis of Salmon Muscle Proteins by an
Enzyme Mixture Extracted from Atlantic Salmon (Salmo Salar) Pyloric Caeca.
Journal ofFood Biochemistry 2000, 24, 177-187.

15. Ohshima, T.; Ushio, H.; Koizumi, C. High Pressure Processing of Fish and Fish
Products. Trends in Food Science and Technology 1993, 4, 370-375.

16. Park, J. W. Surimi and Surimi Seafood; Marcel Dekker Inc.: New York, 2000.

17. Klesk, K.; Yongsawatdigul, J.; Park, J. W.; Viratchakul, S.; Virulhakul, P. Gel
Forming Ability of Tropical Tilapia Surimi as Compared with Alaska Pollock and
Pacific Whiting Surimi. Journal ofAquatic Food Product Technology 2000, 9,
91-104.

18. Kristinsson, H. G.; Demir, N. Functional Fish Protein Ingredients from Fish
Species of Warm and Temperate Waters: Comparison of Acid and Alkali-Aided
Processing Vs. Conventional Surimi Processing. Advances in Seafood
Byproducts, Fairbanks, AK; Alaska Sea Grant College Program. 2003.

19. Yongsawatdigul, J.; Park, J. W. Thermal Denaturation and Aggreegation of
Threadfin Bream Actomyosin. Food Chemistry 2003, 83, 409-416.

20. Theodore, A.; Kristinsson, H. G.; Crynen, S. The Effect of pH and Salt on the
Gelation Properties of Catfish Surimi and Catfish Protein Isolates from Acid and
Alkali-Aided Processing. IFTAnnual Meeting, Chicago; Abstract 76A-10. 2003.

21. Davenport, M. P.; Kristinsson, H. G. Low and High pH Treatments Induce a
Molten Globular Structure in Myosin Which Improves Its Gelation Properties.
IFT Annual Meeting, Chicago; Abstract 42-9. 2003.

22. Kristinsson, H. G.; Crynen, S. The Effect of Acid and Alkali Treatments on the
Gelation Properties of Sarcoplasmic and Myofibrillar Proteins of Channel Catfish.
IFT Annual Meeting, Chicago; Abstract 87-3. 2003.

23. Park, J. W.; Korhonen, R. W.; Lanier, T. C. Effects of Rigor-Mortis on Gel-
Forming Properties of Surimi and Unwashed Mince Prepared from Tilapia.
Journal of Food Science 1990, 55, 353-355,366.




Full Text

PAGE 1

THE USE OF ACID AND ALKALI-AID ED PROTEIN SOLU BILIZATION AND PRECIPITATION METHODS TO PRODUCE FUNCTIONAL PROTEIN INGREDIENTS FROM TILAPIA By BERGROS INGADOTTIR A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Bergros Ingadottir

PAGE 3

This thesis is dedicated to my loving grandparents.

PAGE 4

iv ACKNOWLEDGMENTS My sincere gratitude goes to my major advisor Dr. Hordur G. Kristinsson for his guidance, support and encouragement for the pa st 2 years. His drive and enthusiasm for the field has been an inspiration. I w ould also like to thank my committee members Dr. Jesse F. Gregory III a nd Dr. Dwain Johnson for their valuable suggestions and guidance with this research project. It has been an honor to have the opportunity to work with such talented people. I would like to thank all my lab mates for making my 2 years here in Florida an experience I will never forget. I would like to thank my parents for always believing in me and supporting me through the years, making it possible for me to reach my goals. Finally I would like to thank my fianc for hi s endless love and support over the years.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES...........................................................................................................ix ABSTRACT ...................................................................................................................... xi CHAPTER 1 INTRODUCTION .......................................................................................................1 2 LITERATURE REVIEW ............................................................................................4 Preparation of Fish Protein Ingredients........................................................................4 Raw Material.........................................................................................................4 Conventional Surimi Processing...........................................................................5 The Acid and Alkali-Aided Processes...................................................................8 Functional Properties of Muscle proteins...................................................................12 Solubility.............................................................................................................12 Viscosity..............................................................................................................14 Gelation...............................................................................................................15 Water Holding Capacity......................................................................................18 Protein Denaturation...................................................................................................19 Research Objectives....................................................................................................20 3 MATERIALS AND METHODS ..............................................................................21 Raw Material..............................................................................................................21 Preparation of Muscle Homogenate...........................................................................21 Protein Solubility........................................................................................................21 Solubility Curve...................................................................................................21 Solubility b efore Precipitation............................................................................22 Solubility a fter Isoelectric Precipitation.............................................................22 Protein Measurements.........................................................................................23 Calculation of Solubility......................................................................................23 Viscosity.....................................................................................................................2 4

PAGE 6

66 0 5000 10000 15000 20000 25000 30000 2.52.91111.2 TreatmentG' (Pa) with 2% NaCl without 2% NaCl Figure 7-4. Storage modulus (G ) of protein pastes at 5 C after gelation. The data represents the first replicate. The fina l G for gels was obtained by heating the protein samples from 5 to 80C followed by cooling from 80 to 5C at a rate of 2C/min using a small strain oscillatory procedure. Results are mean SD.

PAGE 7

8 CONCLUSIONS .......................................................................................................86 LIST OF REFERENCES...................................................................................................87 BIOGRAPHICAL SKETCH.............................................................................................93

PAGE 8

viii LIST OF TABLES Table page 3-1. Grading system based on five-point system.............................................................31 7-1. Quality of protein gels as assessed by the folding test.............................................63

PAGE 9

ix LIST OF FIGURES Figure page 2-1. Conventional surimi processing.................................................................................6 2-2. Schematic of the acid and alkali processes................................................................9 3-1. Soluble protein before and after centrifugation........................................................25 3-2. Schematic showing protein isolation from tilapia white muscle..............................27 3-3. The modified Brookfield DV-II viscometer and the milling equipment..................31 4-1. Protein solubility (%) of tilapia homogenate at pH 1.5 to pH 12.0..........................35 4-2. Viscosity (Pa*s) of tilapia homogenate at pH 1.5 to pH 12.0..................................37 4-3. Solubility and viscosity curves demonstrated on the same graph............................38 5-1. Effects of different homogeniza tion times on protein solubility (%).......................40 5-2. Viscosity (Pa*s) of tilapia hom ogenate at low solubilization pH............................41 5-3. Viscosity (Pa*s) for tilapia hom ogenate at high solubilization pH..........................42 5-4. Theoretical and actual % protein recovery after 1s t centrifugation.........................45 6-1. Soluble protein (%) in supe rnatant after 2nd centrifugation....................................47 6-2. Viscosity, (Pa*s) of tilapia prot eins at different precipitation pH............................49 6-3. Shows the same as Figure 6-2, but in a different arrangement................................50 6-4. Theoretical and actual protein r ecovery (%) after 2nd centrifugation.....................51 6-5. Total theoretical and actual protein recovery (%) in final protein isolate...............53 6-6. Electrophoresis for washed tilapia muscle...............................................................54 6-7. Electrophoresis for precipitated proteins without solubilization..............................54 6-8. Electrophoresis for pH 2.5 treated proteins..............................................................55

PAGE 10

x 6-9. Electrophoresis for pH 2.9 treated proteins..............................................................55 6-10. Electrophoresis for pH 11.0 treated proteins............................................................56 6-11. Electrophoresis for pH 11.2 treated proteins............................................................56 7-1. Shear stress values (kPa) of tilapia gels using torsion.............................................58 7-2. Shear strain values of tilapia gels using torsion.......................................................60 7-3. Summary of storage modulus (G) of protein pastes at 5C before gelation...........64 7-4. Summary of storage modulus (G) of protein pastes at 5C after gelation..............66 7-5. Summary of storage modulus (G) of protein pastes at 5C after gelation..............67 7-6. Typical rheogram for washed tila pia muscle with 2% NaCl (w/w).........................69 7-7. Typical rheograms for proteins treated at pH 2.5 with 2% NaCl (w/w)..................70 7-8. Typical rheograms for proteins trea ted at pH 2.9 with 2% NaCl (w/w)..................71 7-9. Typical rheograms for proteins treated at pH 11.0 with 2% NaCl (w/w).................72 7-10. Typical rheograms for proteins trea ted at pH 11.2 with 2% NaCl (w/w)................73 7-11. Typical rheogram for washed tila pia muscle without 2% NaCl (w/w)....................75 7-12. Typical rheograms for proteins trea ted at pH 2.5 without 2% NaCl (w/w).............76 7-13. Typical rheograms for proteins trea ted at pH 2.9 without 2% NaCl (w/w).............77 7-14. Typical rheograms for proteins trea ted at pH 11.0 without 2% NaCl (w/w)...........78 7-15. Typical rheograms for proteins trea ted at pH 11.2 without 2% NaCl (w/w)...........79 7-16. Water loss (%) of tilapia musc le protein pastes on cooking....................................81 7-17. Water loss (%) of tilapia muscle protein gels under pressure..................................82 7-18. Total SH group content in tilapia muscle protein past e before cooking..................84 7-19. Total SH group content in tilapia muscle protein gel after cooking........................84

PAGE 11

xi Abstract of Thesis Presented to the Gra duate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science THE USE OF ACID AND ALKALI-AID ED PROTEIN SOLU BILIZATION AND PRECIPITATION METHODS TO PRODUCE FUNCTIONAL PROTEIN INGREDIENTS FROM TILAPIA By Bergros Ingadottir December 2004 Chair: Hordur G. Kristinsson Major Department: Food Science and Human Nutrition There is great interest in increasing the us e of fish muscle proteins, as food source and food ingredient, due to their functional a nd nutritional properties. Over the years several methods have been developed in order to isolate proteins fr om fish muscle, many of which cause loss of their functional prope rties. Therefore, processing methods that focus on retaining valuable functional prope rties of proteins are preferred. A good example of such a method is production of surimi Surimi is a washed fish muscle used for manufacturing of imitation seafood products. The demand for surimi products is increasing, whereas the traditional resources for surimi making are limited. Other resources, preferably less expensive, have to be utilized, such as dark muscle species or byproducts. The use of such raw material has not been successful in conventional surimi processing until the development of the acid and alkali-aided processes which enables isolation of functional protei ns. The processes are based on solubilization of muscle

PAGE 12

xii proteins at low and high pH, se paration of the soluble protei ns via centrifugation and then precipitation of soluble protei ns at their isoelectric pH. The overall objective of this study was to investigate the use of acid and alkaliaided processing to recover functional proteins from tilapia white muscle. The effects of different low and high solubilization and prec ipitation pHs on solubility, viscosity and protein recovery were determined. In addi tion, quality of gels prepared from acid and alkali treated proteins was compared to ge ls prepared using washed tilapia muscle. The use of the acid and alkali aided processe s on tilapia muscle proteins in regard to isolation parameters showed that the type of proteins recovere d using low and high pH was significantly different whereas the quantity of tota l protein recovered was not. The ability of the protein isolates to fo rm gels upon heating was compared to lab scale conventional surimi pro cessing. Gels were prepared from acid and alkali treated protein isolates with and w ithout the addition of 2% NaCl (w/w) at neutral pH and compared to gels prepared from washed muscle. Gel quality was determined using torsion, folding, oscillatory te sting and water holding capacity. Hardness (shear stress) and elasticity of gels (shear strain) was improved using 2% NaCl (w/w) compared to treatments without salt. Over all, the acid treated proteins exhibited poorer gelling ability compared to alkali treated proteins. Total content of SH groups was measured before and after gelation and S-S bonding did not explain the difference in gel forming ability of different treatments. The results indicate that the alkali aided pro cess can be used to produce high quality protein gels usable fo r manufacturing of imitation seafood products

PAGE 13

1 CHAPTER 1 INTRODUCTION There is a great interest in increasing the use of muscle proteins, as a food source and food ingredient, due to their f unctional and nutri tional properties ( 1 ). Muscle proteins from fish are nutritive, easily digested and exhibit good functionality ( 2;3 ) which makes them desirable for various food applications. However, the use of fish proteins as a food or food ingredient, has been limited due to several reasons, such as, rapid bacterial spoilage, lipid oxidation, protein oxidation, low stability compared to mammalian and vegetable proteins and loss of f unctionality during processing ( 4 ). Over the years several methods have been developed in order to is olate proteins from fish muscle, many of which cause a loss of their functional propert ies. Many of them have been relatively harsh where conditions such as combination of very low or high pH in the presence of organic solvents and high heat have been us ed. These methods were highly unsuccessful since functionality and nutriti onal quality of the products were negatively affected ( 5 ). Processing methods that focus on retaining valu able functional propert ies of the proteins are preferred and have receive d increased attention. A good example of such a method is surimi processing which involves washing fi sh muscle and adding cryoprotectants prior to freezing to stabilize the proteins. Surimi is then used for manufacturing of imitation seafood products by heating. The popularity of surimi based products is gradually increasing both in the US and Europe. The surimi industry is largely based on the utilization of Alaska Pollack, which is now under pressure from over fishing and therefore other species have to be found ( 4 ).

PAGE 14

2 Processing methods which c ould employ inexpensive raw materials to make a quality surimi would be highly desirable both to reduce the pressure of over fishing of currently used species and to reduce the cost of production. Potentia l raw materials could be; industrial fisheries pres ently exploited, seafood proces sing by-products and by-catch or unexploited/under-exploited stocks. Em ploying conventional surimi processing on these raw materials has been met with numerous problems ( 6;7 ). To address the problem of utilization of raw materials unusable for surimi processing two novel processes were developed originally at the University of Massachusetts. These processes, which both work by the same principle, involve acid or alkaline solubiliza tion and isoelectric precipitation of muscle proteins to give a hi ghly functional and stable protein isolate from low value underutilized species and by-products ( 1;7 ). The new process has been shown to work well for various cold water species such as cod ( 8 ), herring ( 9 ), and Pacific whiting ( 10 ) but currently there is little data ava ilable for the potential of using these processes to produce functional proteins from warm water species such as tilapia. Tilapia aquaculture is ra pidly growing worldwide ( 11 ), generating large amounts of by-products (and primary products ) which could be utilized for its protein content, provided the proper process is used. In a ddition, consumption of tilapia is increasing both in the US and globally. Conventional surimi processing from tilapia has been somewhat successful although the yields are fair ly low. To the best of our knowledge the acid and alkali-aided processe s have not been applied on tilapia muscle materials. Therefore, tilapia is a species of great intere st to investigate. To reach the goal of full utilization of tilapia it is esse ntial to investigate the use of the newly developed acid and alkali-aided processes on whole muscle befo re by-products can be utilized. The results

PAGE 15

3 from this research are expected to give im portant information on the production of a high quality protein isolate from aquacultured tilapia.

PAGE 16

4 CHAPTER 2 LITERATURE REVIEW Preparation of Fish Protein Ingredients Raw Material Quality and characteristics of a protein i ngredient or a finished product are highly dependent on the source of muscle protei n and the processing procedures applied ( 5;12 ). Species with white flesh and low fat content are considered most suitable for surimi manufacturing, for example Alaska pollock, Pa cific whiting, hoki, southern blue whiting, northern blue whiting and yellow croaker ( 12 ). The use of fish species with higher content of dark/red muscle and fat has met so me complications such as low grade protein gels, color problems and lipid oxidation ( 12 ). The reason for poorer gelling ability of species with darker muscle has been relate d to its characteristics; higher proteolytic activity, lower muscle pH which can lead to more rapid protein denaturation, higher concentration of sarcoplasmic proteins, highe r lipid content and hi gh concentration of heme proteins in the muscle. All these f actors have been reported as problems when producing high quality surimi from material s containing large amount s of dark muscle ( 6 ). Surimi manufacturers have resorted to several methods to alleviate this problem (e.g., by removing the dark muscle before surimi pr ocessing). The disadvant age of doing this is decreased protein recovery. Th e use of the acid and alkali aided processes has made it possible to produce good quality suri mi from dark muscle species. A major problem facing any pr otein extraction and recovery process is proteolysis by endogenous proteases. Postmortem fish musc le is very prone to proteolysis and the

PAGE 17

5 problem varies greatly with species and s eason. The effect of proteolytic activity on muscle protein gels has detrimental effect on their quality due to rapid degradation of myofibrillar proteins, in particular myosin. Proteolytic activity and types of proteases vary among species. Yongsawatdigul et al. ( 13 ) reported that serine type protease(s) were most probably involved in proteolysis of tilapia surimi resulti ng in gel weakening. Gel weakening due to proteolysis has al so been observed in Arrowtooth flounder, threadfin bream, Atlantic menhaden and lizard fish ( 12 ). Proteolysis can be a significant problem duri ng the acid and alkali-aided processes. The former process is thought to be more problematic since low pHs can activate contaminating gut enzymes (pepsin) and also certain lysosomal muscle enzymes ( 14 ). Undeland et al. ( 9 ) found that proteolysis occurred when herring protei ns were held at pH 2.7, while no proteolysis occurred at pH 10.8. Choi and Park ( 10) also reported proteolytic degradation of muscle proteins during acidaided processing of Pacific whiting. Conventional Surimi Processing Surimi was originated in Japan where it has been a tradit ional food source for centuries. It is a minced fish muscle washed with water and used as an ingredient for imitation seafood products, primarily crab subs titutes. For many years the industry was dependent on supply and avai lability of fresh fish. The discovery of adding cryoprotectants to surimi in order to pr event protein denaturation during freezing, revolutionized the industry ( 12 ) which was no longer dependent on fluctuations in supply of fresh fish.

PAGE 18

6 Figure 2-1. Conventional surimi processing. The fish (skinless fillets, fillets with the skin on, or headed and gutted fish) is minced, mechanically deboned, washed with water, reined, dewatered, cryoprotec tants added ,the product packaged, and finally frozen in blocks until used. This product is termed surimi and is used in finished products, characterized as gels, formed by using heat. In conventional surimi processing as s hown in Figure 2-1, the raw material is minced and then mechanically deboned (bones, skins and scales are removed by pressing soft tissue through small holes in a sc reen). Washing with wa ter concentrates the myofibrillar proteins, removes components that can have negative effects on gelation and compounds that can cause flavor, odor, stab ility and color problems. The amount of water and number of washes depends on the raw material, the desi red final product and water availability. The temperature should be maintained low enough to prevent protein Fish Mince fish meat Wash with water Refine Mechanically debone Dewater Add cryoprotectants, package, freeze = SURIMI

PAGE 19

7 denaturation which varies according to species ( 15;16 ). In the refining step, remaining bone pieces, skin, scales and connective ti ssue are removed. Dewatering is necessary because during the process water is absorb ed (approximately 100% increase) due to repulsion of negatively charged proteins in the washed mince (pH 6.4 to 7.0). The water minimizes the repulsion by separating th e proteins and the a ddition of salt (0.1 to 0.3%), reduces the repulsive forces by sh ielding negative charges which allows the proteins to approach one another. Thus, e xpelling water and reduci ng the tendency of the tissue to absorb water. Finally cryoprotect ants, usually sucrose (4%), sorbitol (4%) and sodium tripolyphosphate (0.2 to 0.3% ) are added in order to pr otect the proteins from denaturation and aggregation during freezi ng which would result in reduced gelation ability of the proteins ( 12 ). Prior to freezing, proteolytic enzyme inhibitors are sometimes added along with cryoprotectants to prevent proteolytic degradation of proteins during heating. For example is the use of Pacific whiting for surimi manufacturing based on addition of enzyme inhibitors and applic ation of fast heating rate to minimize proteolytic degrad ation of muscle proteins ( 12;17 ). Generally, high quality surimi is produced from white fleshed fish such as Alaska pollock but to meet the increasing demand new resources have to be found. Utilization of dark fleshed underutilized species has often le d to products with poor gelation properties. Recent studies show however that the use of warm water species like tilapia have resulted in better gel quality compared to more co mmon surimi species. For example, Klesk and coworkers ( 17 ) compared the gel quality of tropical tilapia surimi to Alaska pollock and Pacific whiting and found that it was compar able to Alaska pollock and better than Pacific whiting (without enzyme inhibition) at when heated at 90C for 15 min.

PAGE 20

8 The Acid and Alkali-Aided Processes The acid and alkali processes, originally developed by Hultin and coworkers are able to overcome some of the problems th at are involved using dark muscle, small species or by-products for surimi producti on. The processes involve using acid or alkaline solubilization of musc le proteins, followed by isoele ctric precipitation of soluble proteins to give highly functional and stable protein isolate used to produce quality surimi ( 1;7 ). As outlined in Figure 2-2, the production is carried out by subjecting a diluted slurry of homogenized muscle tissue to either a very acidic (2.0 to 3.5) or alkaline (10.5 to 11.5) pH to solubilize the majority of muscle proteins via electrost atic repulsion. As a consequence of disruption of the muscle ce ll and solubilization of the proteins the viscosity of the protein solution is drastically lowered. The lowering of viscosity allows for separation of insoluble material, such as, bones, skin, connec tive tissue, cellular membranes and neutral storage lipids from the soluble proteins by centrifugation. The soluble proteins are collected after ce ntrifugation and rec overed by isoelectric precipitation (by adjusting th e pH to 5.2 to 6.0) and then collected by centrifugation. The sediment (protein isolate) is kept and the supernatant discarded. Cryoprotectants are added to the protein isolate prior to freezing to protect the proteins from denaturation. Protein gels made from protein isolat es recovered with the new process from several species have shown to have equa l and sometimes significantly better gelation properties than those pro duced using conventional surimi processing techniques ( 6;9;10;18 ). The process has also shown to improve other functional properties. The process has given excellent resu lts for some cold water species as well as temperate and warm water species and is now in the route of commercialization for North Atlantic and

PAGE 21

9 Pacific species as well as species in the US Southeast. A study done by Kristinsson and Hultin ( 8 ) on cod muscle proteins showed that the alkali treatment improved functional properties (emulsification and gelation) of cod myosin and myofibrillar proteins. This improvement was found to be directly linked to a unique partially unfolded structure of cod myosin after alkali-treatment. Ground Fish Material Homogenization 1 p arts fish : 5 9 p arts water pH reduction or increase p H 2.5 ( HCl ) / p H 11 ( NaOH ) Sediment Layer Membrane li p ids Centrifuge (10,000 xg) Solution Phase S oluble muscle Upper Layer Neutral li p ids Protein aggregation p H ad j usted to 5.5 SEDIMENT = PROTEIN ISOLATE Centrifuge (10,000 xg) Supernatant Mostl y water, can b e Figure 2-2. Schematic diagram of the acid and alkaline process used in the production of functional protein isolates. A different response to the aci d and alkali process can be expected for warm water species compared to cold water species in part because their proteins are more heat stable due to their living environment. A st udy performed on threadfin bream actomyosin indicated that aggregation underwent at higher temperatures than was seen for cold water

PAGE 22

10 species such as cod and pacific whiting ( 19 ). A number of studies at the University of Florida have shown that several warm water sp ecies have a great potential to be utilized via these two processes. Kristinsson and Demir ( 18 ) compared the acid and alkali-aided processes to surimi on channel catfish, Spanish mackerel, croaker and mullet and demonstrated that the two processes led to significantly higher pr otein recoveries and lipid reduction than lab scale conventional surimi processi ng but also that the alkali process resulted in significan tly improved gelling ability, co lor and oxidati ve stability (due to removal of heme proteins) compared to the acid treatment and surimi process. Theodore and Kristinsson ( 20 ) showed that using the alkali-aided processe on catfish led to protein gels of greater strength when compared to using acid-aided process and conventional surimi process over a wide range of pH (pH 6-8) and ionic strengths (0-600 mM NaCl). Results by Davenport and Kristinsson ( 21 ) indicate that there are major changes in myosin (from channel catfish) th at contribute to th e improved gel strength after alkali processing. In a ddition Kristinsson and Crynen ( 22 ) demonstrated that both myofibrillar and sarcoplasmic proteins of muscle (from channel catfish) are positively affected in terms of gelation ability after al kali-treatment, but nega tively affected after acid-treatment. The molecular phenomena responsible for these differences are under investigation. Studies on tilapia using conventional pr ocessing have shown promising results ( 17;23-26 ) whereas few studies have been conducte d with tilapia muscle proteins using the acid and alkali processes to the best of the research ers knowledge. Preliminary significantly greater gel forming ability of the proteins vs. untreated muscle proteins. Treating the tilapia muscle proteins to a lo w pH (2.5) followed by readjustment to pH 7

PAGE 23

11 resulted in significantly poorer gel forming ab ilities compared to untreated proteins. The reasons behind these major differences are unkno wn and are important to determine to be able to successfully produce protein isol ates from tilapia muscle and its by-products. The acid and alkali-aided processes have several advantages over the conventional surimi process in regard to isolating functi onal and quality protein from fish muscle In the acid and alkali-aided processes pr otein recoveries are normally significantly higher due to recovery of both myofibrillar and a sizable amount of sarcoplasmic proteins. Recoveries using conventional te chniques are usually lower due to loss of sarcoplasmic proteins ( 30% of total protein) during washing steps. The acid and alkali-aided processes are si mple and require less labor than surimi processes. For example, li quefaction of the material of interest makes it easier to move it around a processing plant compared to a more solid material such as in surimi processing. Whole fish with skin and bones and fatty fish can be utilized in the acid and alkaliaided processes because proteins are se lectively separated and recovered from undesirable muscle components. This is not feasible usi ng typical surimi processing without negatively affec ting the recoveries and quality ( 5 ). Lipids and cellular membranes can be eff ectively removed in the acid and alkaliaided processes. This significantly increases the color and oxidative stability of the final product compared to proteins produ ced using surimi. Microorganisms are also effectively precipitated out during centrifugation. Removal of heme proteins are more effec tively obtained with the alkali treatment compared to surimi processing resulting in a product that is whiter and more stable to lipid oxidation. Heme proteins are also protected from denaturation and autoxidation during high pH treatment. Acid treatment leads to heme protein denaturation and co-precipitation with mu scle proteins, lead ing to color and oxidative problems. Functional properties are either retaine d, decreased (in few cases for the acid process) and often significantly improved (a lkali process) using the process. The effect on functionality is highly depende nt on species and type of functional property. Problems related to foaming and emulsifica tion of lipids have been encountered in commercialization of the acid and alkali-p rocess and have led to reduced protein recoveries.

PAGE 24

12 The use of high pH in combination of high heat can lead to formation of lysinoalanine. Acid hydrolysis can cause razemization of amino acids. This is normally not a problem in the acid and al kali-processing due to rapid process time and very low process temperature. Functional Properties of Muscle proteins Quality and stability of a final product are affected by func tional properties of proteins ( 27 ). The most important functional pr operties when producing a functional muscle protein isolate for use in food are so lubility, viscosity, wa ter holding capacity and gelation ( 1 ). These properties are discussed in more detail below. Solubility Solubility of myofibrillar proteins is belie ved to play an important role in gelation and water holding capacity of muscle proteins ( 28 ). Therefore, sol ubility of muscle proteins has been the su bject of much research. A change in solubility ca n be obtained in various ways, for example, by varying ionic strength, ion types, pH, and/or te mperature and thus affecting the hydrophobic and/or ionic nature of the proteins. For a long time it was a general belief that solubilization of myofib rillar proteins in high salt con centration (0.3 to 0.6 M) was required to form good fish gels. Studies have show n that this is not necessarily the case, Stefansson and Hultin ( 29 ) showed, that cod muscle myofibrillar proteins were soluble if the ionic strength was sufficiently reduced (< 0.3 mM) at both neut ral and acidic pH. They suggested that the negative charge of myof ibrillar proteins at neutral pH and/or in water or solutions of very low ionic strengt h, repulsive forces from negatively charged side chains are enough to drive the individua l protein molecules apart when sufficient water is available ( 29 ).

PAGE 25

13 Changes in protein solubility can also be obtained by varying the pH of the solution, as is done in the acid and alkali-a ided processes. By changing the pH, the protein acquires a net negative or net positive charge where hydration of the charged residues and electrostatic repulsion results in an increase in solubility ( 30 ). Conditions like lowering the pH near to the isoelectric point reduces the repulsive forces and allows the proteins to associate. Therefore, many proteins exhibit minimum solubility at the isoelectric point (pI) where the lack of el ectrostatic repulsion promotes aggregation between protein molecules. Due to aggregati on of proteins at thes e conditions they can be removed from solution by appropriate centrifugal force. In order to find the most appropriate pH to solubilize and reco ver proteins from a protein solution a solubility curve can be cons tructed (protein concentration vs. pH). For example, a study done by Choi and Park ( 10 ) showed that solubil ity of Pacific whiting proteins was lowest at pH 5.0 which indicated a suitable pH to precipitate the proteins. Maximum solubility was observed at pH 1.5 on the acidic side and at pH 11.0 on the alkali side. The solubility curve of channe l catfish showed that most solubility was observed at pH 2.5 and 11 and l east at pH 5.5, suggesting that the former two pH values would be suitable to solubilize the proteins and the latter pH to precipitate them from solution ( 18 ). The ionic strength of a solution has a dramatic impact on the pH-dependent solubility profile of muscle protei ns. For example, Dagher et al. ( 31 ) showed that solubility of a washed cod muscle mince incr eased dramatically between 8.9 and 9.2 at an ionic strength of 0.001 M. However the effect s of pH at high salt c oncentration were not as dependent as when the salt concentration was low. A recent study by DeWitt et al.

PAGE 26

14 ( 32 ) on using the acid-aided pro cess to solubilize and recove r proteins from beef heart showed that increasing the ioni c strength at low pH led to a decrease in solubility and thus their extractability. This reduction in solubility stems fr om screening of electrostatic repulsive forces betw een the proteins ( 33 ). Viscosity The viscosity of muscle protein homogenate s at low and high pH in the acid and alkali process is important sinc e low viscosity is necessary to separate insoluble material from soluble proteins via centrifugation. Viscos ity of a protein solution is believed to be affected by factors like prot ein concentration, pH, salt, and raw material processing which in turn can affect size, shape, fl exibility and hydration of the proteins ( 33 ). Viscosity of protein solutions usually increases exponentially with protein concentration. This is due to increased interaction between the hydrated protein molecules. When two hydrated molecules are in close proximity, s hort range repulsive in teraction forms where the strength depends on the degree of hydrat ion (more hydration => stronger repulsion and longer range) ( 34 ). An increase in ionic strengt h usually decreases viscosity by affecting the hydration capacity of the proteins ( 33 ). The two most important factors affecting viscosity of a protein solution according to Damodaran ( 33 ) are hydrodynamic size and shape of the protein molecules. Partial denaturation and/or heat induced polymerizat ion, increases hydrodynamic size of proteins and thus increases viscosity. Most macrom olecular solutions do not exhibit Newtonian behavior instead viscosity is decreased w ith increasing shear rate. This is called pseudoplastic behavior or shear thinning. This is due to the fact that the proteins align themselves in the direction of flow and weakly bound dimers and oligomers are dissociated into monomers ( 30 ).

PAGE 27

15 Gelation One of the most important functional propertie s of muscle proteins is their ability to form gels upon heating. A gel is an intermed iate stage between a solid and a liquid where polymers (e.g., proteins) form a three-dimensiona l network that is capable of holding water and other low molecular compounds ( 30 ). Characteristics of pr otein gels are determined by the type and number of prot ein-protein interactions, aggr egation and arrangement of unfolded proteins which are in turn affected by pH, ionic strength, protein concentration, and heating and cooling rates ( 35 ). The pH and ionic strength are by far the mo st important determinants of strength and quality of muscle protein gels. Optimum pH for the gela tion of muscle proteins has often been reported to be between 5.5 and 7.0, and depended on animal species, protein concentration, salt concentration, and instrument used to analyze gel strength. The use of salt (0.3 to 0.6 M) has been regarded a prerequisite in order for myofibr illar proteins to form good gels. Recently, this general belief was challenged as muscle proteins have been found to form excellent gels in the absence of salt provided that elect rostatic repulsion is sufficiently high to create a strong osmo tic pressure in the gel matrix ( 18;36 ). The formation and characteristics of muscle gels are also highly dependent on the heating procedures. Different species re spond very differently to the same heat treatment, which makes it important to determ ine the optimal heating and cooling scheme for good gel formation. As an example, gel strength can be improved for some gels by holding the protein paste at a temperature belo w 50C before the final heat treatment at ca. 80 to 90C, a process called sett ing. However, for some species this results in loss of gel quality if the temperature is held at 50 to 60C for too long due to proteolytic degradation of muscle proteins ( 5 ). In a study performed by Klesk and coworkers ( 37 ),

PAGE 28

16 tilapia formed the best gel when no setting was applied when compared to pollock and pacific whiting. Whereas, using temperature of 60C during setting was found to reduce strain and stress values of the gels and indica ted increased protease activity leading to the degradation of myosin heavy chain. According to Lanier et al. ( 38 ) the setting of fish muscle proteins at temperatures below that at which rapid aggregation occurs (ca. 40C) may be viewed as a process where partially denatured protei ns begin to interact non-covale ntly to form a fine elastic network. Setting below 40 C would allow for slow ordering of the molecules and give gels with greater firmness and cohesiveness. Hermansson ( 39 ) also reported that denaturation of proteins prior to aggregation results in a fi ner gel structure, exhibiting greater elasticity than if random aggregation occurs prior to denatu ration. Park et al. concluded that optimum setting temperature is highly related to ha bitat temperatures ( 40 ). The proteins that have been found to be primarily responsible for gel formation and gel strength are the myofibrilla r proteins most notably myosin and to some extent the complex of actin and myosin (actomyosin) ( 41 ). Grabowska and Sikorski ( 42 ) did a gel study on myofibrillar, sarcoplasmic and stroma fraction and the sarcoplasmic fraction showed no gelling ability. It has been a gene ral believe that sarcopl asmic proteins have a negative effect on gelation. However recent studies point to that that sarcoplasmic proteins may actually improve gelation in some cases. Sarcoplasmic proteins are recovered in the acid and alkali-aided proce sses but not in surimi processing. Recently Kristinsson and Crynen ( 22 ) reported that a very comple x interaction occurs between sarcoplasmic and myofibrillar proteins when mixed in different ratios under different pH and ionic strength conditions. In many cases the sarcoplasmic prot eins had a positive

PAGE 29

17 impact on the gel forming ability of myofibrillar proteins, especially when both had been treated at a high pH (11.0), as done in the alkali-aided process. Gels are complex and their properties a nd the mechanism of their formation has been the subject of many studies. There are several ways available to study the properties of a gel and the mechanism of their fo rmation. Rheology deals with flow and deformation of matter which is induced by applied force ( 43 ). There are mainly two quantities that rheology is con cerned with; stress, which is the force applied divided by an area of matter that it is applied to and strain which is the deformation of a matter induced by the stress. One type of deforma tion is shear, which is when matter changes shape without changing volume ( 44 ). An important property of a gel is deformability, which is how a gel responds to strain without breaking. There are two main rheological studies available in order to obtain information on properties of gels and the mechanism of formation, small strain testing and large strain testing. The combination of these measurements can give valuable information on acceptability of a gel. Small strain testing is a study where a sa mple is deformed without breaking the structure. Heating and cooling of a gel, us ing low frequency and sm all strain oscillatory experiments is one of the best suited methods to follow changes in physical properties relating molecular properties of a gel ( 44;45 ). The primary parameters of interest in small strain testing are; the storage modulus (G') which describes the elastic component of a protein gel, the loss modul us (G'') a measurement of the viscous attribute of the gel and the phase angle ( ), where for a perfectly elastic ma terial stress and strain are in phase or = 0 and for a perfectly viscous material = 90.

PAGE 30

18 Large strain testing is when a sample is deformed until the structure is permanently broken. Large strain testing estimates fundame ntal properties of a gel and have shown to correlate with sensory texture. An example of large strain testing is puncture test and torsion (twisting) which was used in the experiments performed for this thesis ( 44 ). The torsion test is a typical method to test the ha rdness (shear stress) and elasticity (shear strain) of surimi gels. The advantages using the torsion over a compression test e.g. is that pure shear is applied, the volume is unchanged even if the sample is compressable and the shape is maintained during testing. To maintain accuracy of the measurements the cross-sectional diameter needs to have a c onstant length. It wa s observed that shear stress was extremely sensitive (40 to 70 kPa) to changes in gel diameter (0.90 to 1.10 cm) ( 40 ). Water Holding Capacity Water holding capacity (WHC) is an important factor in muscle protein gels as it not only affects the economics of their produc tion but also their qua lity. Water holding capacity can be defined as the ab ility of a matrix (e.g., a protei n gel to retain water against gravitational force) ( 30). The level of water retained in a gel is affected by much the same factors that affect the forma tion of a good gel matrix; pH and ionic strength (i.e., salt). Feng and Hultin ( 36 ) reported that gels with an evenly distributed gel structure showed improved WHC. A poor gel matrix on the ot her hand can lead to synerisis (i.e., a discharge of water from th e gel). Gels prepared in the pH range of 6.4 to 7.4 had increasingly higher WHC as pH increased ( 46 ). An increase in WHC was shown to correlate well with increased negative char ge on the muscle proteins. A comparison between the WHC of protein is olates prepared from the aci d and alkali-aided processes

PAGE 31

19 and those prepared using surimi processing ha s not been conducted, to the best of the writers knowledge. Protein Denaturation It is a commonly held view that denaturi ng fish muscle proteins has a detrimental impact on their functional properties. Denatu ration often results in negative changes in protein functionality such as enzyme activity or loss of functional properties. In other cases, denaturation of the proteins can resu lt in improvement of functional properties such as foaming and emulsification of egg albumin ( 47 ) or improved gelation. Loss of protein functionality has been correlated to loss in ATPase activity, a common indicator of muscle protein denaturation ( 2;48 ). Interestingly, ATPase activ ity is essentially lost in the acid and alkali processing where the prot eins are partially dena tured at low and high pH and then only partially refolded when the pH is readjusted ( 49 ). The fact that acid and alkali-treated proteins often have signi ficantly improved functionality goes against common believe. Kristinsson and Hultin ( 49 ) have shown that the unique structure the proteins possess after pH-treatment is respon sible for improved functional properties such as gelation, emulsification and solubility. For example, th e partially unfolded/folded structure is more flexible and is able to fo rm better protein networks on heating (gelation) and is able to adsorb more readily to interfaces and lower interfacial tension (emulsification). Results indicate that acid treatment has a di fferent effect on the structure of the muscle proteins compared to the alkali treatment, also in a species dependent manner ( 21;49 ). It is of importance to unders tand what specific changes occur with the structure of the prot eins during the process to optim ize the functionality of these proteins.

PAGE 32

20 Research Objectives The overall objective of this study was to in vestigate use of acid and alkali-aided processing to recover functional proteins from tilapia white muscle. The effects of low and high solubilization pH and precipitation pH on solubility, vi scosity and protein recovery was determined. In addition, quality of gels pr epared from acid (pH 2.5 and 2.9) and alkali (pH 11.0 and 11.2) treated protei ns was evaluated and compared to gels prepared using washed tilapia muscle.

PAGE 33

21 CHAPTER 3 MATERIALS AND METHODS Raw Material Tilapia fillets were obtained fresh from a local supplier (Rain Forest Aquaculture, Sunrise, FL), transported on ice to the laborat ory and stored in a cold room at 4C until processed (within 24 hours). Red muscle wa s manually excised from the white muscle and discarded. The remaining white mu scle was ground using a Scoville grinder (Hamilton Beach, Washington, NC) with 6 mm hole s. All preparation of raw material and samples was performed in a cold room at 4C or on ice when applicable to maintain temperature below 5C. Preparation of Muscle Homogenate Ground muscle was mixed with cold (4C ) DI water (1:9, w:w) and homogenized for 60 sec (2x30 sec) with a Waring blende r (Waring Products Di vision, New Hartford, CT) at 40% electrical output. The homogenate was carefully poured into a plastic beaker on ice and was adjusted to the appropriat e pH by adding 2 M HCl or 2 M NaOH followed by stirring with a plastic spatula. Protein Solubility Solubility Curve A protein solubility curv e was constructed from pH 1.5 to 12.0 in 0.5 intervals. Two sets of homogenates were prepared from the same raw material. The first set was adjusted from the native muscle pH (6.5-6.9) to pH 1.5 with 2 M HCl and the second set was adjusted from the native muscle pH to pH 12.0 using 2 M NaOH. At each pH

PAGE 34

22 approximately 30 g of homogenate was accurate ly weighed into centrifuge tubes (50 mL) and centrifuged at 10,000 G for 20 min in a Sorvall RC-5B using a SS-34 rotor ( 29 ), separating insoluble material from the soluble proteins. Total protein was determined by taking a 2 g sample at each pH before centrifugation and dilu ting it 10 times with cold DI water at pH 11.0 to aid in solubilization of the proteins The samples were homogeni zed with a hand held Tissue Tearor (Biospec Products, Inc, Bartlesville, OK) on speed 7 for 20 sec and then analyzed for protein content using the Bi uret Method (see below). So luble protein was determined by taking a 1 mL sample from the supernatant after centrifugation and diluting it 5-fold with cold DI pH 11.0 when applicable (pH 5 to 6 were undiluted). Solubility Before Precipitation From the previously construc ted solubility curve, protei n solubility at low pH (2.3 to 2.9) and high pH (10.8 to 11.4) was studied in mo re detail. Effects of different homogenization times (60, 90, and 120 sec) using these low and high pHs were investigated. Total protein before centrifuga tion and soluble protein after centrifugation were determined as previously described. Solubility After Isoelectric Precipitation Protein solubility af ter isoelectric precipitation, in dicating loss of protein after 2nd centrifugation was determined using four solubilization pHs 2.5, 2.9, 11.0, and 11.2; and four precipitation pHs (5.1, 5.3, 5.5, or 5.7). After solubilization of the proteins, the homogenate was centrifuged at 1000 0g for 20 min using a GS-3 rotor. The supernatant, containing the soluble proteins, was recovered by pouring the content of the cen trifuge bottle through a strainer covered with a doubl e layer of cheesecloth, thereby separating the top layer and sediment from the supernatant. The supernat ant was divided in two parts and the proteins

PAGE 35

23 precipitated using pH 5.1 a nd 5.3, then centrifuged in Sorvall RC-5B at 10,000 G for 20 min using a SS-34 rotor. This was repeated for precipitation of pH 5.5 and 5.7. Soluble protein after the 2nd centrifugation was determ ined by taking a 1 mL sample from the resulting supernatant and analyzed for prot ein content using the Biuret Method (see below). Protein Measurements Protein content was determined using the Biuret Method ( 50 ) with the addition of 10% deoxycholic acid to reduce cloudiness due to lipids. Absorbance was read at 540 nm using an Agilent 8453 UV-VIS spectrophoto meter (Agilent Technologies, Palo Alto, CA) and protein concentration estimated usi ng a standard curve prepared with bovine serum albumin (BSA) in th e range of 1-10 mg/mL. Calculation of Solubility Theoretical Protein Solubility : Defined as the fraction of the total protein soluble after centrifugation, assuming that no protein is lost in the top layer or sediment at centrifugation. protein total mg/mL protein soluble mg/mL Protein Soluble % Actual Protein Solubility: Defined as the fraction of the total protein soluble after centrifugation based on the weight of the homoge nate before centrifugation, the weight of the supernatant recovered after centrif ugation and its protein concentration. 100 % homogenate mL protein total mg/mL nt supernata mL protein soluble mg/mL Protein Soluble

PAGE 36

24 Viscosity Preparation of Samples Preparation of viscosity samples was iden tical to the solubility samples, except centrifugation was skipped for construction of the viscosity curve for pH values between 1.5 and 12 and for viscosity of the solubiliza tion pHs (2.3 to 2.9 and 10.8 to 11.4). The second centrifugation was skipped for preparation of the homogenates solubilized at pHs 2.5, 2.9, 11.0, and 11.2; and precipitated at pHs 5.1 to 5.7 Viscosity Measurements The viscosity was determined using si ngle gap cylinder geometry in an AR2000 Advanced Rheometer (TA Instruments, New Castle, DE). Homogenate sample size was approximately 15 mL. Measurements were pe rformed at 5C using an oscillatory time sweep program with frequency set at 0.1 Hz oscillatory stress at 0.1809 Pa, temperature at 5C and a run time of 2 min (18). Graphs were constructed using the final readings of viscosity (Pa*s). Recovery of Proteins For each solubilization pH (2.5, 2.9, 11.0 and 11.2) 3000 g of homogenate was prepared. The homogenate was divided in tw o parts, one was used for precipitation pHs 5.1 and 5.3 and the other for pHs 5.5 and 5.7. Waiting time for the second part of the homogenate before pH solubilization adju stment was approximately 30 min. After solubilization of the protei ns the homogenate was centrifuged at 10,000 G for 20 min in a Sorvall RC-5B using a GS-3 rotor. The s upernatant was recovered and divided in two parts before precipitation. Precipitated pr oteins were recovered by centrifugation as described for the solubilization step. Recovery of proteins was determined and reported

PAGE 37

25 as theoretical and actual recovery : a) after 1st centrifugation b) after 2nd centrifugation and c) through the whole process. Figure 3-1. A: Soluble protein in initial homogenate (total protein), B: So luble protein in supernatant after 1st centrifugation, C: Soluble protein in supernatant after 2nd centrifugation. Protein Recovered after 1st centrifugation: 100 A B protein % Protein recovered after 2nd centrifugation: 100 100 B C protein % Protein recovered throug h the whole process: 100 A C B protein % Equations for actual recovery are identical, except the weight of the supernatant and homogenate are multiplied with the protein concentration. Electrophoresis Proteins in the initial ho mogenate; top layer, supern atant and sediment after 1st centrifugation; supernatant and sediment after the 2nd centrifugation were separated according to the method described by Laemmli ( 51 ). Samples were prepared in small plastic vials by adding 170 L of diluted pr otein samples to 330 L of Laemmli sample buffer (BioRad Laboratories, In c., CA), mixed well and 25 L -mercaptoethanol added. The vials were placed in bo iling water for 5 min, cooled on ice and frozen at -30C. The samples (~3 mg/mL) were applied to precast gels 4 to 15% or 4 to 20% (BioRad Laboratories Inc., Hercules, CA) and run in a Mini PROTEAN 3 system (BioRad, 1st centrifugation 2nd centrifugation A B C 2.5, 2.9, 11.0 or 11.2 5.5 pH pH

PAGE 38

26 Laboratories Inc., Hercules, CA) using a cons tant current of 200mA (~45 min). A wide range (6.5 to 205 kDa) SigmaMarkerTM molecular weight standard (Sigma Chemical Co., St. Louis, MO) was run for each gel. The wide marker contains thirteen proteins; mysosin (205 kDa), -galactosidase (116 kDa), phosphoryl ase b (97 kDa), fructose-6phosphate kinase (84 kDa), albumin (66 kDa), glutamic dehydrogenase (55 kDa), ovalbumin (45 kDa), glyceradlehyde-3-phos phate dehydrogenase (36 kDa), carbonic anhydrase (29 kDa), trypsinogen (24 kD a), trypsin inhi bitor (20 kDa), -lactalbumin (14.2 kDa), aprotinin (6.5 kDa). The runni ng buffer (pH 8.3) was prepared (500 mL) for each run by diluting a stock solution (30.3 g of Tris base, 144.0 g glycine and 10.0 g of SDS to 1000 mL) ten fold with cold DI wate r followed by thorough mixing. The protein bands were fixed in 12% TCA for 1 hour and stained overnight using an EZBlueTM Gel Staining Reagent (Sigma Chemical Co., St. Louis, MO). Preparation of Protein Isolates Ground muscle was mixed with cold DI water (1:9, w/w), homogenized for 60 sec (2x30 sec) with a Waring ble nder (Waring Products Division, New Hartford, CT) at 40% electrical output and carefully poured (to avoid foaming) into a plastic beaker on ice. The homogenates were adjusted to pHs 2.5, 2.9, 11.0 or 11.2 (~10 mi n) to solubilize the proteins, using 2 M HCl or 2 M NaOH. Th en the homogenate was transferred to centrifuge bottles and centrifuged at 10,000 G fo r 20 min in a Sorvall RC-5B using a GS-3 rotor. Centrifugation resulted in the form ation of 3 layers; th e top layer containing mostly neutral lipids, the supernatant contai ning the soluble proteins and the sediment, containing insoluble material. The supernatan t was separated from the top layer and the sediment by pouring the contents of the centrif uge bottles through a st rainer covered with

PAGE 39

27 two layers of cheesecloth. The top layer and sediment were discarded. The collected supernatant was subjected to isoelectric precipitation by adjusti ng the pH to 5.5, following centrifugation at 10,000 G for 20 min. The resulting supern atant was discarded but the sediment (protein isolate) was placed in a zip-lock bag and stored on ice in a cold room at 4C overnight. Figure 3-2. A schematic showing protei n isolation from tilapia white muscle. Preparation of Washed Muscle Ground muscle was mixed with cold DI wate r (1:3, w/w) and allowed to sit for 15 min, stirring every few minutes with a plastic spatula. The slurry was poured into a strainer covered with a double layer of cheesecloth and the water manually squeezed out. This was repeated 2 times, with the last wash water containing 0.2% NaCl to aid in the 1st centrifugation Top layer Soluble Protein Insoluble Material Protein Isolate 2nd centrifugation Solubilization of muscle proteins Precipitation of soluble protein Protein Isolate A B C D E F G Supernatant Isolate A. Tilapia white muscle was homogenized for 60 sec using a Waring blender, transferred to a plastic beaker on ice and the muscle proteins solubilized by lowering the pH to 2.5 or 2.9, or increasing the pH to 11.0 or 11.2. B. The pH adjusted homogenate was centrifuged for 20 min at 10, 000 G in a Sorvall RC-5B using a GS-3 rotor. C. After 1st centrifugation, three layers are formed; to p layer, sediment and supernatant. The supernatant was recovered by pouring the contents through a strainer covered with two layers of cheesecloth. The top layer and sediment are discarded. D. The soluble proteins are precipitated by adjusting the pH to 5.5 E. The precipitated proteins are centrifuged again for 20 min at 1 ,0000 G in a RC-5B using a GS-3 rotor. F. Two layers are formed after 2nd centrifugation, supernatant (contains soluble proteins) which is discarded and the protein isolate that is recovered and used for subsequent experiments. G. The recovered protein isolate is stored in a zip-lock bags on ice at 4C.

PAGE 40

28 dewatering of the tilapia white muscle protei ns. The washed muscle was placed in a ziplock bag and stored overnight on ice in the coldroom at 4C before gel preparation. Rheology Preparation of Samples To determine moisture content of the isolates and washed muscle a ~5 g sample was placed in a Cenco moisture balance (CSC Scientific Company, Inc., Fairfax, VA). Moisture content was adjusted to 90% by addi ng cold DI water based on formula (3). Tests however revealed that the samples us ually required slightly less water than the formula indicated. 1 X ) X X g X O H gmoisture wanted moisture wanted sample of moisture sample 2( Protein isolate (25 or 30 g) was weighed into a 100 mL plastic beaker, homogenized with a hand held Tissue Tearor (Biospec Products, Inc, Bartlesville, OK). After homogenizing for 1 min on speed 6, 25 mM of sodium phosphate dibasic (pulverized with a mortar and pestle to re duce particle size) wa s added and the paste homogenized again for 2 min. Finally 2% NaCl was added (when applicable) and mixed well with a stainless steel spatula. The pH of the paste was adjusted to 7.1 to 7.2 with 2 M NaOH (~4-5*120 L was needed of 2M NaOH to samples with salt while samples without salt needed ~3*120 L) followed by mixing with a stai nless steel spatula. After the pH adjustment the beaker was covered with parafilm and the paste allowed to sit on ice for 30 min before rheological measurements Total sample preparation including the 30 min setting time was approximately 50 to 60 min.

PAGE 41

29 Measurements Viscoelastic changes on heating and coo ling were determined using single gap geometry in an AR2000 Advanced Rheometer (TA Instruments, New Castle, DE). Approximately 20 g of sample were placed in the sample chamber at 5C and the head set to a specified gap (4950 m). After th e head reaches the gap, excess sample was removed with a stainless steel spatula and a layer of mineral oil wa s placed on top of the sample to prevent evaporation on heating. The opening was covered with a metal moisture trap also to prevent evaporation. The samples were heated from 5 to 80 C at a rate of 2 C/min and cooled from 80 to 5 C at the same rate and measurement conducted using a oscillatory mode with constant frequency set at 0. 1 Hz and maximum strain at 0.01( 8 ). Gel Preparation Protein isolates were adjusted to 83 % moisture content by squeezing the water manually out of the isolates. The washed muscle was centrifuged at 10,000 G for 20 min in a RC-5B centrifuge using a GS-3 rotor to re ach the right moisture content. Moisture was determined as described before. Appr oximately 130 g of isolate/washed muscle were accurately weighed and placed in a Mini Chopper (Sunbeam Products Inc., Boca Raton, FL) with 25 mM Sodium Phosphate Di basic buffer added after blending for 20 sec and 2% NaCl added after ~1 min (when appli cable). The pH was adjusted to 7.1 to 7.2 by adding 2M NaOH dropwise. The paste was mixed for a total of 4 min with all steps performed in a cold room at 4C. The paste was then manually stuffed into steel tubes (diameter 19 mm) and the ends sealed with a rubber cap and fastened with a hose clamp. The paste was cooked for 30 min at 80C in a Precision water bath (Precision Scientific, Winchester, VA) and cooled in ice water for 15 min. Gels were removed from the

PAGE 42

30 stainless steel tubes and placed in zip-lock ba gs and stored in a cold room at 4C for 48 hours prior to testing. Gel Quality Analysis Torsion Testing After storing the gels at 4 C for 48 hours they were cut into 28.7 mm long samples using a cutting motive. After reaching room te mperature (~40 min) the gels were milled into a dumbbell shape with a minimum center diam eter of 1.0 cm. Gels were tested using a modified Brookfield DV-II viscometer (Br ookfield Engineering Laboratories, Inc., Stoughton, MA) or (Gel Consultants, Raleigh, NC) and the samples twisted at 2.5 rpm until structure failed. Shear stress (resistance to breakage) and shear strain (distance until breakage) of the gels were obtained using co mputer software linke d to the viscometer. A B

PAGE 43

31 C D Figure 3-3. M odified Brookfield DV-II viscometer. A) Brookfield Engineering Laboratories, Inc., Stoughton, MA; or Gel Consultants, Raleigh, NC. B to D) M illing equipment. Folding Folding test was performed within 60 hour s of storage at 4C, according to the method of Kudo et al. ( 52 ). Approximately 3 mm slices were cut and folded by hand at room temperature. The ability of the gels to fold was graded usi ng a five point system (Table 3-1). Table 3-1. Grading system based on five-point system. Grade Description 5 No crack occurs even if folded in four 4 No crack when folded in two but forms a crack when folded in four 3 No crack when folded in two but splits when folded in four 2 Cracks when folded in two 1 Splits when folded in two Water Holding Capacity (WHC) Water holding capacity on cooking was dete rmined by analyzing the moisture content of the paste before cooking and mois ture content of the gels after cooking.

PAGE 44

32 Moisture content was determined based on we ight before and after drying the samples overnight at 106C in an oven. WHC of the gels on pressing was determined using the method of Feng ( 36 ) where pressing loss is defined as the water lo ss of a 3 mm thick slice under 3000 G pressure (using a 3L beaker full of water) for 1 mi n. The sample was sandwiched between five layers of Whatman filter paper (Whatman Inc., Clifton, NJ) which absorbed the expressible water. Weight before and afte r pressing was recorded and moisture content was determined using an oven as described before. 100 (g) Weight pressed Pre (g) weight pressed After (g) weight pressed Pre (%) Water e Expressibl 100 (g) Sample Pressed Pre of Moisture Total (g) Content Water e Expressibl (%) WHC Sulfhydryl Content Total sulfhydryl (SH) cont ent was determined on the paste before cooking and on the gels after cooking by using the method of Choi and Park ( 10 ) with slight modifications. The pastes and gels were dilu ted 100 times to give a protein concentration between 1 and 2 mg/mL. A 0.25 mL sample of the protein solution was added to 2.5 mL of 8 M urea, 2% sodium dodecylsulfate (SDS) and 10 mM EDTA in 0.2 M Tri-HCl buffer at pH 7.1. To this solution 50 uL of 10 mM Ellmans reagent (10 mM 5,5-dithiobis (2-nitrobenzoic acid) was added, mixed and heated in a water bath at 40C for 15 min. After the reaction the absorbance of the so lution was measured at 420 nm using an Agilent 8453 UV-VIS spectrophotometer (Agilent Technologies, Palo Alto, CA) and total SH content was calculated using a mola r extinction coefficient of 13600 mol/cm.

PAGE 45

33 Statistics Results are expressed as means SD. Analysis of variance (ANOVA) was used to determine significant differences (p < 0.05) between treatments. Transformation using arcsin was used for precipitation solubility All results were analyzed by using the Statistical Analysis System (SAS).

PAGE 46

34 CHAPTER 4 SOLUBILIZATION OF MUSCLE PRO TEINS OVER WIDE PH RANGE Solubility The basis for utilizing the acid and alkali processes on tilapia white muscle is to solubilize the muscle proteins at low and high pH. In or der to achieve high protein recovery it is important to obt ain high solubility at certain pH values to separate the soluble proteins from undesirable constituents of muscle. Low solubility at certain pH values is also important to precipitate (i.e., r ecover) the solubilized proteins. Therefore, to understand how the muscle proteins react to di fferent pH values a solubility curve was constructed ranging from pH 1.5 to 12.0 and changes in percent soluble protein were observed (Figure 4-1). Minimum protein sol ubility was observed between pH 5.0 and 6.0, which is in the range of the isoelectric point for the majority of proteins in muscle ( 2 ). Lowering the pH away from the isoelectric po int resulted in a dramatic increase in protein solubility up to pH 4.0 where the proteins were ~94% soluble. Maximum solubility at low pH was at pH 2.5 or ~96% but decreased dow n to 90% at pH 1.5. This decrease could possibly be due to anion induced aggrega tion, since more HCl would increase the ionic strength of the soluti on and may reduce some of the electrostatic repulsion between the proteins ( 53;54 ). An increase in pH away from the isoelectric point resulted in a slight increase in protei n solubility until pH 9.0 to 10.0 where it rapidly increased between pH 10.0 and 11.0; and reached a maximum ~99% at pH 12.0. This solubility curve is similar to what Undeland et al ( 55 ) observed for herring and Kim et al ( 56 ) observed for Pacific whiting.

PAGE 47

35 0 20 40 60 80 100 02468101214pHProtein Solubility (%) Figure 4-1. Shows protein sol ubility (%) of tilapia light muscle homogenate at pH 1.5 to 12.0. Protein content was determined using the Biuret method with the addition of 10% deoxycholic acid usi ng a BSA standard curve. Protein solubility was defined as the fractio n of the total protein soluble after centrifugation, assuming no protein was lo st in the top laye r or sediment. Results are mean SD. As Figure 4-1 shows, approximately 20% of the proteins were soluble at physiological pH (6.5 to 7.0) at the 10 fold d ilution commonly used in acid and alkali processes. Sarcoplasmic proteins which are so luble in water or solutions of low ionic strength usually constitute a bout 20 to 30% of fish muscle (57). The proteins which were soluble in the isoelectric point range were mostly sarcoplasmic proteins along with some dissociated actin (Figure 6-7). At the isoelectric pH (5.0 to 6.0) where minimum solubility was observed, lack of electro static repulsion between charge d residues results in closer contact of the proteins which promotes hydrophobic interactions (protein-protein interaction) and aggregation. However, at pH below or above isoelctric pH, the proteins become positively or negatively charged which results in electrostatic repulsion between

PAGE 48

36 the molecules and hydration of charged residues which in turn promotes solubility of the proteins ( 30 ). Mechanism of solubiliza tion of myofibrillar protei ns is believed to take place in two steps; first depolymerization of the thick filaments and then dissociation of actin from myosin ( 58 ). Addition of acid or base wh ich raises ionic strength is not believed to have significant e ffects on protein solubility as suggested by Undeland et al. ( 9 ). The rapid increase in solubility at low pH compared to a more steady increase at higher pH might be attributed to more ioni zable groups with pKa values between 2.5 and 7.0 than between 7.0 and 11.0 ( 6;9 ). Viscosity Low viscosity of the tilapia muscle homogena te is important in the acid and alkali process since it facilitates se paration of soluble protein fr om insoluble material via centrifugation. As an example, Undeland et al. ( 9 ) showed that total lipid removal was reduced with increasing viscos ity at low pH for homogenate prepared from herring light muscle. Low viscosity was also shown to be important to separate membrane phospholipids from muscle proteins, which could lead to increased oxidative stability of the final protein isolate ( 6;9 ). A viscosity curve of homogenate prepar ed from tilapia white muscle was constructed in the pH range of 1.5 to 12.0 (Figure 4-2) in accordan ce to the solubility curve. The viscosity of the solubilization pHs of interest for the acid and alkali processes were low, or < 140 mPa*s. Two viscosity peak s were observed, at pH 5.0 and pH 9.5, respectively. At low pH, viscosity started to increase at pH 4.5, reaching a maximum ~2 Pa*s at pH 5.0. At high pH, the viscosity started to increase around pH 8.0 and reached a maximum ~ 10 Pa*s at pH 9.5. Homogenate prepared from catfish showed similar

PAGE 49

37 results ( 59 ) whereas Undeland and coworkers ( 55 ) observed similar viscosity peaks at low and high pH for herring light muscle homogenate. 0.000 2.000 4.000 6.000 8.000 10.000 12.000 0.02.04.06.08.010.012.014.0pHViscosity (Pa*s) Figure 4-2. Shows viscosity (Pa*s) of tilapi a light muscle homogenate at pH 1.5 to pH 12.0 in 0.5 increments. Viscosity was measured using oscillatory time sweep in an AR2000 Rheometer. Results are mean SD. The large peak observed at pH 9.5 could be attributed to greater water binding capacities of proteins at pH 9.0 to 10.0 than at any other pH as a resu lt of ionization of sulfhydryl and tyrosine residues ( 30 ). Interactions between hydrated residues are short range repulsive interactions that become stronger with increasi ngly hydrated molecules ( 34 ). This might contribute to increased hydrodynamic size of the proteins which in turn leads to an increase in viscosity. Decreased water binding above pH 10 could be due part ly to loss of positively charged -amino groups of lysyl residues and incr eased solubility of the proteins ( 30 ). Lower viscosity could also be contributed to large myofibrillar assemblies, which have a high hydrodynamic volume, but when they break up and the effective hydrodynamic volume significantly decreases. At pH 5.0 whic h is close to the is oelectric pH of the proteins a smaller peak was observed. Movi ng away from the isoel ectric point increases

PAGE 50

38 net charge and repulsive forces resulting in facilitation of the prot eins to bind water and swell. Even just below to the isoelectric pH, proteins carry a ne t positive charge which results in repulsion and hydrat ion of residues that could increase the hydrodynamic size of the proteins and cause an increase in viscosity ( 30 ). The drop in viscosity after pH 5.0 could be related to solubilization of the pr oteins which would re sult in lowering of viscosity. Figure 4-3 shows that at low pH, a peak in viscosity was observed just before a dramatic increase in solubility was observed. At high pH solubility was more gradually increasing with increasing pH and viscosity c ould not be directly re lated to increase in viscosity. 0 10 20 30 40 50 60 70 80 90 100 02468101214 pHProtein Solubility (%)0.0 2.0 4.0 6.0 8.0 10.0 12.0Viscosity (Pa*s) Solubility Viscosity Figure 4-3. Shows solubility and viscosity curves previously demonstrated on the same graph. Results are mean SD.

PAGE 51

39 CHAPTER 5 SOLUBILIZATION OF MUSCLE PROTEINS AT LOW AND HIGH PH BEFORE 1ST CENTRIFUGATION Effect of Homogenization Times It has been reported that di fferent homogenization times may affect muscle protein solubility ( 29 ) which could in turn lead to variations in protein recovery. For this reason, effects of different homogenization times (60, 90, 120 sec) on protein solubility (%) were investigated at solubilization pHs of interest to the acid (pH 2.3 to 2. 9) (Figure 5-1 A) and alkaline process (pH 10.8 to 11.4) (Figure 5-1 B). Overall, increasing the homogenization time did not have considerable effects on protein solubility. Longer homogenization times did not seem to result in lower solubility as Stefansson and Hultin ( 29 ) observed when homogenizing cod muscle for 120 sec. At low pH, using 90 sec at pH 2.3 led to significantly lower solubility (p < 0.01) comp ared to other pH values and homogenization times. For high pH, using 120 sec at pH 11.2 le d to significantly higher protein solubility than was seen for other treatments (p < 0.01). The value obtained was above 100%, which could be in part be reflected by expe rimental error or deviation. Based on these results, further studies (solubility, vi scosity and recovery) were performed by homogenizing the tilapia white muscle for 60 sec because of good stability in solubility and minimum foaming. To minimize the ex tent of protein dena turation due to shear force or temperature increase, which in turn leads to foaming (and le ss protein recovery), the tilapia muscle was homogenized in two steps of 30 sec at 4C.

PAGE 52

40 0 20 40 60 80 100 120 2.22.32.42.52.62.72.82.93.0pHProtein Solubility (%) 60 sec 90 sec 120 sec(A) 0 20 40 60 80 100 120 10.710.810.911.011.111.211.311.411.5pHProtein Solubility (%) 60 sec 90 sec 120 sec(B) Figure 5-1. Effect of different homogeniz ation times (60, 90 and 120 sec) on protein solubility (%) at low (A) and high (B ) pH. The tilapia white muscle was homogenized at 4 C for 30 sec at a time with a 30 sec rest time in between. Results are mean SD. Viscosity Examination of viscosity for low and high solubilization pHs showed that at low pH viscosity was significantly higher (p < 0.05) for pH 2.9 compared to pH 2.3 and 2.5 (Figure 5-2). At high solubilization pH ther e was no significant di fference (p > 0.05) in

PAGE 53

41 viscosity (Figure 5-3). Lower viscosity was observed for certain hi gh pH values over the range tested compared to viscos ity at low pH. Ther e was also more vari ation in viscosity at low pH compared to high pH. Adjustme nt to low pH (2.3 to 2.9) after storage of homogenate over night resulted in a drama tic lowering of viscosity, down to similar values as were seen for alkali solubilized pr otein (data not shown). This decrease could possibly have been attributed to limited hydrolysis which was observed at low pH (Figures 6-8 and 6-9, lanes 3 to 4). -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.22.42.62.83.0pHViscosity (Pa*s) Figure 5-2. Shows viscosity (Pa*s) of tilapia homogenate at low solubilization pH (2.3 to 2.9) determined with oscillatory tim e sweep with controlled stress in an AR2000 Rheometer. Results are mean SD.

PAGE 54

42 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 10.610.811.011.211.411.6pHViscosity (Pa*s) Figure 5-3. Shows viscosity (Pa*s) for ti lapia homogenate at high solubilization pH (10.8-11.4) determined with oscillatory time sweep with controlled stress in an AR2000 Rheometer. Results are mean SD. At extreme low and high pH the proteins are partially denatured. Partial denaturation of proteins can increase thei r hydrodynamic size and thus result in an increase in viscosity ( 33 ). It has been suggested that pr oteins at extreme low pH are more extensively denatured than pr oteins at extreme high pH ( 49 ) and work on herring light muscle homogenate ( 55 ) and cod myosin ( 49 ) supported that suggestion. Kristinsson and Hultins work with myosin suggested that at pH 2.5 the myosin heavy chains were dissociated whereas at pH 11.0 the heavy chai ns were still associated, which led to a higher viscosity of myosin at low pH co mpared to high pH. Increased hydrodynamic volume of the myosin molecule at low pH c ould contribute to the high viscosity observed at low pH. When a protein unfolds, hydr ophobic groups buried in the in terior of a protein are exposed to the water surrounding them. The wa ter tries to minimize contact to the nonpolar groups resulting in strong attractive forces between non-polar groups ( 60 ). The

PAGE 55

43 proteins could form aggregates or other el ongated structural assemblies and resulting in an increase in their hydrodynamic volume ( 34 ) resulting in higher viscosity at pH 2.9 compared to pH 2.3. With increasing charge these forces are reduced and the viscosity is lowered as was seen for pH 2.3. The large variation in viscosity at low pH was interesting to note. When proteins are denatured at low or high pH they can take on many different structural states, depending pH and other solution conditions ( 61 ). These different states may thus have different viscosities. It has also been hypothesized that at low pH proteins may take on more numerous structural states than at high pH, and there may be rapid conversions between one state to another ( 61 ). This could be one explanation between the great variations seen in viscosity at low pH in contrast to high pH Time might also contribute to the variability seen at low pH compared to high pH. Undeland and coworkers ( 55 ) showed that viscosity was reduced when aci dified herring homogena te was stored on ice for up to 25 min. More volume was needed of 2 M HCl than 2 M NaOH to adjust the pH. The chloride ion (1.81) has twi ce the radius of the sodium ion (0.98) and this size difference as well as larger volume could po ssibly contribute to increased volume of the proteins and therefore viscosity ( 60). The trend towards decrease in viscosity at low pH with decreasing pH could be due to dilution of the solution when 2 M HCl is added. Diluting the homogenate may result in decreased interactions between hydrated pr otein molecules and thei r ability to absorb water and swell which in turn results in a decrease in viscosity ( 33 ).

PAGE 56

44 Recovery of Soluble Proteins The amount of proteins recovered at lo w and high pH was studied by comparing two acid solubilization pHs (2.5 and 2.9) a nd two alkali solubilization pHs (11.0 and 11.2). The basis for the selection of these solubilization pHs was as follows: pH 10.8 was very difficult to work with due to instab ility, pH 11.4 was considered too high due to amount of base needed to adjust the pH a nd possible deamination of the proteins (e.g. indicated by the smell). For the low pH sol ubilization pHs, pH 2.5 and 2.9 were selected because one of the most widely used pH in the acid aided process is pH 2.5 ( 10;18 ) and pH 2.9 has been used with good results in other labs ( 62 ), pH 2.3 was excluded due to very large quantities of acid needed to lower the pH ( 55;56 ). In Figure 5-4, results are summarized and e xpressed as theoretical and actual protein recovery (%) after 1st centrifugation based on total pr otein in the homogenate. The difference between theoretical and actual rec overy lies in difference in calculation and representation of the results. Theoretical pr otein recoveries are calculated based on the concentration of protein pres ent in the supernatant after centrifugation compared to before centrifugation. On the other hand, actua l recovery takes into account proteins that are lost in the separation proce ss of soluble proteins from inso luble material, e.g. proteins lost in the fat layer and in the cheesecloth th at covers the strainer used for separation. The bottom layer containing the insoluble materi al can be reprocessed to recover some of the proteins that are lost dur ing separation especially when using low solubilization pH ( 63 ). After the 1st centrifugation, actual recovery for alkali treated proteins was significantly higher (p < 0.01). Theoretical recovery of proteins was higher than actual protein recovery and ranged from 89% for pH 11.0 to 95% for pH 11.2. Actual recovery

PAGE 57

45 ranged from 62% for pH 2.5 to 73% for pH 11.2. In regard to actual recovery, no significant difference was observed within the acid or alkali treated proteins. There was no significant difference (p > 0.05) in regard to theoretical recovery between low and high solubilization pH. A difference between the acid and alkali treated proteins was observed using electrophoretic separation of the proteins. Hydrolysis of the myosin h eavy chain (~205 kD) was observed of supernatants at pH 2.5 (F igure 5-8, lanes 3 to 4) and 2.9 (Figure 5-9, lanes 3 to 4) after 1st centrifugation. This was not seen for the supernatants recovered after 1st centrifugation for solubilizat ion pHs 11.0 (Figure 5-10, la nes 3 to 4) and 11.2 (Figure 5-11, lanes 3 to 4). Apart from that simila r polypeptides were recovered after 1st centrifugation using differe nt solubilization pHs. 89.8 91.888.7 94.5 73.0 71.4 63.8 61.9 0 20 40 60 80 100 2.52.91111.2 Solubilization pHProtein Recovered (%) Theoretical Recovery Actual Recovery Figure 5-4. Theoretical and actual % protein recovery afte r 1st centrifuga tion for four different solubilization pHs (2.5, 2. 9, 11.0, 11.2). Results are mean SD. Different letters indicate significant difference (p < 0.05) between treatments. aaaabbcc

PAGE 58

46 CHAPTER 6 PRECIPITATION OF MUSCLE PROTEINS Solubility Soluble proteins which are se parated from neutral fat and insoluble material in 1st centrifugation are subjected to is oelectric precipitation in the acid and alkali processes to recover the proteins ( 1 ). Therefore, it is of great interest to investigate how different solubilization pHs respond to various precipitation pHs and to determine the combination which gives the highest recovery. Based on resu lts from the solubility curve (Figure 3-1) where minimum solubility was observed at pH 5.0-6.0, four pHs (pH 5.1, 5.3, 5.5, and 5.7) were selected to pr ecipitate the proteins. Precipitation of proteins sol ubilized at low pH values re sulted in significantly (p < 0.01) lower amount of soluble protein after 2nd centrifugation (Figur e 6-1) compared to high solubilizing pH values. This is in ag reement with results for catfish reported by Kristinsson and Demir ( 18 ). At high solubilization pH (11.0 and 11.2), the use of precipitation pH 5.5 and 5.7 was not significantly different (p > 0.05) Proteins precipitated at pH 5.1 containe d more soluble protein after 2nd centrifugation for all treatments except for pH 11.0 probably becau se pH 5.1 is further away from the isoelectric point. The large difference seen between low and high solubilization pHs in the precipitation is likely to be partly explaine d by different degree of denaturation of the proteins at low and high pH a nd then subsequently different degree of refolding as pH is readjusted to pH 5.1 to 5.7. A more exte nsively denatured protein has more hydrophobic

PAGE 59

47 areas exposed and is better able to form mo re and stronger proteinprotein interactions. Work with trout hemoglobin by Kristinsson and Hultin ( 64 ) demonstrated that the protein was more extensively denature d at low pH compared to high pH, and was also less able to refold on pH readjustment to 5.5 and 7, leaving more expos ed hydrophobic groups. This led to substantially more protein aggr egation for hemoglobin refolded from low pH compared to high pH, and thus less hemoglobi n remained in solution after centrifugation, in accordance with that seen here. 7.7 7.1 6.9 6.6 8.2 6.7 6.2 6.4 15.7 14.3 13.413.4 13.4 13.4 16.0 14.30 2 4 6 8 10 12 14 16 18 5.15.35.55.7 Precipitation pHProtein Solubility (%) pH 2.5 pH 2.9 pH 11.0 pH 11.2 Figure 6-1. Soluble protein (%) in supernat ant after the 2nd centrif ugation, using four different precipitation pHs. Protein solubility was defined as the fraction of the total protein soluble after centrifugation, assuming no protein was lost. Results are mean SD. Different letters in dicate significant difference (p < 0.05). Viscosity Differences in viscosity between low a nd high pH using various precipitation pHs before 2nd centrifugation were not as evident as was observed in solubility after precipitation (Figure 6-1). This indicated that the differences seen in solubility after a bc d e d c f f g hj j j j k

PAGE 60

48 precipitation were not necessarily due to diffe rences in viscosity. However, pH 2.9 did seem to give significantly higher viscosity at precipitation pHs 5.1 and 5.3 (p < 0.05). This difference did not reflect the results seen for precipitation solubility. Viscosity of precipitated proteins solubilized at pH 2.9 formed larger un its than when solubilized at the other pH values, presumably due to di fferent protein-protei n interactions which resulted from unfolding and refolding for that pH treatment. The viscosity values after readjusting the proteins to pH 5.1 to 5.7 resulted in an enormous increase in viscosity values compared to native protein at similar pH values (Figure 3-2). The reason for this dramatic increase is probably due to the fact that the proteins have not refolded to their original state leading to increased interacti on and aggregation, lead ing to an increase in hydrodynamic volume and therefore increase in viscosity. These changes were also observed in a study performed on acidified a nd alkalized sarcoplasmic proteins from herring white muscle did not show an increase in viscosity indicating that the myofibrillar proteins are probably responsible for this changes in viscosity ( 55 ). Then it can be concluded that the use of different solubilization pH (low vs. high) has more impact on protein conformation than the use of different precipitation pHs.

PAGE 61

49 0 50 100 150 200 250 300 5.15.35.55.7 Precipitation pHViscosity (Pa*s) pH 2.5 pH 2.9 pH 11.0 pH 11.2 Figure 6-2. Viscosity, (Pa*s) of tilapia mu scle proteins at diffe rent precipitation pH before the 2nd centrifugation. Tilapia white muscle proteins were solubilized using four different pHs (2.5, 2.9, 11.0, and 11.2) and precipitated using four different pHs (5.1, 5.3, 5.5, and 5.7). Results are mean SD. Different letters WITHIN each precipitation pH indica te significant difference (p < 0.05). a a a a a a a a a b b b b a b b

PAGE 62

50 0 50 100 150 200 250 300 2.52.91111.2 Precipitation pHViscosity (Pa*s) pH 5.1 pH 5.3 pH 5.5 pH 5.7 Figure 6-3. Shows the same as Figure 6-2, but in a different arra ngement, categorized based on solubilization pH not preci pitation pH. Results are mean SD. Different letters WITHIN each solu bilization pH indicate significant difference (p < 0.05). Protein Recovery Between 1st and 2nd Centrifugation Protein recovery after 2nd centrifugation using differe nt precipitation pHs was expressed as proteins recovered after 2nd centrifugation as a fraction of total soluble protein after 1st centrifugation. Theoretical and actual protein recovery (%) were calculated (Figure 6-4). Opposite to what was seen after 1st centrifugation, recoveries were higher for the acid treat ed proteins, or 92 to 94% wher eas recovery for the alkali treated proteins ranged from 84 to 89%. These data are in direct agreem ent with the data on the amount of soluble proteins after precipitation. Where the proteins treated at low pH resulted in lower amount of soluble protein in the supernatant after 2nd centrifugation compared to proteins treated at high pH, indicating higher am ount of protein recovered in the protein isolate. The reason for the highe r level of precipitation can be explained as a a a a a b ab a a ab b ab a ab b ab

PAGE 63

51 before that the proteins adju sted to low pH were more extensively denatured and less refolded and thus had a higher level of aggregation The use of different precipitation pH therefore does not seem to have impact on protein recoveries since similar percentages are observed for all precipitation pHs. This finding indicates that selecting a specific precipitation pH is not critical for tilapia musc le when using the acid and alkali processes. 93.4 94.194.2 94.3 92.1 93.0 93.2 93.1 70 75 80 85 90 95 100 5.15.35.55.7 Precipitation pHProtein Recovery (%) Actual Recovery Theoretical Recovery A. 93.8 92.5 93.2 93.793.6 94.4 94.694.5 70 75 80 85 90 95 100 5.15.35.55.7 Precipitation pHProtein Recovery (%) Actual Recovery Theoretical RecoveryB. 86.4 85.8 86.9 88.287.6 87.7 85.8 84.8 70 75 80 85 90 95 100 5.15.35.55.7 Precipitation pHProtein Recovery (%) Actual Recovery Theoretical RecoveryC. 89.1 87.0 88.1 86.7 85 87 86 84 70 75 80 85 90 95 100 5.15.35.55.7 Precipitation pHProtein Recovery (%) Actual Recovery Theoretical RecoveryD. Figure 6-4. White muscle tilapia proteins were solubilized at pH 2.5 (A), 2.9 (B), 11.0 (C) and 11.2 (D) and precipitated at four different pHs (5.1, 5.3, 5.5, and 5.7). Theoretical and actual pr otein recovery (%) after 2nd centrifugation were expressed as the percenta ge of total soluble protein recovered after 1st centrifugation. Different capital le tters indicate significant difference (p < 0.05) for actual recovery. Different small letters indicate a significant difference for theoretical recovery. aA A A A aaaaA B AB AB A B AB B A A A A aaaababababa b

PAGE 64

52 Total Protein Recovery Protein recoveries through the whole process were dete rmined for all treatments including recoveries from washed muscle which was used as a control. The washed muscle mimics the current industry standard of making surimi. Figure 6-55 summarizes the results and showed that ac tual recoveries were lower th an theoretical recoveries, as would be expected. No one treatment gave st atistically significantly (>0.05) better results than other treatments. Theoretical rec overies ranged from 83% to 88% for low solubilization pH but more va riability was observed at high solubilization pH where it ranged from 75 to 85%. This was al so observed for herring light muscle (9 ) as well as catfish, mackerel, mullet and croaker ( 18 ). Lower recoveries (theoretical) using the acidic treated proteins could be explained by larger amount of proteins lost in the 1st centrifugation. Kristinsson and Hultin ( 8 ) found that emulsification ability of pH-treated treated proteins was significan tly improved compared to untreated proteins due to higher hydrophobicity. The acid treated proteins may t hus have emulsified with lipids in the tilapia and therefore less was r ecovered on the first centrifugation. Actual recoveries for high solubilization pH were on the other hand slightly higher when compared to acidified proteins, or ranging from 61 to 68% compared to 56 to 61%, respectively. There were however no stat istically significant (p > 0.05) differences between low and high solubilizati on pHs. That could be expl ained by the fact that after 1st centrifugation more proteins are lost duri ng the acidic treatment than alkali but after 2nd centrifugation more proteins are lost in the alkali treatment which evens out the difference. Protein recoveries from washed fish muscle were 65.1% which was similar to the results obtained for high solubilization pH s. Preparation of wa shed muscle involves

PAGE 65

53 washing most of the sarcoplasmic proteins ( 2 ) possibly along with some of the myofibrillar proteins ( 65 ). 56.4 56.8 57.7 57.8 83.7 83.6 83.7 83.0 0 10 20 30 40 50 60 70 80 90 100 5.15.35.55.7Precipitation pHProtein Recovery (%) Actual Recovery Theoretical Recovery A. 58.9 59.3 61.361.2 83.0 83.7 87.8 87.9 0 10 20 30 40 50 60 70 80 90 100 5.15.35.55.7Precipitation pHProtein Recovery (%) Actual Recovery Theoretical RecoveryB. 68.0 84.3 62.1 62.6 62.0 76.5 74.9 75.0 0 10 20 30 40 50 60 70 80 90 100 5.15.35.55.7Precipitation pHProtein Recovery (%) Actual Recovery Theoretical Recovery C. 61.3 62.3 67.0 65.3 84.8 81.2 79.2 77.5 0 10 20 30 40 50 60 70 80 90 100 5.15.35.55.7Precipitation pHProtein Recovery (%) Actual Recovery Theoretical RecoveryD. Figure 6-5. White muscle tila pia proteins were solubilized at pH 2.5 (A), 2.9 (B), 11.0 (C) and 11.2 (D) and precipit ated at 4 different pHs 5. 1, 5.3, 5.5 and 5.7. Theoretical and actual protein recovery (%) in final protein isolate was based on total protein in initial ho mogenate. Results are mean SD. Different capital letters indicate significant diffe rence (p < 0.05) for actual recovery. Different small letters indicate a signifi cant difference (p < 0.05) for theoretical recovery. An SDS-PAGE analysis was performed on samples collected during washing of the muscle (Figure 6-6). The final washed muscle (lane 2) showed that the washed muscle mainly contained protein bands tentatively identified as myosin heavy chain or MHC (~205 kD), actin (~43 kD) and possibly some tropomyosin (35.5 kDa). During washing AAAAA A A A A A B B A A B B aaaaaaaaaabbaaaa

PAGE 66

54 of the muscle a large amount of actin was washed out a long w ith proteins with tentative weights; ~100 kDa, ~69 kDa, ~56 kDa (desmin), ~41 kDa (troponin T), ~38 kDa (tropomyosin-beta) and ~17 kDa (unidentified). 123 456 7 Figure 6-6. SDS page for wash ed tilapia muscle. Lane 1: sigma wide marker, lane 2: washed muscle after 3 washes, lane 3: 1st wash, lane 4: 2nd wash, lane 5: 3rd wash (0.2% NaCl), lane 6: sk ip, lane 7: initial sample. 1 4 10 6 5 23789 Figure 6-7. Precipitation of prot eins at pH 5.1-5.7 without so lubilization. Lane 1: Sigma wide marker, 2: Initial homogenate, lane s 3-6: Protein isol ate precipitated at pHs 5.1, 5.3, 5.5, 5.7, lanes 5 to 7: supernatants after centrifugation at pHs 5.1, 5.3, 5.5, and 5.7.

PAGE 67

55 1 2 3 45 6 78 9 10 11 12 Figure 6-8. Lane 1: Wide Sigma Marker, lane 2: Initial homogenate at pH 2.5, lane 3: supernatant at pH 2.5 after 1st centrifugation, lane 4: supernatant at pH 2.5 after 1st centrifugation, lanes 6-8: isolates precipitated at pH 5.1, 5.3, 5.5 and 5.7, respectively. Lanes 9-12: S upernatants after precipitation and 2nd centrifugation at pH 5.1, 5.3, 5.5, 5.7. 1 2 3 4 5 6 7 89 10 11 12 Figure 6-9. Lane 1: Wide Sigma Marker, lane 2: Initial homogenate at pH 2.9, lane 3: supernatant at pH 2.9 after 1st centrifugation (used for precipitation pH 5.1 & 5.3), lane 4: supernatant at pH 2.9 after 1st centrifugation (used for precipitation pH 5.5 & 5.7), lanes 6-8: is olates precipitated at pH 5.1, 5.3, 5.5 and 5.7, respectively. Lanes 9-12: Supernatants after precipitation and 2nd centrifugation at pH 5.1, 5.3, 5.5, 5.7.

PAGE 68

56 1 2 3 45 678 10 11 12 9 Figure 6-10. Lane 1: Wide Sigma Marker, lane 2: Initial homogenate at pH 11.0, lane 3: supernatant at pH 11.0 after 1st centri fugation (used for pr ecipitation pH 5.1 & 5.3), lane 4: supernatant at pH 11.0 after 1st centri fugation (used for precipitation pH 5.5 and 5.7), lanes 6-8: is olates precipitated at pH 5.1, 5.3, 5.5, and 5.7, respectively. Lanes 9-12: Su pernatants after precipitation and 2nd centrifugation at pH 5.1, 5.3, 5.5, and 5.7. 1 234 5 6 7 9 8 10 11 12 Figure 6-11. Lane 1: Wide Sigma Marker, lane 2: Initial homogenate at pH 11.2, lane 3: supernatant at pH 11.2 after 1st centrifugation (used for precipitation pH 5.1 and 5.3), lane 4: supernatant at pH 11.2 after 1st centrifugation (used for precipitation pH 5.5 and 5.7), lanes 6-8: is olates precipitated at pH 5.1, 5.3, 5.5, and 5.7, respectively. Lanes 9-12: Supernatants after precipitation and 2nd centrifugation at pH 5.1, 5.3, 5.5, and 5.7.

PAGE 69

57 CHAPTER 7 GELATION Gel forming ability of tilapia white muscle protein isolates prepared with the acid and alkali processes were determined and co mpared to washed tilapia white muscle (analog to conventional surimi processing). Gel characteri stics were determined using torsion, rheology (viscoelastic changes), fold ing and water holding capacity. Gels were prepared with and without addition of 2% Na Cl (w/w) at pH 7.1 to 7.2. These are common gelation conditions for fish muscle gels. Adjus ting the pH of the prot ein isolates from pH 5.5 to slightly above neutrality leads to in creases electrostatic re pulsion between proteins, thus giving a more even dist ribution of proteins in the gel matrix resulting in good gel quality ( 36;66 ). The use of 2.0% NaCl (w/w) is commonly used in gelation since it partially solubilizes the muscle proteins. Previous work with washed tilapia muscle has shown higher strength and deformation of gels prepared with 2% NaCl (w/w) or higher compared to gels with 0.5 to 1.5% NaCl added ( 24 ). Recently, salt free muscle protein gels at pHs above neutrality have also been found to have good gelation properties (8 ), which made it of interest to study these conditions for the tilapia muscle proteins. Torsion In Figure 7-1, it can be seen that the shear st ress (i.e., resistance to breakage) of gels prepared using 2% NaCl (w/w) was significantly higher (p < 0.01) than that seen for gels without added NaCl, except for gels prepared fr om washed muscle. For the gels with 2% added salt, the lowest shear stress, thus l east resistance to breakage, was obtained for washed muscle gels (32.1 4.3 kPa). Gels fr om isolates made using the pH 2.5, 11.0, and

PAGE 70

58 11.2 treatments gave the highe st shear stress values, ranging from 80 to 84 kPa. There was however no significant difference between these gels (p > 0.05). For the gels without added salt, the highest stress value (69.1 12. 0) was obtained for the isolates made using pH 11.0 treatment while the lowest stress valu e was obtained for the isolates made using pH 2.9 treatment and the washing procedure (30.4 7.0 and 26.0 2.0, respectively. 80.2 32.1 83.6 58.7 77.9 26.0 47.5 69.1 30.4 45.2 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 2.52.911.011.2Washed muscle TreatmentShear stress (kPa) with 2% NaCl Without 2% salta b a a c A B CA B Figure 7-1. Shows shear stress values (kPa) of gels produced from white muscle proteins of tilapia. The use of low and high solubilization pH treatment was compared to a three cycle washing treatment (contro l). Gels were cooked in steel tubes at 80C for 30 min. The gels were st ored in a cold room at 4 C for 48 hours prior to testing with a Torsion Gelometer. Results are mean SD. Different capital letters indicate significant diffe rence (p < 0.01) for treatments without 2% NaCl. Different small letters indicate a significant difference for treatments with 2% NaCl. For each treatment gels with 2% NaCl (w/w) had a significantly higher stress value, except for washed muscle. Resistance of the gels to deformation or sh ear strain showed that the addition of 2% NaCl (w/w) resulted in significantly higher (p < 0.01) strain values for all treatments compared to gels without added NaCl (Fi gure 7-2). For samples without added salt, isolates made using pH 11.0 and 11.2 treatments gave significantly (p < 0.05) higher strain

PAGE 71

59 values (1.6 0.1 and 1.5 0. 2, respectively) compared acid treated proteins and washed muscle. On the other hand, isolates made using pH 2.9 and the washed muscle had the lowest shear strain values (1.1 0.1 and 1.2 0.1, respectively), which is in line with the results seen for shear stress. If isolates ma de using low pH treatment are compared it is clear that pH 2.5 treatment with and without 2% NaCl (w/w) gave significantly higher strain values than pH 2.9 treatment, thus being significantl y more elastic. The strain values obtained for the treatme nts were lower or borderline for the ideal shear strain values for surimi from fish mu scle, which are expected to be between 2 and 3. However, the reason for lower values could be explained by the absence of cryoprotectants and setting which are used in surimi production. Bakir et al. ( 67 ) observed that gels prepared from Atlan tic mackerel and Bluefish without using cryoprotectants resulted in significantly lower st rain values possibly due to the preventive action towards the proteins during heati ng which could result in improved gelation ability. Shear stress and strain values were signi ficantly higher for samples with 2% added NaCl (w/w). Salt is believed to improve gelation ability and water retention by solubilizing myofibrillar proteins ( 2 ). Addition of salt above 300 mM (~1%) solubilizes the myofibrillar proteins by breaking up the in teractions between myosin in the thick filament and actin in the thin filaments, along with other cytoskeletal proteins ( 68 ). In a concentrated gel paste however, the osmotic pressure is probabl y too high to obtain completed solubilization of the myofibrillar proteins. Uniform dispersion of partially solubilized proteins is likely to be more important in a gel paste. Addition of salt is also

PAGE 72

60 believed to contribute to a more elastic gel by dispersing the proteins more evenly, which is a consequence of partial or full solubilization. 2.0 1.7 1.9 1.6 1.9 1.2 1.5 1.6 1.1 1.3 0.0 0.5 1.0 1.5 2.0 2.5 2.52.911.011.2Washed muscle TreatmentShear strain with 2% NaCl Without 2% saltab a a b D C C B A Figure 7-2. Shows shear strain values of gels produced from white muscle of tilapia The use of low and high solubilization pH treatment was compared to a three cycle washing treatment (cont rol). Gels were cooked in steel tubes at 80C for 30min. The gels were stored in a cold room at 4C for 48 hours prior to testing with a Torsion Gelometer. Results are mean SD. Different capital letters indicate significa nt difference (p < 0.01) for treatments without 2% NaCl. Different small letters indicate a significant difference for treatments with 2% NaCl. For each treatment gels with 2% NaCl (w/w) had a significantly higher strain value when compared to gels without 2% NaCl (w/w). It was interesting to note the difference between the two low-pH treatments. A difference in gel performance between differe nt low pH treatments has been seen with other species. A study performed by Kim et al. ( 56 ) on Pacific whiting showed that deformation (mm) of gels (contained 1.5% b eef plasma protein) using a puncture test were higher for proteins treate d at pH 2.0 compared to pH 3.0. This was in part

PAGE 73

61 explained by the authors by increased hydropho bicity of the prot eins at low pH. Davenport and coworkers ( 69 ) also found a significant difference in gel forming ability among three different low pH treatment s which could on the other hand not be explained by hydrophobicity changes or ch anges in protein conformation using tryptophan as a structural probe. The pH 2.9 treatment in particular resulte d in the formation of weak gels. During isoelectric precipitation of proteins solubilized at pH 2. 9 it was observed that they seemed to form larger aggregates compared to pH 2.5 and pH 11.0. Larger aggregates do not form as ordered three dimensional structur es and thus form a weaker gel. To many protein-protein interactions might result in an hard and inelastic gel whereas to many protein-water interactions might result in a soft and fragile gel ( 63 ). Hydrolysis of the myosin heavy chain (Figures 6-8 and 6-9) wa s observed for low pH, which could in part explain the reduced gel forming ability of the protein isolates made with acid treatment. This is in agreement with results by Undeland et al. ( 9 ) who observed significantly higher stress values for gels prepared from alkali processed isolates (w ith cryoprotectants and 2% NaCl (w/w)) from herring light muscle co mpared to acid produced isolates. The authors suggested that reduced amount of myosin heavy chain (possibly due to proteolysis) contributed to lower stress values at acidic conditions. It is interesting to note however that the pH 2.5 treatment with t ilapia proteins was better than the pH 2.9 treatment. It is possible the proteases in quest ion were more active at pH 2.9 than pH 2.5. Differences in protein conformation between the two treatments may also be causing the difference, as discussed above.

PAGE 74

62 It is also worth noting that the washed muscle, acid isolat es and alkali isolates had different protein composition, which may contri bute to the different gel forming ability of these treatments (Figures 6-6 to 6-11). Interestingly the alkali process and the washing process removed similar protei ns but had significantly diffe rent gel forming abilities. The very low stress values obtained for the washed muscle compared to the acid and alkali treated proteins indi cates that conformational diffe rences contribute to these differences. Folding The fold test is a common test used in the surimi industry for a quick evaluation of gel quality. The fold test was performed on 3 mm slices of gels which were subjected to the strain and stress tests mentioned before. All treat ments with 2% added NaCl exhibited excellent folding ability and received the highest score available, or 5 (Table 7-1). Gels prepared using al kali treated proteins without 2% NaCl (w/w) exhibited excellent folding ability and were double folded without breaking. On the other hand, the salt free gels prepared using acid treated pr oteins and washed muscle performed very poorly. The lowest folding score was obtained fo r gels prepared from isolates made with pH 2.9 treatment which split in two during the first fold (score 1). Gels prepared from washed muscle and pH 2.5 cracked without sp litting during the first fold and received a score of 2.

PAGE 75

63 Table 7-1. Quality of protein gels as assessed by the fold test. Gel quality was estimated by folding approximately 3 mm thick ge l slices by hand at room temperature and grading the quality using a five point system., 5: No crack occurs even if folded in four, 4: No crack when folded in two but forms a crack when folded in four, 3: No crack when folded in tw o but splits when folded in four, 2: Cracks when folded in two, 1: Splits when folded in two. Results are mean SD. Different letters within each colu mn indicate significant difference (p < 0.05). Different numbers within each ro w indicate significant difference (p < 0.05). Added NaCl 2.5 2.9 11 11.2 Washed Muscle 2% 5.0 0.0 a, 1 5.0 0.0 a, 1 5.0 0.0 a, 1 5.0 0.0 a, 1 5.0 0.0 a, 1 0% 2.0 1.5 b, 1 1.0 0.0 b, 2 5.0 0.0 a, 3 5.0 0.0 a, 3 2.0 1.2 b, 1 Rheology Small strain oscillatory rheological testing was used to follow changes in viscoelastic properties of gels during heating and cooling. Initial and final storage modulus (G) during gelation wa s determined for all trea tments in two separate experiments (replicate 1 and 2) The results for initial G for both replicates are summarized in Figure 7-3. Due to large variation in data obtained for final G between the replicates, the results are represented in two separate bar graphs (Figure 7-4 and Figure 7-5). All treatments exhibited higher initial G (P a) in the absence of NaCl (Figure 7-3) compared to samples with 2% NaCl (w/w) adde d. A higher G translates to a more rigid system. The protein paste was at pH 7.1 to 7.2 which would give the muscle proteins a substantial negative charge ( 70 ). This negative charge cr eates strong repulsion forces between the proteins, creating more space for water to enter, and a more expanded and rigid system. When 2% NaCl (w/w) is added the ions screen some of these repulsive forces bringing the proteins closer together and decreasing the sp ace available for water to enter ( 46 ) and therefore decrease hydration and make the system less expanded. This

PAGE 76

64 explains the higher initial G for sample s without added salt. There were some differences in initial G between treatments For example, the washed muscle showed highest value among the samples that contained salt. This is likely due to more structure in the washed muscle, since some of the myofib rils would still be intact for that system. 0 1000 2000 3000 4000 5000 6000 7000 8000 2.52.91111.2Washed Muscle TreatmentG' (Pa) with 2% NaCl without 2% NaCl Figure 7-3. Storage modulus (G ) of protein pastes at 5C before gelation. Results are mean SD. The final G represents samples that have undergone thermally induced gelation along with setting on cooling (Figure 7-4 and Figure 7-5). The results for final G showed that the acid treated samples exhibited better gelling ability in replicate 2 than in replicate 1, indicating significant variations, wh ich could be attributed to the procedures used, the isolate or the raw material. Ther e is however evidence that the acid process may lead to larger variations in gelation when using small strain oscillatory testing

PAGE 77

65 compared to the alkali process ( 20 ). The reason for this is unknown, but is hypothesized to be due to unstable structural protein c onformations that can form at low pH, thus leading to different refolded st ructures at pH 5.5, where the proteins are precipitated. Although variations were large, the results indicate that the proteins from the alkali process have the ability to form stronger gels in both replicates compared to the proteins from the acid process, which partly agrees with the torsion results. In contrast to the torsion results wh ere samples containing 2% NaCl (w/w) exhibited better gelling ability, samples w ithout salt seemed to form stronger gels, especially in replicate 2, ex cept for the isolates made with the pH 2.9 treatment. The small scale oscillartory testing and the torsi on testing are not necessa rily expected to go hand in hand, since they are different tests done at different protein concentrations. The small scale oscillatory testing gives more insight into the gel forming mechanism and protein-protein interaction poten tial of the muscle proteins at lower concentrations, while the torsion test measures gel strength and quali ty at high protein con centration. It is interesting to note the higher G for the isolat es from the alkaline pr ocess, suggesting it has substantially more protein-protein inter actions and higher gel forming potential at lower protein concentrations than higher con centrations compared to the isolates from acid process which performed poorer at lo wer concentrations compared to higher concentrations. The rheology re sults also indicated that th e washed muscle performs better relative to the alkali isolates, at low compared to high protein concentrations.

PAGE 78

66 0 5000 10000 15000 20000 25000 30000 2.52.91111.2 TreatmentG' (Pa) with 2% NaCl without 2% NaCl Figure 7-4. Storage modulus (G ) of protein pastes at 5 C after gelation. Data represent the first replicate. The fina l G for gels was obtained by heating the protein samples from 5 to 80 C followed by cooling from 80 to 5C at a rate of 2C/min using a small strain oscillatory procedure. Results are mean SD.

PAGE 79

67 0 5000 10000 15000 20000 25000 30000 2.52.91111.2Washed Muscle TreatmentG' (Pa) with 2% NaCl without 2% NaCl Figure 7-5. Storage modulus (G ) of protein pastes at 5 C after gelation. Data represent the second replicate. The fi nal G for gels was obtained by heating the protein samples from 5 to 80 C followe d by cooling from 80 to 5 C at a rate of 2C/min using a small strain osci llatory procedure. Results are mean SD It was of interest to observe the change s in G during heating and cooling for the acid and alkali treated proteins and washed tilapia white muscle to gain an insight into possible differences in gel forming mechanis ms between the samples (Figures 7-6 to 7-15). Washed tilapia muscle with 2% (w/w) added NaCl (Figure 7-6) had a stable G, from 5 to 38C, where it started to increas e until peaking at a pproximately 44C and then declining down to a minimum at 53C. The G then rose steadily until 80C was reached. A similar curve for heating was observed by Klesk and coworkers ( 37 ) who studied the effect of state of rigor on the gel-forming ability of washed tilapia muscle. The increase up to 44C has been attributed to cross-linking of myosin and the drop in

PAGE 80

68 G after that to denaturation of light mero myosin, leading to increased fluidity. The second increase in G has been attributed to the formation of permanent cross-linked myosin filaments. A rapid increase in G was observed during cooling of the gel indicating formation of a firm gel structure Myofibrillar proteins, especially myosin and actin are believed to be largely responsible for gelation. Therefore, ge lation of myosin, actin and their complex actomyosin has been the subject of many st udies. For myosin, the heavy chain is believed to be the main subunit involved in the gelation. The role of the light chain seems to be dependendent on ionic strength wh ere at high ionic strength they are not very important. However, at low ionic strength, re moval of the light chains the rigidity of myosin gels was substantially lowered ( 2 ). Results from these studies have been used to try to elucidate gelation mechanism of co mplex mixture of myofibrillar proteins ( 2 ). The increase in G that was observed for the wa shed muscle on heating is attributed to dissociation of light chains from the heavy ch ains resulting in conf ormational changes to the molecule. Cross-linking of myosin is also believed to be responsible for the increase in G since it forms a strong elastic network. The subsequent drop is postulated to relate to dissociation of the actin-myosin complex and unfolding of the al pha-helix portion of the myosin rod ( 35 ). This transformation leads to a large increase in fluidity and might disrupt some protein networks that have already formed ( 35 ). All these structural changes lead to rapid aggregation and forma tion of gel networks resulting in a steady increase in G ( 2 ).

PAGE 81

69 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 0102030405060708090 Temperature (C)G' (Pa) Cooling Heating Figure 7-6. Example of a typical rheogram obtained for washed tilapia muscle with 2% NaCl (w/w) that shows storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating. Acid treated proteins with 2% NaCl (w/w) behaved quite differently from the washed muscle. The pH 2.5 treated proteins showed a continuous increase in G which started around 36C and leveled off at appr oximately 65C (Figure 7-7). A similar pattern was observed for the heating phase of the pH 2.9 treatment (Figure 7-8). The onset of gelation was thus at a significantly lo wer temperature that that seen for washed tilapia muscle. This could be due to a more unfolded structure of the acid treated proteins, thus requiring a lower temperat ure to further unfold and inte ract on heating. The same

PAGE 82

70 was seen for cod muscle proteins by Kristinsson and Hultin ( 8 ). The large difference observed between replicates of the acid tr eated proteins has been discussed in the previous section and maybe attributed to unsta ble protein structures at low pH leading to different refolded structures which can affect the gelation properties. 0 2000 4000 6000 8000 10000 12000 14000 16000 0102030405060708090 Temperature (C)G' (Pa) Replicate 1 Replicate 2 Cooling Cooling Heating Figure 7-7. Examples of typical rheograms from two independent experiments (replicate 1 and 2) obtained for proteins treated at pH 2.5 with 2% NaCl (w/w). The rheograms show storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating.

PAGE 83

71 0 2000 4000 6000 8000 10000 12000 14000 16000 020406080100 Temperature (C)G' (Pa) Replicate 1 Replicate 2 Cooling Cooling Heating Figure 7-8. Examples of typical rheograms from two independent experiments (replicate 1 and 2) obtained for proteins treated at pH 2.9 with 2% NaCl (w/w). The rheograms show storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating. The alkali treated proteins showed a sim ilar viscoelastic curve on heating as the acid treated proteins. The G started to gradually increase around 40C but as presented previously had a larger final G indicating th e formation of a firmer gel compared to the acid treated proteins (Fi gure 7-9 and Figure 7-10). The slightly higher point of gelation for the alkali treated proteins compared to ac id treated proteins s uggested differences in conformation. The higher onset temperature for the alkali treated proteins may suggest that they are more refolded than the acid treated proteins. The protein composition

PAGE 84

72 differences between the two isolates might also partly account for the differences in temperature sensitivity 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 22000 0102030405060708090 Temperature (C)G' (Pa) Replicate 1 Replicate 2 Cooling Cooling Heating Figure 7-9. Examples of typical rheograms from two independent experiments (replicate 1 and 2) obtained for proteins treated at pH 11.0 with 2% NaCl (w/w). The rheograms show storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating.

PAGE 85

73 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 22000 24000 26000 0102030405060708090 Temperature (C)G' (Pa) Replicate 1 Replicate 2 Cooling Cooling Heating Figure 7-10. Examples of typical rheograms from tw o independent experiments (replicate 1 and 2) obtained for proteins treated at pH 11.2 with 2% NaCl (w/w). The rheograms show storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) were adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating. The acid and alkali treated proteins with added 2% Na Cl (w/w) did not show the same transitions as were observed for washed muscle possibly due to denaturation of the muscle proteins during proces sing. It has been found that major changes occur with myosin on acid and alkali treatment ( 8 ). It was for example found that acid treatment led to complete dissociation of cod myosin while alkaline treatment only led to the dissociation of the light chai ns from the myosin head. As a result the proteins had different viscoelastic behavior on heating and cooling compared to a washed cod muscle

PAGE 86

74 ( 8 ). It was believed that the absence of light chains led to protein aggregation and crosslinking at lower temperatures due to a more exposed myosin head group. Yongsawatdigul and Park ( 71 ) also observed that acid and alkali treated proteins from rockfish were denatured during treatment while still retaining ability to form gels. The lower gelation ability of the acid-treated proteins compared to the alkali treated could be due to different conformational changes; partly because of loss of myosin heavy chain during processing, or because of unfavorable conformation of the proteins during acidic treatments (e.g., too many hydrophobic groups, l eading to bigger aggregates and a less ordered gel). Another explanation for the poor gelling ability of the acid-treated proteins could be the presence of denatured sarcoplas mic proteins, many which were retained in the acid process but not in the alkali a nd washed process. Crynen and Kristinsson demonstrated with catfish proteins that the acid process not only negatively affect the myofibrillar proteins but also lead to change s in the sarcoplasmic proteins that, when mixed with the myofibrillar proteins, have a very detrimental effect on gelation. Washed muscle without 2% added NaCl (Figure 7-11) exhibited different viscoelastic behavior on heati ng compared to the washed muscle containing salt. Initial G was higher and decreased with increasing temperature until it reached a minimum at 46C, which was 7C lower than the minimum observed for washed muscle with 2% NaCl (w/w). This decrease suggests that pr otein complexes may have dissociated instead of forming cross-links as in the system with salt. A steady increase, similar to the one seen for the proteins in salt, was then obs erved with increasing h eating excluding a small bump at approximately 56C. The incr ease in G on cooling was similar for both washed muscle in the pres ence and absence of salt.

PAGE 87

75 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 0102030405060708090 Temperature (C)G' (Pa) Cooling Heating Figure 7-11. An example of a typical rhe ogram obtained for washed tilapia muscle without 2% NaCl. (w/w) th at shows storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) were adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating. When the viscoelastic curves for samples without added NaCl were compared it to samples with 2% NaCl (w/w) they were quite different and resembled the curves for washed tilapia muscle. The acid treated proteins showed a slight decrease in G on heating until they reached approximately 40C, where the G dropped down to a minimum at 47C. The G then increased ag ain and leveled off at higher temperatures. Similar to the samples containing salt, the acid treated proteins wit hout added salt showed large variability and only one curve from both re plicates was usable for the isolates made

PAGE 88

76 with the pH 2.5 treatment (Figur e 7-12). Protein isolates made with the pH 2.9 treatment showed a very similar trend, however with a lower final G (Figur e 7-13). The alkali treated proteins without added salt showed the same trend as the acid treated proteins except for a higher final G and less va riability (Figure 7-14 and Figure 7-15). 0 2000 4000 6000 8000 10000 12000 14000 16000 0102030405060708090 Temperature (C)G' (Pa) Heating Cooling Figure 7-12. Examples of typi cal rheograms from obtained for proteins treated at pH 2.5 without 2% NaCl (w/w). The rheogr ams show storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) ad justed to pH 7.1 to 7.2 we re allowed to sit for 30 min on ice before heating.

PAGE 89

77 0 2000 4000 6000 8000 10000 12000 14000 16000 0102030405060708090 Temperature (C)G' (Pa) Replicate 1 Replicate 2 Cooling Heating Figure 7-13. Examples of typical rheograms from tw o independent experiments (replicate 1 and 2) obtained for proteins treated at pH 2.9 without 2% NaCl (w/w). The rheograms show storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating.

PAGE 90

78 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 22000 24000 26000 0102030405060708090 Temperature (C)G' (Pa) Replicate 1 Replicate 2 Cooling Heating Figure 7-14. Examples of typical rheograms from tw o independent experiments (replicate 1 and 2) obtained for proteins treated at pH 11. 0 without 2% NaCl (w/w). The rheograms show storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating.

PAGE 91

79 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 22000 24000 26000 0102030405060708090 Temperature (C)G' (Pa) Replicate 1 Replicate 2 Heating Cooling Figure 7-15. Examples of typical rheograms from tw o independent experiments (replicate 1 and 2) obtained for proteins treated at pH 11. 2 without 2% NaCl (w/w). The rheograms show storage modulus (G) during heating at 5 to 80C followed by cooling from 80 to 5C at 2C/min. Samples (~10% protein (w/w)) adjusted to pH 7.1 to 7.2 were allowed to sit for 30 min on ice before heating. Not many gelation studies have been performe d with fish proteins in the absence of salt, in part due to the long held believe that salt was necessary for gelation. The results here show that gels can form in the absence of salt, and higher final G values were in most cases seen in the absence of salt. Th e absence of salt leads to more electrostatic repulsion between the proteins at pH 7.1 to 7.2 compared to samples with salt, which may explain the higher G. The onset of thermal gelation was also different for samples in the absence of salt. For example the rheologi cal curve of washed muscle without salt

PAGE 92

80 exhibited a similar pattern reaching a minimu m 7C lower than compared with salt. Another observation was that the rheology curv es for acid and alkali treated proteins without salt exhibited all simila r drop in G around 40C. This drop could be explained by a breakup of protein aggreg ates with higher temperatures and the subsequent increase in G due to proteins de naturing and forming permanent cross-links at higher temperatures. Washed muscle consists mainly of myofibrillar prot eins because much of the sarcoplasmic proteins are re moved during the washing step ( 12 ). The acid-produced protein isolate contained a si zable amount of sarcoplasmic proteins, while the alkali isolate had significantly less of the sarcoplasmic proteins. The types of proteins recovered therefore does not seem to explain this similar behavior of the acid and alkali isolates on heating. The above emphasizes that the differences between the two systems are far from trivial and mo re investigation is needed to understand the underlying mechanisms for gelation in salt and no salt. Water holding Capacity Water Holding Capacity on Cooking The same gels which were tested with the Torsion Gelometer were analyzed for water loss upon cooking (Figure 7-16). All treatments gave gels with good water-holding properties. Addition of 2% NaCl (w/w) si gnificantly reduced (p < 0.05) water loss for protein isolates made with the pH 2.9 and pH 11.0 treatment compared to other treatments. Most water loss or ~1% was obs erved for acidified proteins at pH 2.9 without the addition of 2% Na Cl (w/w) which is still quite good. A study performed by Feng and Hultin ( 36) on washed chicken muscle protein at pH 7.0 with 0.15M NaCl resulted in 1.7% loss of water during c ooking. Another study done by Kristinsson and

PAGE 93

81 Hultin ( 46 ) on chicken breast muscle showed that sa mples without salt lost more water on cooking compared to samples containing salt. -4.0 -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 2.5 2.9 11.0 11.2Washed muscle Treatments% loss water With 2% added salt Without 2% salt Figure 7-16. Water loss (%) of tilapia muscle protein pastes on cooking at 80C for 30 min. Results are mean SD. Water Holding Capacity on Pressing The same connection between salt concen tration and water holding capacity was seen for the gels when they were subjected to a press test (i.e., gels without 2% salt lost more water than those containing salt, except for gels from the washed muscle where no difference was seen). Again, the gel prepared from proteins treated at pH 2.9 lost most water on pressing than any othe r treatment. The gel without added salt from the other pH-treatments had similar water loss. The gel from the washed muscle without added salt lost the least amount of water upon pressing. A smaller difference between treatments was seen for samples containing 2% NaCl (w/w).

PAGE 94

82 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 2.52.91111.2Washed Treatments% loss of water with 2% NaCl without 2% NaCl Figure 7-17. Water loss (%) of heat induced tilapia muscle protein gel with and without 2% NaCl under pressure (3kg) pressing. Results are mean SD. Gel structure is believed to greatly influence the ability of a gel to hold water. In a well ordered gel, movement of water might be restricted compared to a less ordered structure ( 45 ). This is partly due to strong capilla ry forces within the ordered structure and also high gel pressure if water is bei ng held due to strong repulsive forces between negative charges in the protein matrix ( 46 ). Between pH 7.1 and 7.2 there are strong repulsive forces between muscle proteins and these repulsive forces strengthen even more in the absence of salt ( 46 ). The water-holding results agree with the gelation results, in that the weaker and more brittle gels had lower water-holding abililty. Formation of localized aggregates is believed to co ntribute to lower gel forming ability ( 36 ) and a correlation between a gel elasticity and water holding capacity has been seen for chicken breast gels ( 46 ). For example the gels from the pH 2.9 treatment had the poorest gel

PAGE 95

83 forming ability (especially in the absence of sa lt), thus having a poor er gel structure, and also had the least ability to hold water than the other treatme nts. The salt free gels in general had poorer gel forming ab ility than the gels with salt, and these did also have worse water-holding ability. The addition of salt to the proteins thus appears to have led to a better ordered structure than the absen ce of salt. The salt is able to partially solubilize the muscle proteins, and this could have been important for the formation of a ordered structure since protei ns would have been disperse d better prior to gelation. SH-Content Changes in free sulfhydryl (SH) groups and disulfide bonds (S-S) are often associated with gelation of muscle proteins It is also known th at SH groups become more reactive at the high pH values re presentative of the alkaline process ( 71 ). It was thus of interest to see if the processes led to a change in SH and S-S groups. Therefore, total SH content of protein pastes before cooking and gels after cooking was determined by unfolding the proteins using urea and thus exposing SH gr oups that might be located in the interior of the proteins (Figures 3-1 and 3-2). Washed muscle exhibited the highest total SH content before cooking for samp les without added NaCl, whereas washed muscle with 2% NaCl (w/w) had significantly lower (p < 0.05) total SH content. There were no significant differences (p > 0.05) between gels with no added salt and added salt between the other treatments. There were also no significant difference (p > 0.05) in total SH-groups after cooking between treatments and salt concentrations. It was also interesting to see that there were no signifi cant (p > 0.05) differen ce in total SH-groups before and after gel formation.

PAGE 96

84 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 2.52.91111.2washed muscleTreatmentsumol/g protein with 2% salt without 2% salt Figure 7-18. Shows total SH group content in tilapia muscle protein paste before cooking. Results are mean SD. 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 2.52.91111.2washed muscleTreatmentsumol/g protein With 2% NaCl Without 2% NaCl Figure 7-19. Shows total SH group content in tilapia muscle protein gel after cooking. Results are mean SD.

PAGE 97

85 A reduction in total SH groups indicate s the formation of S-S bonds between proteins, and has been found on many occasio ns on gelation of muscle proteins ( 68 ). A significant difference between the low and high pH treatments was not found with the tilapia protein isolates. Other studies with the acid and alkali pr ocesses have either reported differences or no differences in re active and total SH-g roups. Kristinsson and Hultin ( 49 ) showed that alkali treat ed myosin had more reac tive SH groups than acid treated myosin, and also had better get fo rming ability. A study by Yongsawatdigul and Park ( 71 ) performed on rockfish muscle showed th at washed muscle had lower content of total SH compared to unwashed muscle. Th ese authors observed a hi gher amount of total SH groups in proteins made with the alkaline process compared to acid process, thus suggesting S-S bonds formed during the acid proc ess. The proteins made with the alkali process had the larges t decrease in SH groups, sugges ting that more S-S bonds were formed during gelation for that treatment comp ared to other treatments. This was not seen with the tilapia muscle proteins. Another study performed on Pacific whiting showed a significantly higher tota l SH content of isolates made at low pH compared to high pH, suggesting SH groups became more re active on alkali treatmen t. Even though differences were seen in SH groups with Pacific whiting proteins there were no link between these and gel forming ability. A study performed on threadfin bream showed that total sulfhydryl groups were stable up to 30C but decreased from 40 to 80C which indicates formation of disulf ide linkages during heating ( 19 ). Again, the data with the tilapia proteins indicate th at no significant S-S bonds were formed on heating and thus differences in SH groups coul d not explain the different f unctionality of the proteins.

PAGE 98

86 CHAPTER 8 CONCLUSIONS The use of the acid and alkali aided processes on tilapia musc le proteins regarding isolation parameters showed that the type of proteins recovered using low and high pH was significantly different whereas the quantity of tota l protein recovered was not different. Gel forming ability of the acid and alka li treated proteins was determined and compared to washed tilapia muscle. Using larg e strain testing (torsi on) showed that gels with 2% added NaCl were stronger and more elastic compared to samples without 2% NaCl. The alkali treated protei ns also gave better results when compared to acid treated proteins and washed muscle. The use of sm all strain oscillatory testing gave very interesting results. Samples without 2% NaCl performed in some cases better than samples with 2% NaCl. Variability at low pH was great possibly due to a more unstable structure of the proteins. The small and larg e strain testing are not necessarily expected to go hand in hand because they are differe nt tests performed at different protein concentration. Torsion determines fundame ntal gel quality and has been related to acceptability of a product texture whereas osci llatory small strain testing gives more insight in gel forming ability of the protei ns at lower protein c oncentration. More investigation is needed to understand the underlying mechanism and the molecular properties involved in gelation of tilapia muscle proteins. The findings from this research indicate th at alkali treated tila pia muscle proteins can form strong and elastic gels upon heating.

PAGE 99

87 LIST OF REFERENCES 1. Hultin, H. O.; Kelleher, S. D. Process for Isolating a Protein Composition from a Muscle Source and Protein Composition. U.S. Patent 6,005,073, December 21, 1999. 2. Xiong, Y. L. Structure-Function Re lationship of Muscle Proteins. In Food Proteins and Their Applications ; S. Damodaran and A. Paraf, Eds.; Marcel Dekker: New York, 1997; pp 341-392. 3. Kristinsson, H. G.; Rasco, B. A. Fish Protein Hydrolysates: Production, Biochemical, and Functional Properties. Critical Reviews in Food Science and Nutrition 2000, 40 43-81. 4. Mackie, I. M. Fish Protein. In New and Developing Sources of Food Proteins ; B. J. F. Hudson, Ed.; Chapman and Hall: New York, 1994; pp 95-143. 5. Hultin, H. O. Recent Advances in Surimi Technology. In Recent Advances in Marine Biotechnology ; M. Fingerman and R. Nagabhushanam, Eds.; Science Publisher, Inc.: Enfield, Ne w Hampshire, 2002; pp 241-251. 6. Hultin, H. O.; Kelleher, S. D. Surimi Processing from Dark Muscle Fish. In Surimi and Surimi Seafood ; J. W. Park, Ed.; Marcel Dekker: New York, 2000; pp 59-77. 7. Hultin, H. O.; Kelleher, S. D.; Feng, Y. M.; Kristinsson, H. G.; Richards, M. P.; Undeland, I. A. High Efficiency Alka line Protein Extraction. U.S. Patent 6,136,959, October 24, 2000. 8. Kristinsson, H. G.; Hultin, H. O. Effect of Low and High pH Treatment on the Functional Properties of Cod Muscle Proteins. Journal of Agricultural and Food Chemistry 2003, 51 5103-5110. 9. Undeland, I. A.; Kelleher, S. D.; Hultin, H. O. Recovery of Functional Proteins from Herring ( Clupea Harengus ) Light Muscle by an Acid or Alkaline Solubilization Process. Journal of Agricultural and Food Chemistry 2002, 50 7371-7379. 10. Choi, Y. J.; Park, J. W. Acid-Aided Protein Recovery from Enzyme-Rich Pacific Whiting. Journal of Food Science 2002, 67 2962-2967.

PAGE 100

88 11. Fitzsimmons, K. Tilapia Aquaqulture in the 21st Century. 2004; pp power point presentation. 12. Park, J. W.; Lanier, T. C. Processing of Surimi and Surimi Seafoods. In Marine & Freshwater Products Handbook ; R. E. Martin, Ed.; Technomic Publishing Company: Lancaster, NH, 2000. 13. Yongsawatdigul, J.; Park, J. W.; Viru lhakul, P.; Viratchakul, S. Proteolytic Degradation of Tropical Tilapia Surimi. Journal of Food Science 2000, 65 129133. 14. Kristinsson, H. G.; Rasco, B. A. Hydr olysis of Salmon Muscle Proteins by an Enzyme Mixture Extracted from Atlantic Salmon (Salmo Salar) Pyloric Caeca. Journal of Food Biochemistry 2000, 24 177-187. 15. Ohshima, T.; Ushio, H.; Koizumi, C. Hi gh Pressure Processing of Fish and Fish Products. Trends in Food Science and Technology 1993, 4 370-375. 16. Park, J. W. Surimi and Surimi Seafood ; Marcel Dekker Inc.: New York, 2000. 17. Klesk, K.; Yongsawatdigul, J.; Park, J. W.; Viratchakul, S.; Virulhakul, P. Gel Forming Ability of Tropical Tilapia Surimi as Compared with Alaska Pollock and Pacific Whiting Surimi. Journal of Aquatic Food Product Technology 2000, 9 91-104. 18. Kristinsson, H. G.; Demir, N. Functio nal Fish Protein Ingredients from Fish Species of Warm and Temperate Waters: Comparison of Acid and Alkali-Aided Processing Vs. Conventional Surimi Processing. Advances in Seafood Byproducts, Fairbanks, AK ; Alaska Sea Grant College Program. 2003. 19. Yongsawatdigul, J.; Park, J. W. Th ermal Denaturation and Aggreegation of Threadfin Bream Actomyosin. Food Chemistry 2003, 83 409-416. 20. Theodore, A.; Kristinsson, H. G.; Crynen, S. The Effect of pH and Salt on the Gelation Properties of Catfish Surimi and Catfish Protein Isolat es from Acid and Alkali-Aided Processing. IFT Annual Meeting, Chicago ; Abstract 76A-10. 2003. 21. Davenport, M. P.; Kristinsson, H. G. Low and High pH Treatments Induce a Molten Globular Structure in Myosin Which Improves Its Gelation Properties. IFT Annual Meeting, Chicago ; Abstract 42-9. 2003. 22. Kristinsson, H. G.; Crynen, S. The E ffect of Acid and Alkali Treatments on the Gelation Properties of Sarcoplasmic and Myof ibrillar Proteins of Channel Catfish. IFT Annual Meeting, Chicago ; Abstract 87-3. 2003. 23. Park, J. W.; Korhonen, R. W.; Lanier T. C. Effects of Rigor-Mortis on GelForming Properties of Surimi and Unwa shed Mince Prepared from Tilapia. Journal of Food Science 1990, 55 353-355,366.

PAGE 101

89 24. Huang, C.-H.; Lai, H.-T.; Weng, Y.-M. Suitability of Hybrid Tilapia ( Oreochromis Niloticus X Oreochromis Aureus ) Muscle for Gel Formation. International Journal of Food Science and Technology 1998, 33 339-344. 25. Kongpun, O. The Gel Forming Ability of Washed and Unwashed Fish Meat (Lizardfish and Nile Tilapia). Nat. Sci 1999, 33 258-269. 26. Onibala, H.; Takayama, T.; Shindo, J. ; Hayashi, S.; Miki, H. Influence of Freshness on Occurrence of Setting and Disi ntegrating in Heat-Induced Gels from Tilapia. Fisheries Science 1997, 63 276-280. 27. Xiong, Y. L. Meat Processing. In Food Proteins, Processing Applications ; S. Nakai and H. W. Modler, Eds.; W iley-VCH: New York, 2000; pp 89-145. 28. Hultin, H. O.; Feng, Y. M.; Stanley, D. W. A Re-Examination of Muscle Protein Solubility. Journal of Muscle Foods 1995, 6 91-107. 29. Stefansson, G.; Hultin, H. O. On the Sol ubility of Cod Muscle Proteins in Water. Journal of Agricultural and Food Chemistry 1994, 42 2656-2664. 30. Damodaran, S. Amino Acids Peptides and Proteins. In Food Chemistry ; O. R. Fennema, Ed.; Marcel Decker: New York, 1996; pp 321-429. 31. Dagher, S. M.; Hultin, H. O.; Liang, Y. Solubility of Cod Muscle Myofibrillar Proteins at Alkaline PH. Journal of Aquatic Food Product Technology 2000, 9 49-59. 32. DeWitt, C. A. M.; Gomez, G.; James, J. M. Protein Extraction from Beef Heart Using Acid Solubilization. Journal of Food Science 2002, 67 3335-3341. 33. Damodaran, S. Food Proteins: An Overview. In Food Proteins and Their Applications ; S. Damodaran and A. Paraf, Ed s.; Marcel and Dekker, Inc.: New York, 1997; pp 1-24. 34. Bryant, C. M.; McClements, J. Molecu lar Basis of Protien Functionality with Special Consideration of Cold Set Gels Derived from Heat Denatured Whey. Trends in Food Science and Technology 1998, 9 143-151. 35. Xiong, Y. L.; Blacnchard, S. P. D ynamic Gelling Properties of Myofibrillar Protein from Skeletal Muscles of Different Chicken Parts. Journal of Agricultural and Food Chemistry 1994, 42 670-674. 36. Feng, Y. M.; Hultin, H. O. Effect of PH on the Rheological and Structural Properties of Gels of Water-Washed Ch icken-Breast Muscle at Physiological Ionic Strength. Journal of Agricultural and Food Chemistry 2001, 49 3927-3935. 37. Klesk, K.; Yongsawatdigul, J.; Park, J. W.; Viratchakul, S.; Virulhakul, P. Physical Behavior of Tropical Tilapia Su rimi Gels at Various Thermal Treatments

PAGE 102

90 as Compared with Alaska Pollock and Pacific Whiting Surimi. Journal of Aquatic Food Product Technology 2000, 9 91-104. 38. Lanier, T. C.; Lin, T. S.; Liu, Y. M. ; Hamman, D. D. Heat Gelation Properties of Actomyosin and Surimi Prepar ed from Atlantic Croaker. Journal of Food Science 1982, 47 1921. 39. Hermansson, A. M. Physico Chemical Aspects of Soy Pr oteins Structure Formation. Journal of Texture Studies 1978, 9 33-58. 40. Park, J. W.; Lin, T. M.; Yongsawatdi gul, J. New Developments in Manufacturing of Surimi and Surimi Seafood. Food Review International 1997, 13 577-610. 41. Shimizu, Y.; Nishioka, F. M.; Shien, C. M. Gelation Characteristics of SaltAdded Myosin Sol. Japan Soc. Sci. Fish 1983, 49 1239. 42. Grabowska, J.; Sikorski, Z. E. The Gel Forming Ability of Fish Myofibrillar Proteins. Lebensm.-Wiss. u Technol. 1976, 9 33-35. 43. Dickinson, E. Rheology. In An Introduction to Food Colloids ; E. Dickinson, Ed.; Oxford University Press: Oxford, UK, 1992; pp 51-78. 44. Hamman, D. D. A Tool for Understa nding Thermally Induced Protein Gelation. In Interactions of Food Proteins ; P. Nicholas and R. Barford, Eds.; American Chemical Society: Washington, DC, 1991; pp 212-227. 45. Hamann, D. D.; MacDonald, G. A. Rheo logy and Texture Properties of Surimi and Surimi Based Foods. In Surimi Technology ; T. C. Lanier and C. M. Lee, Eds.; Mercel Dekker: New York, 1992; pp 429-500. 46. Kristinsson, H. G.; Hultin, H. O. Role of pH and Ionic Strength on Water Relationships in Washed Minced Chicken-Breast Muscle Gels. Journal of Food Science 2003, 68 917-922. 47. Ingadottir, B.; Kristinsson, H. G. Ac id and Alkali Unfolding and Refolding Strategies Improve the Foami ng Properties of Egg Albumen. IFT Annual Meeting, Chicago ; Abstract 42-2. 2003. 48. Vissessanguan, W.; An, H. Effects of Proteolysis and Mechanism of Gel Weakening in Heat-Induced Ge lation of Fish Myosin. Journal of Agricultural and Food Chemistry 2000, 48 1024-1032. 49. Kristinsson, H. G.; Hultin, H. O. Changes in Conformation and Subunit Assembly of Cod Myosin at Low and High pH and after Subsequent Refolding. Journal of Agricultural and Food Chemistry 2003, 51 7187-7196. 50. Torten, J.; Whitaker, J. R. Evaluation of the Biuret and Dye-Binding Methods for Protein Determination in Meats. Journal of Food Science 1964, 29 168-174.

PAGE 103

91 51. Laemmli, U. K. Cleavage of Structural Proteins During the A ssembly of the Head of Bacteriophage T4. Nature 1970, 227 680-685. 52. Kudo, G.; Okada, M.; Miyauchi, D. Gel-Forming Capacity of Washed and Unwashed Flesh of Some Paci fic Coast Species of Fish. Marine Fisheries Review 1973, 32 10-15. 53. McClements, J. Rheological Properties of Emulsions. In Food Emulsions. Principles, Practice and Techniques ; CRC Press: Boca Raton, 1999; pp 254-264. 54. Goto, Y.; Fink, A. L. Acid Indu ced Folding of Heme Proteins. In Methods in Enzymology ; Academic Press, Inc.: New York, 1994; pp 3-15. 55. Undeland, I. A.; Kelleher, S. D.; Hultin, H. O.; McClements, J.; Thongraung, C. Consistency and Solubility Changes in Herring ( Clupea Harengus ) Light Muscle Homogenates as a Function of pH. Journal of Agricultural and Food Chemistry 2003, 51 3992-3998. 56. Kim, Y. S.; Park, J. W.; Choi, Y. J. New Approaches for the Effective Recovery of Fish Proteins and Their P hysiochemical Characteristics. Fisheries Science 2003, 69 1231-1239. 57. Haard, N. F.; Simpson, B. K.; Sun Pan, B. Sarcoplasmic Proteins and Other Nitrogenous Compounds. In Seafood Proteins ; Z. E. Sikorski; B. Sun Pan and F. Shahidi, Eds.; Chapman and Hall: London, 1994; pp 13-39. 58. Parsons, N.; Knight, P. Origin of Variab le Extraction of Myosin from Myofibrils Treated with Salt and Pyrophosphate. Journal of The Science of Food and Agriculture 1990, 51 71-90. 59. Kristinsson, H. G.; Theodore, A.; Demi r, N.; Ingadottir, B. A Comparative Study between Acid and Alkali-Aided Proc essing and Surimi Processing for the Recovery of Proteins from Channel Catfish Muscle. Journal of Food Science 2004, Submitted 60. Chang, R. Macromolecules. In Physical Chemistry for the Physical and Biological Sciences ; R. Chang, Ed.; University Science Books: Sausalito, 2000; pp 895-940. 61. Kristinsson, H. G. Conformational an d Functional Changes of Hemoglobin and Myosin Induced by pH. Dissertation, University of Massachusetts, 2002. 62. Hultin, H. O., The Use of Acid Treated Proteins for Gelation.2003 63. Cortes-Ruiz, J. A.; Pacheco-Aguilar, R.; Garcia-Sanches, G.; Lugo-Sanches, M. E. Functional Characterization of Protein Concentrates from Bristly Sardine under Acidic Conditions. Journal of Aquatic Food Product Technology 2001, 10 2-23.

PAGE 104

92 64. Kristinsson, H. G.; Hultin, H. O. Ch anges in Trout Hemoglobin Conformations and Solubility after Exposure to Acid and Alkali pH. Journal of Agricultural and Food Chemistry 2004, 52 3633-3643. 65. Lin, D. D.; Park, J. W. Extraction of Proteins from Pacific Whiting Mince at Various Washing Conditions. Journal of Food Science 1996, 61 432-438. 66. Chang, H. S.; Feng, Y. M.; Hultin, H. O. Role of pH in Gel Formation of Washed Chicken Muscle at Low Ionic Strength. Journal of Food Biochemistry 2001, 25 439-457. 67. Bakir, H. M.; Hultin, H. O.; Kelleher, S. D. Gelation Properties of Fatty Fish Processed with or without Added S odium Chloride, Cryoprotectants and Antioxidants. Food Research International 1994, 27 443-449. 68. Lanier, T. C. Surimi Gelation Chemistry. In Surimi and Surimi Seafood ; J. W. Park, Ed.; Marcel Dekker, Inc: New York, 2000; pp 237-265. 69. Davenport, M. P.; Theodore, A.; Kristinss on, H. G. Influence of Insoluble Muscle Components on the Gelation Properties of Catfish Protein Isolates Made with Acid and Alkali Aided Processing. IFT Annual Meeting, Las Vegas ; Abstract 83A-21. 2004. 70. Feng, Y. M.; Hultin, H. O. Solubility of the Proteins of Mackerel Light Muscle at Low Ionic Strength. Journal of Food Biochemistry 1997, 21 479-496. 71. Yongsawatdigul, J.; Park, J. W. Eff ects of Alkali and Ac id Solubilization on Gelation Characteristics of Rockfish Muscle Proteins. Journal of Food Science 2004, 69 499-505.

PAGE 105

93 BIOGRAPHICAL SKETCH Bergros Ingadottir was born and raised in Reykjavik, Iceland. Bergros graduated from the University of Iceland in June 2002 where she received a B.Sc. in food science. She was accepted in the Department of Food Science and Human Nutrition at the University of Florida in August 2002, where sh e was awarded a resear ch assistanceship. There she pursued a masters degree in food sc ience, specializing in protein biochemistry under the supervision of Dr. Hordur G. Kristin sson. While at the University of Florida she was invited to join Phi Tau Sigma, the honorary society of food scientists. Bergros graduated in 2004 and returned to Iceland.