Citation
Effects of Low and High pH on Lipid Oxidation in Spanish Mackerel

Material Information

Title:
Effects of Low and High pH on Lipid Oxidation in Spanish Mackerel
Creator:
PETTY, HOLLY TASHA ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Antioxidants ( jstor )
Food ( jstor )
Gels ( jstor )
Hemoglobins ( jstor )
Juices ( jstor )
Lipids ( jstor )
Mackerel ( jstor )
Oxidation ( jstor )
pH ( jstor )
Reactive oxygen species ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Holly Tasha Petty. Permission granted to University of Florida to digitize and display 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:
7/12/2007
Resource Identifier:
660034023 ( OCLC )

Downloads

This item has the following downloads:

petty_h ( .pdf )

petty_h_Page_040.txt

petty_h_Page_008.txt

petty_h_Page_105.txt

petty_h_Page_129.txt

petty_h_Page_039.txt

petty_h_Page_087.txt

petty_h_Page_001.txt

petty_h_Page_064.txt

petty_h_Page_146.txt

petty_h_Page_014.txt

petty_h_Page_021.txt

petty_h_Page_031.txt

petty_h_Page_082.txt

petty_h_Page_092.txt

petty_h_Page_052.txt

petty_h_Page_076.txt

petty_h_Page_093.txt

petty_h_Page_048.txt

petty_h_Page_112.txt

petty_h_Page_117.txt

petty_h_Page_015.txt

petty_h_Page_121.txt

petty_h_Page_154.txt

petty_h_Page_025.txt

petty_h_Page_113.txt

petty_h_Page_101.txt

petty_h_Page_157.txt

petty_h_Page_145.txt

petty_h_Page_060.txt

petty_h_Page_017.txt

petty_h_Page_094.txt

petty_h_Page_158.txt

petty_h_Page_043.txt

petty_h_Page_098.txt

petty_h_Page_055.txt

petty_h_Page_144.txt

petty_h_Page_074.txt

petty_h_Page_020.txt

petty_h_Page_089.txt

petty_h_Page_116.txt

petty_h_Page_120.txt

petty_h_Page_028.txt

petty_h_Page_016.txt

petty_h_Page_091.txt

petty_h_Page_104.txt

petty_h_Page_012.txt

petty_h_Page_061.txt

petty_h_Page_009.txt

petty_h_Page_163.txt

petty_h_Page_107.txt

petty_h_Page_068.txt

petty_h_Page_018.txt

petty_h_Page_056.txt

petty_h_Page_073.txt

petty_h_Page_010.txt

petty_h_Page_110.txt

petty_h_Page_069.txt

petty_h_Page_088.txt

petty_h_Page_096.txt

petty_h_Page_041.txt

petty_h_Page_071.txt

petty_h_Page_118.txt

petty_h_Page_035.txt

petty_h_Page_152.txt

petty_h_Page_070.txt

petty_h_Page_057.txt

petty_h_Page_026.txt

petty_h_Page_067.txt

petty_h_Page_150.txt

petty_h_Page_077.txt

petty_h_Page_135.txt

petty_h_Page_003.txt

petty_h_Page_160.txt

petty_h_Page_159.txt

petty_h_Page_123.txt

petty_h_Page_131.txt

petty_h_Page_033.txt

petty_h_Page_024.txt

petty_h_Page_019.txt

petty_h_Page_013.txt

petty_h_Page_148.txt

petty_h_Page_134.txt

petty_h_Page_085.txt

petty_h_Page_072.txt

petty_h_Page_044.txt

petty_h_Page_063.txt

petty_h_Page_084.txt

petty_h_Page_058.txt

petty_h_Page_156.txt

petty_h_Page_122.txt

petty_h_Page_004.txt

petty_h_Page_050.txt

petty_h_Page_137.txt

petty_h_Page_103.txt

petty_h_Page_161.txt

petty_h_Page_036.txt

petty_h_Page_030.txt

petty_h_Page_139.txt

petty_h_Page_153.txt

petty_h_Page_032.txt

petty_h_Page_080.txt

petty_h_Page_143.txt

petty_h_Page_127.txt

petty_h_pdf.txt

petty_h_Page_097.txt

petty_h_Page_066.txt

petty_h_Page_090.txt

petty_h_Page_045.txt

petty_h_Page_054.txt

petty_h_Page_005.txt

petty_h_Page_065.txt

petty_h_Page_149.txt

petty_h_Page_142.txt

petty_h_Page_079.txt

petty_h_Page_006.txt

petty_h_Page_126.txt

petty_h_Page_132.txt

petty_h_Page_151.txt

petty_h_Page_051.txt

petty_h_Page_059.txt

petty_h_Page_078.txt

petty_h_Page_109.txt

petty_h_Page_141.txt

petty_h_Page_007.txt

petty_h_Page_046.txt

petty_h_Page_102.txt

petty_h_Page_114.txt

petty_h_Page_162.txt

petty_h_Page_062.txt

petty_h_Page_108.txt

petty_h_Page_027.txt

petty_h_Page_138.txt

petty_h_Page_081.txt

petty_h_Page_002.txt

petty_h_Page_111.txt

petty_h_Page_037.txt

petty_h_Page_136.txt

petty_h_Page_029.txt

petty_h_Page_049.txt

petty_h_Page_075.txt

petty_h_Page_053.txt

petty_h_Page_023.txt

petty_h_Page_038.txt

petty_h_Page_133.txt

petty_h_Page_124.txt

petty_h_Page_047.txt

petty_h_Page_106.txt

petty_h_Page_083.txt

petty_h_Page_125.txt

petty_h_Page_100.txt

petty_h_Page_128.txt

petty_h_Page_011.txt

petty_h_Page_042.txt

petty_h_Page_155.txt

petty_h_Page_034.txt

petty_h_Page_022.txt

petty_h_Page_119.txt

petty_h_Page_099.txt

petty_h_Page_086.txt

petty_h_Page_140.txt

petty_h_Page_147.txt

petty_h_Page_130.txt

petty_h_Page_115.txt

petty_h_Page_095.txt


Full Text





EFFECTS OF LOW AND HIGH pH ON LIPID OXIDATION IN SPANISH MACKEREL


By

HOLLY TASHA PETTY













A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007


































2007 Holly Tasha Petty

































To all with a passion for knowledge and the compassion to share it.









ACKNOWLEDGMENTS

I thank my supervising committee chair (Hordur G. Kristinsson) for the exceptional

privilege of conquering a future world problem, for his mentorship, for his friendship, and for

giving me the courage to realize my potential and dreams. I thank my supervisory committee

(Steve Talcott, Marty Marshall, and Charlie Sims) for their valuable expertise, instruction, and

revolutionary contributions to the academic field of food science. I thank the external member on

my committee (Christian Leeuwenburgh) for helping me realize my passion for applied

physiology and kinesiology and for providing worldly understanding of free radicals, aging, and

life. I especially thank Wei Huo for his statistical expertise and generous help. Others in my

graduate department who especially provided me with support include Maria Plaza, Dr. Ross

Brown, Susan Hillier, Carmen Graham, Bridget Stokes, Maria Ralat, Marianne Mangone, and

Walter Jones.

I would like to thank all of my family and many influential friends for encouraging me to

fulfill this dream. I specifically thank Andrew Piercy for his kindness and patience and Sean

McCoy for his unconditional inspiration and contribution. I thank all of my lab peers for lending

me knowledge, support, tears, and smiles. I thank Captain Dennis Voyles and his family for their

friendship and the Spanish mackerel capture of a lifetime. I thank those I have not specifically

acknowledged or who crossed my graduate path, for their positive influences and support. Lastly,

I am thankful, for the true lessons I have learned, and irreplaceable treasures I have earned.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IS T O F T A B L E S ................................................................................. 9

LIST OF FIGURES ................................... .. .... .... ................. 10

A B S T R A C T ............ ................... ............................................................ 12

CHAPTER

1 INTRODUCTION ............... ................. ........... .............................. 14

2 L ITE R A TU R E R E V E IW ........................................... ......... .................. ........................... 16

O obtaining a Fish Protein Isolate ........................................... ....................................... 16
Conventional Surimi Processing .................................. .....................................16
A cid/A lkali-A ided Processing...................... ...................................... ............... 17
Advantages of the Acid/Alkali-Aided process ............................................ 17
Lipids in Fish M uscle ................................... .. ........... .. ............19
L ipid C om position ...................... .. .. ......... .. .. ................................................. 19
Lipid Oxidation in Fish............. ..... ...................................20
Autocatalytic M mechanism .............................................................................. 20
Transition Metals: Catalysts of Lipid Oxidation...................................................21
Heme vs. Non Heme Iron and Lipid oxidation .................................... ............... 21
H em e Proteins and A utoxidation...................................................... .......................... 22
Hemoglobin vs M yoglobin and Lipid Oxidation......................................... ............... 23
Lipid Content and Lipid Oxidation................................................................................ 23
Effects of Extreme pH on Lipid Oxidation in Fish Muscle ................................. 24
Lipid Oxidation and Antioxidant Implementation .............................................. 25
L ipid oxidation and P protein s ........................................ .............................................26
Protein Oxidation ................................................ 27
M echanism s of Protein Oxidation ........................... ..............................................28
P protein S mission ................................................................2 9
Protein Polym erization ..................................... .. ........... .. ............29
Protein Carbonylation ...... .... ...... ........................................ ..... ............. 30
Lipid M ediated Protein O xidation ...................... .... ......... ........................ ............... 31
Protein O xidation A ccum ulation ...................... .... ............... ........................ ............... 31
Protein Oxidation and Muscle Food Quality ........... ........... ............. ...............32
Protein oxidation and conformational changes in proteins ..........................................33
Functional properties and protein oxidation......................................... ............... 34
S ig n ifican ce .................................................................................3 6

3 EFFECTS OF ALKALIACIDIC ADJUSTMENT ON LIPID OXIDATION IN
SPANISH M ACKEREL HOM OGENATE ................................................ .....................37









Introdu action ................ ......... .............. .............................................37
M materials and M methods ..................................... ... .. .......... ....... ...... 38
R aw M materials ................................ ......................................................3 8
C hem icals ......................... ............... ...................................... 39
Storage Studies and Preparation of Homogenate ............. ... .............................39
A antioxidant Im plem entation ......... ......... ......... .......... ........................... ............... 40
Total Lipid Analysis ................................... .. ........... .. ............40
Lipid Hydroperoxides........................ ... ....... ..... ............. 40
Thiobarbituric Acid Reactive Substances (TBARS)................................... .................41
A analysis of M moisture and pH ....................................................................... ............... 41
Statistical A nalysis................................................... 41
Results ........................... .................................42
Changes in Spanish Mackerel Lipid Oxidation................... ....... ............. ........ ..42
W after and lipid Content................. ............ ........ ...... ................ 42
Lipid Oxidation in Whole and Minced Spanish Mackerel ............................................. 43
Lipid Oxidation in pH Treated Spanish Mackerel Homogenates.............................43
Changes in homogenate lipid oxidation with storage................................................44
Changes in lipid oxidation with pH and storage .................................. ............... 45
Changes in lipid oxidation with initial quality and pH............................................. 45
Changes in lipid oxidation with holding time and pH................... ........... ................... 46
O overall pH effects ................... ....... ...... ..................................... ............ 47
Lipid Oxidation of Homogenates with Antioxidant Implementation............... ........... 47
Lipid Oxidation of Homogenates after Readjustment to pH 5.5 and 7 ..............................49
Batch and homogenate pH readjustment ................................... .................50
Time and homogenate pH readjustment.................................... .....................51
Discussion ........................... ... ............. ... ...... .. ............. ............... 51
Lipid Oxidation and Storage Stability ................... ......... ..................... 51
Lipid Oxidation in Acid/Alkali Adjusted Homogenates ................................................52
H om ogenate pH R eadjustm ent....... ......... ......... .......... ....................... ............... 58
A antioxidant Im plem entation ......... ................. ............................................................. 60
EDTA and Citric Acid Implementation at Low pH..................................... ............... 64
C conclusion ................ ..... ..... ......... .. ............................................65

4 THE EFFECTS OF THE ALKALI/ACIDIC AIDED ISOLATION PROCESS ON
LIPID AND PROTEIN OXIDATION ........................................................ ............... 75

In tro du ctio n ................... ...................7...................5..........
M materials and M methods ................................... ... .. .......... ....... ...... 76
R a w M ate ria ls ............................................................................................................ 7 6
C hem icals .................. ................. ........ ... .................. 76
Storage studies and Preparation of Homogenate and Isolate .......................................77
Thiobarbituric Acid Reactive Substances (TBARS).................................................77
Protein Quantification and Characterization ........................................ .....................78
Rheology ............................... ............ ...................... ......... 79
Protein Oxidation ............................................... 79
A analysis of M moisture and pH ........... .............. ...................... ................. ............... 80
Statistical A analysis ............................................. 80


6









R e su lts ...................................... ................... ....................................... .. 8 1
Oxidative Stability of Acid/Alkali Derived Protein Isolates..................................81
G el Form ing A ability of Isolates.................... ........................................... ........84
Protein Composition of Homogenates and Recovered Isolate Fractions ........................84
D iscu ssio n ................... .... ... ............................ .................... ................ 8 5
Changes in pH of M ince w ith Storage ....................................................... ...................85
Lipid Oxidation During Extreme pH Adjustment and Readjustment to 5.5 ...................86
P ro te in O x id atio n ....................................................................................................... 8 9
G el F orm ing A ability ........................ ........... ........................... ............... ............. 92
Textural Changes and Protein and Lipid Oxidation.................................................95
S D S p a g e ................................................................................................................... 9 7
C o n clu sio n ................... ...................9...................8..........

5 EFFECT OF ACID PROCESSING ON LIPID OXIDATION IN A WASHED
SPANISH MACKEREL MUSCLE MODEL SYSTEM ............................................. 107

Introdu action ................ .... ..... ......................................................................... 107
M materials an d M eth od s .............................................................................. ..................... 10 8
Raw Materials .................................................108
C h e m ic a ls ......... ........................................................................................... 1 0 8
W washed Fish ...................................... ....................108
Preparation of Spanish Mackerel Hemolysate ........................................ .....109
Hemoglobin Quantification .............. ......... ........ ........110
Spanish mackerel Pressed juice ................................... ...................... ............. 110
Thiobarbituric Acid Reactive Substances (TBARS) .........................110
Spanish mackerel Pressed Juice protein Quantification and Characterization ............110
Preparation of Low Moisture Washed Spanish Mackerel Mince Oxidation Model
System ................ ........... .... ................. ..... .....................111
Preparation of Hemoglobin and Pressed Juice Oxidation Model System ..........1......12
Oxygen Radical Absorbance Capacity (ORAC) ...............................................113
Total Antioxidant Potential ............... .........................114
Moisture Content and pH .................. .. ......... ........ .........115
Statistical A naly sis ......... ............ ............................................... 115
R e su lts .......... ..... .... ... ... ..................................................... 1 1 6
One and Two Way adjustment .............. ............... .... .........117
Two and Three Component Adjustment ........... ........... .............. ............ 118
Comparison of one way and multi-component pH adjustment in Washed System ...... 118
Antioxidant mechanisms of non-substrate components WFM model system ...........1...19
O R A C T est ................ ............................... .....................119
TEA C A ssay .............. ............................................................................ 120
SD S Page ................... .................................................. ................... 121
D iscu ssion ............... ............. .... ....... ......................... ...............12 1
No pressed juice treatment (control without pressed juice) ............... ..................122
Lipid oxidation in the control sam ples ............................................. ......... ...... 123
The Effect of pH Adjustment on Tissue Components................................................125
The pH effect on hem oglobin.......................................................................... 125
Hemoglobin adjustment to low pH and readjustment to pH 6.8 ...................................127


7









Pressed Juice adjustm ent ................... ....................................................... ............... 129
Lipid oxidation in the adjusted washed fish muscle..........................................131
Antioxidant Capacity as related to oxidation in washed muscle................................135
C o n clu sio n ................... ...................1...................3.........9

6 CONCLUSION................ ..... .. .......... ........... ............... .. 145

L IST O F R E F E R E N C E S .................................................................................. ..................... 14 8















































8










LIST OF TABLES


Table page

3-1 Initial average mince values for moisture content, pH, and lipid content .......................67

4-1 Moisture content (%) and protein content (%) of isolates ...........................................99









LIST OF FIGURES


Figure page

2-1 Conventional surimi processing schem atic................................... ......................... 16

2-2 Flow schematic of acid and alkaline process ............................... ................................. 18

3-1 Lipid oxidation in Spanish M ackerel mince .............................. ................................ 67

3-2 Development of lipid oxidation in homogenates, freshly prepared..............................68

3-3 Development of lipid oxidation in homogenates after 1 day of storage............................69

3-4 Development of lipid oxidation in homogenates after 2 days of storage ..........................70

3-5 The effect of ascorbic acid and tocopherol on pH 3 and pH 6.5 homogenates. ................71

3-6 The effect of EDTA and citric acid on acid homogenates............... ..........................72

3-7 The effect of pH readjustment in homogenates at pH 3 ..............................................73

3-8 The effect of pH readjustment in homogenates at pH 11. ............. ..................... 74

4-1 Schematic of the observed centrifugation layers .......................................... ................99

4-2 Lipid oxidation of acid, alkaline, and control isolates after one hour ............................100

4-3 Lipid oxidation acid, alkaline, and control isolates after 8 hours ..................................100

4-4 Lipid oxidation of acid, alkaline, and neutral isolates, 1 and 8 hours and 3 days ...........101

4-5 Levels of protein oxidation after 1 hour and 8 hrs pH, and 3 days..............................101

4-6 Storage Modulus (G') of isolate pastes derived from 1 and 8 hrs, and 3 days ..............102

4-7 Examples of rheograms from acid (pH 3) protein isolates ......................... ...........102

4-8 Examples of rheograms from alkaline (pH 11) protein isolates ............... ...............103

4-9 Examples of rheograms from control (pH 6.5) protein isolates................................... 103

4-10 Protein composition of isolates from adjusting for 1 hour ...........................................104

4-11 Protein composition of isolates from adjusting for 8 hour ...........................................104

4-12 Protein composition of isolates, 8 hours of holding followed by 3 days......................105

4-13 Protein com position of initial hom ogenates ........................ .................. ...................105









4-14 Protein composition of homogenates held for one hour ................ ............... 106

4-15 Protein composition of homogenates held for 8 hours ................................................106

5-1 The role of acidified (pH 3) and pH6.8 tissue components on lipid oxidation..............141

5-2 The role of two and three acidified (pH 3) and pH (6.8) tissue components.................. 142

5-3 Antioxidant capacity of Spanish mackerel hemoglobin and pressed juice......................143

5-4 Protein composition of untreated Spanish mackerel pressed juice ..............................144









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EFFECT OF LOW AND HIGH PH ON LIPID OXIDATION IN SPANISH MACKEREL

By

Holly Tasha Petty

May 2007

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

Because of over fishing, traditional sources cannot satisfy the demand for fish proteins.

The production of protein isolate from underutilized fish sources using a new high and low pH

process is being investigated. Underutilized fish sources include pelagic species, by-catch, and

fish byproducts. The objective of this study was to investigate how the acid/alkali-aided process

influences lipid oxidation development in homogenates, isolates, and washed fish prepared from

Spanish mackerel muscle.

Studies of homogenates at acid and alkaline pH showed that lipid oxidation increased with

age of raw material, mince cold storage, pH exposure time, and acidification. Lipid oxidation in

the acid homogenates was effectively inhibited by ascorbic acid and tocopherol, but was

unaffected by the addition of chelators. The pH readjustment from acid to neutral pH did not

significantly affect lipid oxidation in Spanish mackerel homogenates; however, the addition of

antioxidants in combination with pH readjustment was problematic.

Isolates prepared from acid and alkaline treatment had comparable levels of lipid

oxidation, regardless of homogenate holding time. High levels of lipid oxidation were observed

in the supernatant fractions from acid, alkaline, and control treatments. Overall, alkaline isolates

contained greater levels of protein oxidation compared to acid and alkaline isolates. Protein









degradation occurred during acid and alkali processing of Spanish mackerel. Rheological testing

showed that there was greater gel strength in alkaline isolates, followed by the control and acid

isolates. Protein oxidation could explain differences in gel quality between acid and alkaline

isolates.

The adjustment from pH 3 to pH 6.8 greatly affected all components in a hemoglobin-

mediated lipid oxidation washed Spanish mackerel model system. The addition of pH treated

hemoglobin to the washed Spanish mackerel system resulted in lipid oxidation levels comparable

to the control. Increases in lipid oxidation compared to the control were observed when pressed

juice was pH treated. The greatest levels of lipid oxidation in the Spanish mackerel model

system occurred when washed fish muscle was pH treated. These results show that lipid

oxidation incurred during acidic processing could be primarily due to phospholipid changes,

secondary to changes in hemoglobin and/or aqueous extracts.









CHAPTER 1
INTRODUCTION

Because of the large demand for fish protein and inability to satisfy the need by traditional

means, the utilization of fatty pelagic species and by-catch is under current study. By-catch from

shrimp vessels, pelagic species, and by-products are underutilized as human food (1). By-catch

includes low value fish that are caught and discarded back into the sea or that are further used for

fish meal (2). It is estimated that four pounds of by-catch result for every pound of shrimp

caught. In addition, dark muscle fish species compose 40-50% of total catch worldwide (3, 4)

Pelagic species present numerous obstacles if processed conventionally, due to small size, shape,

bone structure, seasonal availability, and most importantly the prevalence of unstable lipids and

pro-oxidants (4-6).

Extracting functional and high quality fish proteins from fish sources by solubilizing the

proteins at low and high pH values and recovering solubilized proteins with isoelectric

precipitation has been conducted (7). From these findings, isolation of functional proteins from

underutilized, low value fish sources was developed (7, 8). These newly developed processes

produce a protein isolate low in lipids. This is of major importance, due to the abundance of

triacylglycerols (neutral storage lipids) and membrane phospholipids in mitochondria of the dark

fish muscle (5). Phospholipids are the primary substrates in lipid oxidation due to their highly

unsaturated nature (9, 10).

Protein isolates made from white muscle using the alkali process have been found to

possess greater oxidative and color stability over unprocessed or acid treated muscle (11).

Studies have shown that high pH inactivates hemoglobin, thereby preventing oxidation of lipids

and formation of an undesired yellow-brown color as well as off odors and flavors. In contrast,

acid treatment denatures hemoglobin, enhancing its pro-oxidative activity (12). Pelagic fish









species and fish byproducts are rich in blood and heme proteins. The blood and heme proteins

are used for oxygen transport for production of long-term energy in the dark muscle via

oxidative metabolism. In white muscle, heme proteins are used for more short-term energy (5).

A number of other changes in homogenized fish muscle can occur at low and high pH

values that can contribute to the acceleration or delay of lipid oxidation. Currently there is a

limited understanding of how low and high pH influences oxidative processes in fish muscle.

Further investigation is therefore needed regarding the stability of fish muscle as affected by

extreme pH. Due to the abundance of heme proteins in pelagic fish species, it is imperative to

study their oxidative stability and to control their pro-oxidative effects. These proposed studies

could further the implementation of pelagic fish protein isolates as a food source and functional

food ingredient, and an understanding of oxidation mechanisms in general.










CHAPTER 2
LITERATURE REVEIW

Obtaining a Fish Protein Isolate

Conventional Surimi Processing

The production of surimi has proved to be somewhat useful in utilizing fish proteins.

Briefly, fish muscle is ground and washed at approximately 1 part fish to 2-5 parts water. The

washing step can be repeated several times depending upon processing conditions (i.e. type of

fish and plant location). During the washing step, 30% of the total proteins in the form of

sarcoplasmic proteins are lost. The fish muscle is then dewatered using a screw press. Normally,

cryoprotectants are added to prevent protein denaturation during freezing and thawing. This

process (Figure 2-1). Unfortunately, high quality surimi from conventional processing has only

resulted from white flesh fish. Attempts to create surimi from dark pelagic species has resulted in

products having poor color and lipid stability due to pigments, pro-oxidants, and an abundance of

lipids (4, 6).

Ground Fish Muscle



Wash with 2- 5 parts water
(3 times with 0.3% NaC1 in last step)


Refine and Dewater



Stabilize
(Add crvoprotectants and freeze)


Surimi



Figure 2-1. Conventional surimi processing schematic. The final surimi is heated to form a gel,
resulting in various seafood analog products (13).









Acid/Alkali-Aided Processing

Solubilization and recovery of protein by pH rather than salt adjustment, led to an

economical protein isolate from low value fish sources (7). At high and low pH values, the

proteins become solubilized due to large net charges. It was also discovered that at extreme pH,

the cellular membranes surrounding the myofibrillar proteins are disrupted (13, 14). As a

consequence of these changes in protein solubility and membrane integrity, the viscosity of a

muscle protein homogenate at very low and high pH values was found to decrease significantly.

This viscosity decrease enables soluble proteins to be separated from undesirable materials in the

fish muscle via high speed centrifugation (e.g.bones, scales, connective tissue, microorganisms,

membrane lipids, neutral lipids etc.). The soluble proteins can then be recovered and

precipitated by adjusting the pH into the range of their isoelectric point (-pH 5.3 to 5.6) (Figure

2-2). This process uses a very low (2.5 to 3.0) or high pH (10.8 to 11.5), thus enabling for a

simple and economical recovery of low value fish species (7, 8, 15). Studies have demonstrated

that of the two processes, the alkali-aided process is substantially better than the acid-aided

process since oxidation and color problems are reduced and protein functionality is superior (13,

16, 17).

Advantages of the Acid/Alkali-Aided process

Using the acid/alkali-aided process versus conventional processing is beneficial for various

reasons. Protein recoveries can range from 90-95% using the acid/alkali-aided process versus

conventional surimi recoveries ranging from 55-65% from fish fillets. Conventional processing

only retains the myofibrillar proteins, in contrast, this new process allows for the recovery of

both myofibrillar and sarcoplasmic proteins. Processing is more economical using the

acid/alkali-aided process, reducing labor, expenses, production complexity, and handling









homogenatee compatibility in processing). Additionally, minimally processed and lower quality,

less utilized


Figure 2-2. Acid and alkaline process used to produce protein isolate from low- and high-value
fish protein sources.

sources of fish protein can be implemented. Before processing, fish sources can be of lower

quality and can be previously frozen. Lipids can be less of problem due to the separation of

triacylglycerols and phospsholipids from the protein homogenate, enabling the production of a

more stable product with better color and flavor; however, there is not absolute separation of

heme proteins from the final protein isolate. Due to the abundance of mitochondria in dark

muscle, the acid/alkali process is advantageous in the separation of phospholipids from dark

muscle (5). Functional properties including gelation, water holding capacity, and solubility are

all improved during the acid/alkali-aided process due to a desirable partial denaturation of the









fish proteins. Overall, the protein isolate produced is safer due to elimination or reduction of

toxins (PCBs and mercury) and microbial loads, and no pollution due to processing (6, 13, 16,

18, 19).

Lipids in Fish Muscle

Lipid Composition

For dark muscle fish species, fat composition can vary in response to environment,

seasonal variation, migratory behavior, sexual maturation, and feed (20). Fish living in colder

water will have more dark muscle. Dark muscle tends to vary in composition, especially with

season and location; therefore, when sampling dark muscle or fatty fish species, it is important to

consider the possible variations in lipid content (20).

The two types of lipids contained within teleost (bony) fish muscle are phospholipids and

triacyglycerides. Phospholipids or structural lipids are the major components of cell membranes

within the fish tissue (myofibrillar protein); while triaglycerides are found in adipocytes and

serve as storage energy. Phospholipids are found in various membranes including those of the

cell and organelles within the cell.

In general, muscle from lean fish is composed of approximately 1% lipid, primarily

phospholipid. The majority of lipids in lean fish are stored in the liver. Some lean species

include the bottom-dwellers: cod, saithe, and hake. Major lipid deposits in fatty fish species are

found in tissue other than the liver: subcutaneous tissue, in the belly flap, and throughout the

muscle tissue. Fatty species include the pelagic species such as herring, mackerel, and sprat.

Most importantly, lipids are widely distributed within muscle tissue, primarily next to the

connective tissue (myocommata) and between light and dark muscle (21). In pelagic species, the

dark muscle uses triacylglycerides for continuous aerobic energy in migrations, whereas white









muscle uses glycogen as the main source of energy and is quickly fatigable due to accumulation

of lactic acid during anaerobic respiration (22).

Lipid Oxidation in Fish

Lipid oxidation serves as a mediator of important processes in living biological systems

(23). However, post-mortem lipid oxidation can only be controlled to a certain extent, before

antioxidants and their systems are exhausted (24). Lipid oxidation is one of the primary causes

of deterioration in cold stored fish muscle (25) and negatively affects color (26), odor and flavor

(27), protein functionality and conformation (28), and overall nutritional content of fish muscle

(29, 30). This quality deterioration is due to the high content of polyunsaturated fatty acids

contained within fish muscle (9, 10, 25) along with highly active pro-oxidants (5).

Major reactants in lipid oxidation are oxygen and unsaturated fatty acids. Exposure of

lipids to atmospheric oxygen results in unstable intermediates (31). Ultimately, these

intermediates then degrade into unfavorable compounds attributing to off flavor and aroma.

Lipid oxidation has many plausible mechanism of initiation, including non-enzymatic and

enzymatic reactions. Non-enzymatic mechanisms include autoxidation (free-radical mechanism)

and photogenic oxidation (singlet oxygen mediated) (31). Enzymatic mechanisms include

actions by lipoxygenase and cylooxygenase. Lipid oxidation most commonly occurs by a free

radical mechanism, involving the formation of a reactive peroxyl radical (31). In addition, it has

been reported that the detrimental effects due to heme (autoxidation mechanisms) are greater

than effects due to lipoxygenase, mainly due to longer-lasting pro-oxidative activity (32).

Autocatalytic Mechanism

The autocatalytic mechanism is deemed the primary cause of lipid oxidation in post

mortem fish (31) and is historically referred to as lipid peroxidation (33). The free radical

mechanism has three classic stages: initiation, propagation, and termination. Initiation begins









with hydrogen abstraction from an unsaturated fatty acid, particularly those having a pentadiene

structure (31). The lipid radical (Le) then reacts with molecular oxygen and a peroxyl radical

results (LOOe). Propogation is the phase in which the peroxyl radical (LOO*) abstracts

hydrogen from neighboring lipid molecules, thereby forming more lipid radicals and lipid

hydroperoxide (LOOH) molecules. Lipid hydroperoxides cannot be detected through sensory

analysis, which is the reason why peroxides do not correlate with "off flavor". The degradation

of these lipid hydroperoxides, due to metal catalysts, results in the unfavorable secondary

intermediates of shorter chain length: ketones, aldehydes, alcohols, and small alkanes. The

secondary products give rise to off aromas, flavors, and yellow color in fish. Aldehydes are

normally associated with off odor. Termination occurs when a radical reaction is quenched by a

reaction with another radical or antioxidant. It is important to note that these reactions are

occurring simultaneously and that the mechanism of lipid oxidations can become quite complex.

(22, 31).

Transition Metals: Catalysts of Lipid Oxidation

Decomposition of lipid hydroperoxides at refrigerator and freezing temperatures is due to

electron transfer from metal ions. Transition metals serve as primary oxidation catalysts in lipid

oxidation, mainly iron. Iron content varies among muscle food sources (31). The main catalysts

in biological systems include heme and non-heme iron (5). Non-heme sources include released

iron from heme proteins, protein complexes (transferritin, ferritin, and metalloprotiens), and

other low molecular weight complexes (diphosphonucleotides, triphosphonucleotide, or citrate

chelates) (34).

Heme vs. Non Heme Iron and Lipid oxidation

White muscle contains less iron than dark muscle, due to less myoglobin and hemoglobin

(5). Through research, it has been shown that heme proteins serve as potent pro-oxidants (7, 35).









Further more, it is believed that heme proteins serves as the main pro-oxidant in post mortem

fish muscle (36, 37). It is believed that hemoglobin is the main initiator of oxidation in vitro

systems (12, 38); whereas, hydrogen peroxide mediated metmyoglobin has been shown to

initiate oxidation in vitro-model systems (39, 40). In addition, Kanner et al (41) found that low-

molecular weight iron also initiates lipid oxidation. Nonetheless, studies have shown that

hemoglobin is the leading pro-oxidant (16, 35, 42, 43). The basic mechanism of lipid oxidation is

known; however, the predominant initiators(s) of oxidation are still debated.

Heme Proteins and Autoxidation

If most of the hemoglobin is not removed through bleeding, the fish muscle will become

more susceptible to oxidation. Mincing or mechanical mixing of fish muscle can promote

oxidation through the incorporation of hemoglobin, hemoglobin subunits, free heme, and iron.

(38). The presence of hemoglobin and/or myoglobin during the acid and alkali-aided process and

in the end product (protein isolate) could lead to serious lipid oxidation (42, 44, 45).

Whether oxidized or reduced, heme proteins can serve as potent pro-oxidants (46). Post

mortem, heme proteins are in their ferrous state (Fe+2). In this state, oxygen can be bound or un-

bound to the heme thus giving oxyhemoglobin/myoglobin and deoxyhemoglobin/myoglobin,

respectively. The heme iron can become oxidized to the ferric state (Fe+3) and result in

metmyoglobin/hemoglobin, with water binding to the 6th coordination site (47). This oxidation

will also give rise to the superoxide anion. Subsequent superoxide dismutation can result in the

highly reactive hydroperoxyl radical, hydroxyl radical, and hydrogen peroxide (6, 24)

Autoxidation of hemoglobin or myoglobin by superoxide can lead to the formation of the

hypervalent ferryl-hemoglobin/myoglobin radical. The formation of this porphyrin cation radical

is deemed the primary decomposer of lipid hydroperoxides, resulting in the secondary

compounds, associated with off-flavors and odors (37, 48).









Hemoglobin vs Myoglobin and Lipid Oxidation

Myoglobin is located in the muscle cell itself while hemoglobin is found in the blood cells

of the muscle tissue vascular system (49). Hemoglobins from various fish species are

differentiated from how they react to changes in temperature, pH, etc (50). It has been found that

with decreases below pH 7.6 (43) and as low as pH 1.5 (51), either through physiological

changes post mortem and/or experimental manipulation, a shift from oxy to deoxy-hemoglobin

occurs, called the Bohr shift. The change to the deoxy state is believed to increase hemoglobin's

pro-oxidative activity due to heme iron exposure (52). Physiologically, this phenomenon may be

advantageous in the release of oxygen in the production of pyruvate and consequently lactate, in

pelagic species due low oxygen affinity (53).

Recently, the comparison of myoglobin and hemoglobin from trout as catalyst for lipid

oxidation was presented by Richards and Dettmann(54) The researches found that hemoglobin

has a greater tendency to induce lipid oxidation due to the mechanism of heme dissociation

hversus myoglobin at neutral pH (55). The oxidative nature of myoglobin versus hemoglobin at

low pH is being investigated.

Lipid Content and Lipid Oxidation

According to Richards and Hultin (56), only low levels of phospholipid (0.01%) are

required for hemoglobin mediated lipid oxidation and for off flavors to develop. According to

Undeland et al. (57), it is the presence and amount of hemoglobin that affects the extent of

oxidation and not the amount of lipid. In a study conducted by Richards and Hultin (58), even a

six-fold increase in membrane phospholipids did not affect the extent of lipid oxidation. Despite

bleeding, hemoglobin still remains within the fish muscle and can still pose problems (59).









Effects of Extreme pH on Lipid Oxidation in Fish Muscle

At extreme acidic pH, heme proteins are denatured, increasing the propensity of lipid-pro-

oxidant interactions. Lipid membranes become more susceptible to oxidation when in the

presence of heme proteins adjusted to low pH or readjusted from low pH (3.0) to neutral pH (7)

(18). This is due to unfolding of the heme protein and exposure of the heme iron upon

denaturation, or lack of refolding with pH readjustment (51). In addition, transition metals and

free heme become active at low pH (56). Autoxidation and ferryl-Hb/Mb production are

enhanced at low pH (56, 60). Furthermore, the heme crevice of hemoglobin is unfolded at low

pH, increasing reactions with lipid substrates, and thus oxidation (12). With acidic pH, there is

also a greater likelihood of subunit dissociation of the hemoglobin and thus loss of the heme (51,

61). In contrast, the pro-oxidative activity and autoxidation of hemoglobin is reduced at alkaline

pH due to increased stabilization of the heme group (51, 60).

Studies of the effect of extreme pH on hemoglobin and oxidation have been conducted in

model systems by Kristinsson and Hultin (51). Trout hemoglobin was added to a washed cod

model system. This model muscle system contain virtually all membrane phospholipids and

proteins, but is free of soluble pro- and antioxidants through the washing step. The ability of

hemoglobin to oxidize cod at pH values 2.5, 3.5, 10.5 and 11 was studied. It was found that there

was considerably more oxidation at low pH values compared to higher pH values.

Undeland et al. (62) studied the minimization of oxidation of herring muscle homogenate

at low pH (pH 2.7) by reducing exposure times, using high speed centrifugation, and

implementing antioxidants or chelators. Storage stability of the protein isolates was only

improved by antioxidant and chelator implementation, erythrobate and

(ethylenediaminetetraacetic acid) EDTA, respectively (62).









Various techniques to reduce lipid oxidation exist and include heme separation, lipid

membrane centrifugation, pH exposure time, and antioxidant implementation. However, the

quantification of lipid oxidation in the homogenate made from fish of varying quality in certain

conditions requires attention such as extended holding times (12 hours) and an expanded pH

range (2.5 to 11.5).

Lipid Oxidation and Antioxidant Implementation

The greatest defense against lipid oxidation is the use of antioxidants. Inhibition of lipid

oxidation can occur directly or indirectly. Inhibitors can prevent initiation and propagation steps

in the free radical mechanism (31). Antioxidants occur exo- and endogenously. Antioxidants

donate hydrogen to the peroxyl radical and eliminate it as a reactive substance in the radical

mechanism (31).

Past research evaluated the effects of antioxidants on lipid oxidation in mackerel fillets

(63). Exogenous antioxidants included: ascorbic acid, tocopherol, and butylated-hydroxytoluene

(BHT). The antioxidants were applied separately and in combination. At frozen storage

temperatures (-200C and -300C), the BHT and BHT combined with ascorbic acid proved to be

the most inhibitive. This implies that there is a synergistic effect between lipid and water-soluble

antioxidants in delaying lipid oxidation in whole fish muscle (63).

Concentration effects of antioxidants are well known in vivo. At low ascorbic acid

concentrations, there is a reduction of transition metals (iron), promoting hydroxyl radical

formation, and at high concentration, ascorbic acid scavenges hydroxyl and superoxide radicals

in vivo (49). According to Witting (64) and Chow (65), tocopherol serves as a pro-oxidant at

high concentration and an antioxidant at low concentrations.

At low pH versus high pH, ascorbic acid and tocopherol can inhibit lipid oxidation.

Together, ascorbic acid, in high concentrations, and tocopherol effectively prevent oxidation in









vitro. Tocopherol by itself is especially effective in reducing lipid oxidation in vivo (66). In

herring fish homogenate (pH 3), there was a significant reduction in oxidation as measured by

malondialdehyde when implementing ascorbic acid and tocopherol (62). In addition, when

incorporating antioxidants erythrobate (iso-ascorbic acid) and STPP (sodium tripolyphosphate),

there was a significant reduction in lipid oxidation in the final protein isolate (67). According to

previous studies, it is essential that antioxidants be added early in the acid aided process to obtain

a good lipid stable final product (67-69).

According to Undeland et al. (67), low molecular weight aqueous extracts, also know as

"press juice," from fish muscle are effective in inhibiting lipid oxidation in cod lipid membranes

caused by hemoglobin. Possible inhibitors in the "press juice" could include nucleotides and

other reducing agents. The role of these compounds at low or high pH is not well understood.

Lipid oxidation and Proteins

The functionality of fish protein isolate preparations relate mostly to the behavior and

interaction between myofibrillar proteins (most notably actin and myosin). Sarcoplasmic and

myofibrillar proteins contribute to varying portions of the overall protein isolate, depending on

whether the alkaline or acidic process is used. Myofibrillar proteins are the primary components

of protein isolates produced from the alkaline process, and sarcoplasmic proteins are the primary

proteins in isolates prepared from acid processing (14).

The effect of lipid oxidation and reactive oxygen species (ROS), on protein stability and

function in muscle foods is poorly understood. Howell and Saeed (70) reported texture changes

in frozen Atlantic mackerel (Scomber scombrus) stored at 20 and 30C for up to 2 years, which

could be related to an increase in lipid oxidation. Increases in G' (elastic) and G" (viscous)

moduli were observed. With the implementation of antioxidants in the muscle, a correlation

between the kinetics of lipid oxidation and theological behavior was observed. In another study









by Saeed and Howell (63), it was reported that lipid oxidation with frozen storage caused a

reduction in ATPase activity, solubility, and myosin heavy chains. The incorporation of

antioxidants in Atlantic mackerel prevented/delayed the deleterious effects caused by lipid

oxidation. Hitherto, studies have not addressed the effects of acid and alkali processing on

protein oxidation development.

Protein Oxidation

The relationships between muscle food quality and protein oxidation are of great interest

and have yet to be fully understood. Consequence of oxidative stress, proteins undergo covalent

modification and incur structural changes. Reported protein modifications include and are not

limited to back bone disruption, cross-linking, unfolding, side-chain oxidation, alteration in

hydrophilic/hydrophobic character and conformation, altered response to proteases, and

introduction of newly formed reactive groups, or amino acid conversion (71-74). Although

protein functionality and quality changes are commonly observed with increases in lipid

oxidation, the actual mechanisms by which proteins are oxidized in muscle foods are not well

understood..

Extensive research shows a positive correlation between protein oxidation and diverse

physiologically deteriorative events, such as aging and disease. Biological events are highly

dependant upon radicals and oxidative processes (75). Oxidative stress arises when antioxidant

mechanisms are unable to counteract oxidant loads (76). Oxidative stress is secondary to partial

and/or complete exhaustion or degradation of antioxidant systems (24). In the event of oxidative

stress, generated radicals and non-radical species can oxidize proteins directly or indirectly. Due

to oxidatively reliant metabolism in vivo, direct protein oxidative attack occurs most frequently

in the presence of reactive oxygen species (ROS): hydrogen peroxide, superoxide, nitric oxide,

peroxynitrite, hydroxyl radical, peroxyl, and/or alkoxyl radicals, etc. (77). Similarly, these ROS









are produced in muscle foods during mechanical processing and storage (78-80). Indirect protein

oxidation can occur in the presence of secondary lipid and protein oxidation products (81-83).

Oxidizing agents common to living and post mortem muscle could include and are not

limited to transition metals, protein-bound transition metals (Heme), U.V., X, X irradiation,

ozone, lipid and protein oxidation products, and oxidative enzymes (84, 85). Neutrophils,

macrophages, myloperoxidase (hypochlorous acid), and oxireductase enzymes are just a few

possible mediators specific to in vivo protein oxidation (84, 85). Rational for oxidative

mechanisms post mortem are extrapolated from those which occur physiologically. The

introduction of oxygen during muscle food processing is analogous to reperfusion in vivo (86).

Pro-oxidative environments present themselves in the aging and restructuring of meat

products. With storage, endogenous antioxidant levels decrease in whole muscle; whereas, an

increased exposure to oxygen and endogenous oxidants as well as decreasing endogenous

antioxidants occur in processed muscle. Thus, radical species and derivatives induce greater

levels of oxidative stress with decreases in antioxidant capacity (87, 88).

Mechanisms of Protein Oxidation

A multitude of mechanisms give rise to protein oxidation. Oxidatively susceptible protein

sites in the presence of irradiation and/or metal generated ROS have been studied (89-92).

Structural protein units most prone to attack include the a-carbon and side chains of amino acid

residues as well as the peptide backbone. Amino acid residues most susceptible to oxidation in

vivo and in vitro are cysteine and methionine due to the presence of a readily oxidizable sulfur

atom and vulnerability to all oxidizing sources (76). Oxidation of cysteine can lead to disulfides

(S-S), glutathiolation, and thyl radicals (93, 94), while the oxidation of methionine

predominately leads to methionine sulfoxide (95). Oxidative modifications do not always result









in deleterious effects. In fact, negligible effects or antioxidative capabilities may arise with

protein oxidation (96-98).

Reactive oxygen species (ROS) can directly oxidize proteins by abstracting a hydrogen

atom from the a-carbon atom of an amino acid, producing a carbon centered radical. Further

reaction of these radicals with oxygen gives a peroxyl radical. Peroxyl radicals can then abstract

hydrogens or react with hydroperoxyl radicals (protonated superoxide), forming the alkoxyl

radical. A hydroxyl derivative can be produced from the reaction of the hydroperoxyl radical and

the alkoxyl radical (89-92).

In conjunction with protein radical formation, oxidation of proteins, peptides, and amino

acids can give rise to various protein derivatives. Protein oxidative derivatives common to both

physiological and post mortem tissue result from scission, polymerization, carbonylation, and

protein-lipid interaction. Specific oxidation modifications more common in vivo include

hydroxylation, nitrosylation, nitration, sulfoxidation, and/or chlorination of protein residues (99).

Protein Scission

Scission of peptide backbones and/or amino residues can occur with protein oxidation.

Alkoxyl radicals and alkoxyl peroxides can be cleaved through two major mechanisms: the a -

amidation pathway and the a -diamide pathway. Carbon radicals or derivatives are cleaved to

form amide and a-keto-acyl derivatives in the a -amidation pathway; while diamide and

isocyante derivatives form in the diamide cleavage pathway (89). According to Garrison (89) and

Uchida et al. (100) hydroxyl radical mediated cleavage/fragmentation of individual residues such

as glutamate and aspartate (89) and proline (100) can occur.

Protein Polymerization

Along with peptide cleavage, protein and/or amino radicals and derivatives can undergo

polymerization, forming cross-linked protein residues (82-85). Protein cross-linkage can occur









through covalent and non-covalent interactions. Non-covalent protein cross-linking is due to

hydrophilic and hydrophobic bonding. Oxidation of proteins can result in greater exposure of

hydrophobic groups, and thus, greater hydrophobic interactions among residues. Increased

interaction of hydrophobic residues can lead to aggregation. Likewise, hydrogen bonding can

facilitate protein and protein-lipid aggregation (101). Covalent intra- and inter-protein

polymerization are due to radical mediated mechanisms. ROS induced protein cross-linking in

vivo and in muscle foods can occur under several primary conditions: (1) homolysis of two

carbon, protein, or tyrosine radicals (89), (2) thio-disulfide exchange of oxidized cysteine

sulfhydryl groups (89, 102), (3) interaction of oxidatively derived protein carbonyl groups with

lysine/arginine residues (87, 103), (4) cross-linking of dialdehydes (malondialdehyde or

dehydroascorbate) with two lysine residues (104, 105)

Protein Carbonylation

Metal catalyzed oxidation (MCO) is a primary mechanism by which proteins are directly

oxidized in vivo to form carbonyl derivatives. Likewise, considerable protein oxidation is due to

MCO in muscle foods (106). In the vicinity of proteins and or amino acid residues, protein bound

transition metals illicit the production of ROS and oxidation of amino acid side-chains.

Hydrogen peroxide is reduced to ROS species (hydroxyl radical and ferryl radical) in the

presence of transition metals (107-110). MCO can lead to the formation of 2-oxo-histidine from

histidine and carbonyl derivatives from lysine, arginine, proline, and threonine (107, 111).

Carbonyl derivatives are frequently investigated in physiological and post mortem systems,

due to convenience and elucidation of overall protein oxidation. Carbonyl groups aldehydess,

ketones) are indicative of protein oxidization induced by various sources of ROS and oxidative

stress. Carbonyls can be detected in various ways and the assay involves the derivitization of

carbonyl groups with 2,4-dinitropheylhyrazine (DNP), producing a 2,4-dinitrophynl (DNP)









hydrazone (112). Following covalent derivatization, various quantitative techniques can be

utilized including and not limited to 2-D page, spectrophotometric methods, and high

performance liquid chromatography (HPLC) (112). Additionally, carbonyl measurement is

applicable to mixed samples: homogenate tissue, plasma, and proteins. Other methods require

purification, but allow for differentiation of the oxidative source (Shacter, 2000).

Lipid Mediated Protein Oxidation

Lipid mediated protein oxidation is believed to occur by various mechanisms (113). One

proposed mechanism describes a protein or amino residue radical and lipid oxidation product

(malondialdhyde) condensation reaction, forming a lipid-protein complex. Proteins could react

with lipid oxidation products alkyll radicals, hyderoperoxides, and secondary oxidation products)

giving rise to a protein-lipid radicals (63). Another mechanism for lipid-protein oxidation is

described: (1) free radicals react with proteins side chains producing protein free radicals, (2)

protein radicals could then react with oxygen and produce a peroxy radical, and (3) hydrogen

peroxide radicals formation could thus follow and induce formation of carbonyl products. The

extent of protein oxidation could theoretically develop faster than lipid oxidation, due to the

cytosolic associated location of proteins in muscle foods and possible exposure to free radicals

(Srinivasan and Hultin, 1995). Specific amino acids are highly susceptible to lipid induced

oxidation due to the presence of reactive side chains: sulfhydryl, thioether, amino group,

imidazole ring, and indole ring. Corresponding amino acids include cysteine, methionine, lysine,

arginine, histidine, and trypotphan (101, 114).

Protein Oxidation Accumulation

Physiologically, oxidatively modified proteins can accrue over time. Reduced activities

and/or levels of specific proteases, low rates of cell turnover, depletion of antioxidant

mechanisms, or inherently undetected oxidative modifications can lead to protein oxidation









accumulation in vivo (115). In instances where an amino acid conversion occurs, the cell is

unable to discriminate between oxidized and un-oxidized residues. MCO can lead to

modifications that go unnoticed including the conversion of proline to hydroxyproline, or the

conversion of glutamate to asparagine or aspartate (76). The cell predominantly relies upon the

protease known as the proteosome, to degrade oxidatively modified proteins (116). Methionine

sulfoxide reductase is one of the few repairing proteases, enabling for the conversion of

methionine sulfoxide to methionine (MET) in vivo (117). A decrease in proteosome activity

occurs with aging and is believed to be predominantly responsible for the accumulation of

oxidized proteins (106, 118, 119). Accumulation of protein oxidation products increases the

likelihood of cross-linking. Furthermore, polymerized proteins can not be degraded by the

proteosome and can prevent degradation of other oxidized proteins (73). Unlike living systems,

reparation measures for protein oxidation in post mortem tissue do not exist. Thus an

understanding of protein oxidation mechanisms in muscle foods and how functionality is

affected by protein oxidation is well warranted.

Protein Oxidation and Muscle Food Quality

Protein oxidation can induce physiochemical changes in proteins. Oxidative modifications

can lead to decreases and or loss in protein activity and stability, negatively impacting cellular

processes in vivo and functionality in vitro (120). In post mortem muscle foods, the intrinsic

property primarily affected by oxidative modification is conformation and/ or structure. The

functional property in muscle foods most affected by this intrinsic change is hydration.

Alteration in hydration capacity can ultimately affect other functionally properties: gelation,

emulsification, viscosity, solubility, digestibility, and water holding capacity (121-125). Due to

the significant contributions to muscle functionality, myosin is extensively studied in protein

oxidation studies (86). Numerous muscle food studies conclude that changes in protein structure









and functionality, under oxidizing conditions, are due to protein oxidation. Understanding the

cause of structural or functional protein changes as they relate to protein oxidation is important

to muscle food quality.

Protein oxidation and conformational changes in proteins

Changes in conformational stability in muscle food proteins is attributed to changes in free

energy. As a result of oxidative modification, intra- and intermolecular interactions are altered

within a protein's structure; thereby, decreasing the thermal energy and conformational stability.

Liu and Xiong (126) studied thermal stability changes in the Fe (II)/ascorbate/H202 induced

oxidation of myosin, derived from chicken pectoralis muscle. Using differential scanning

calorimetry (DSC), the protein oxidation by hydroxyl radicals and effects on myosin

conformational stability were measured. Significant reduction in myosin thermal stability was

observed as indicated by decreases in enthalpy (A H). Additionally, decreases in transition

temperature (Tm) were observed (126). The un-oxidized myosin transition temperature (550C)

designates the myosin conformational change to a randomly coiled, 'broken net,' structure upon

heating. Thus, decreases in Tm are indicative of decreased conformation stability (127). Relative

to the native myosin, a new transition (60-800C) temperature was observed among oxidized

samples upon heating. This newly formed heating transition was attributed to cross-linking

among oxidized cysteine residues as the Tmwas not observed among the oxidized samples

containing N-ethyl-maleimide, a thiol blocking reagent(126). According to Srinivasan and Xiong

(128), decreases in conformation stability were observed in bovine cardiac muscle myofibrillar

proteins when subjected to hydroxyl radical attack (128). Similarly, Saeed and Howell (70)

observed decreases in enthalpy (AH) and Tm in Atlantic mackerel fillets stored at -200C versus

those stored at -300C for 2 years. These changes were indicative of aggregation and denaturation

within the mackerel fillets (70). In another study conducted by Saeed and Howell (63),









denaturation of myofibrillar proteins and lipid oxidation of docosahexaenoic and

eicosapenaenoic acid were observed in frozen storage of Atlantic Mackerel and were

significantly inhibited by incorporations of antioxidants: namely BHT and ascorbic acid in a

minced system (63).

Increases in hydrophobic character and ultra violet absorption (U.V.) serve as markers of

protein destabilization and are presumed indicators of protein oxidation. With structural change

and denaturation, hydrophobic residues located within the interior of a protein become exposed.

Li and King (129) observed an increase in chicken myosin hydrophobicity in an iron (II)

ascorbatee mediated oxidation system (129). Wang et al.(130) observed an increase in

hydrophobicity in cold stored (-15 or -290C) beef heart surimi along with increases in lipid

oxidation and loss of sulfhydryl groups. Wang et al.(130) elucidated that increases in lipid

oxidation may cause or be mechanistically related to protein oxidation development. Past studies

have concluded that lipid oxidation induces protein oxidation; however, the mechanisms of

protein oxidation in the presence of lipids remain inconclusive.

Functional properties and protein oxidation

Increased hydrophobicity and cross-linking as a result of protein oxidation give rise to

decreased solubility. According to Buttkus (131) and Dillard and Tappel (132), significant

solubility decreases were attributed to malonaldehyde protein adduct products in muscle foods.

Gel electrophoretic patterns revealed protein aggregation with free-radical induced myosin

oxidation. These patterns were analogous to electrophoretic patterns for processed meat (88, 121,

133). Jarenback and Liljemark (134) observed extensive decreases (90%) in cod myosin

solubility in a linoleic hydroperoxide system. Significant decreases in solubility have been

observed in various other transition/metal mediated myosin (Turkey, beef) oxidation systems

(123, 133). According to Decker et al (133), transition metal/ascorbate induced protein oxidation









of turkey myosin resulted in decreased gelling ability, solubility, and hydration. Decreases in

functionality were attributed to increased levels of carbonyls and most likely cross-linking of

oxidized proteins (133).

Both a loss (135) and increase in gelling ability (121, 136) have been observed in the

oxidation of myosin. In a storage study conducted by Saeed and Howell (70), viscoelastic and

elastic modulus were measured in myosin prepared from matched Atlantic Mackerel fillets

stored at -20 C and -30C, for up to two years. The storage/elastic modulus (G') and viscous

modulus (G") were higher in the -200C versus the -300C derived myosin paste and were higher

with increased storage. This is most likely due to covalent and/or other interactions in

actomyosin (70). In this study, protein denaturation and aggregation correlated well with the loss

of ATPase activity and protein solubility shown in previous studies conducted by Saeed and

Howell (63). The elastic modulus was significantly decreased with the incorporation of

tocopherol (500 ppm), this was attributed to effective lipid oxidation prevention. Transfer of free

radicals or protein-lipid interactions could proceed from lipid oxidation (70). It was proposed

that protein G' and G" were reduced due to less lipid oxidation, induced aggregation. The

samples with greater lipid oxidation had greater levels of protein aggregation.. It was inferred

that the use of tocopherol would quench the radicals and reduce the formation of lipid and

protein oxidation product and thus reduce the gelling ability. Cross-linking can occur through

protein free radicals in muscle systems possibly causing aggregation and may improve gel

formation and strength (63).

Liu and Xiong (137) found that in an iron/ascorbate/H202 mediated myosin oxidation

system, gel elasticity was significantly reduced (40%) compared to the non-oxidized myosin.

This decrease in elasticity was attributed to protein oxidation. Myosin cross-linking and scission









were exibited by the presence of protein fragments and aggregates (137). In contrast to this

study, Srinivasan and Hultin (121) found that the iron/ascorbate induced oxidation of washed and

minced cod muscle resulted in increased gelling ability and carbonyl content. Hitherto, few

studies have correlated measures of protein oxidation with protein functionality and structure.

Significance

The development of lipid oxidation at low and high pH in fish homogenates and in protein

isolates obtained from the alkali-acid aid process has only been investigated to a limited extent.

This research could be very helpful in improving the stability of the homogenates and fish

protein isolates in processing and utilization of pelagic species, which otherwise are highly prone

to oxidation. In addition, the impact of lipid oxidation on proteins during the acid and alkali-

aided process has not been fully investigated. Contradictions exist as to what are the key

catalysts of lipid oxidation and very limited knowledge is available on the role of different key

catalysts or mediators of oxidation at low and high pH in muscle homogenates. This work will

further strengthen the argument that heme proteins are the prime mediators of lipid oxidation in

fish muscle homogenates and protein products thereof. Basic work in this area is expected to

greatly further our knowledge on oxidation processes at low and high pH and allow for effective

controls to minimize oxidation.









CHAPTER 3
EFFECTS OF ALKALI/ACIDIC ADJUSTMENT ON LIPID OXIDATION IN SPANISH
MACKEREL HOMOGENATE

Introduction

Current fish protein sources throughout the world dwindle and methods to efficiently use

existing fish supplies are well warranted. A novel acid/alkali-aided process enables for the

isolation of fish proteins from traditional and unconventional (by-catch and pelagic) sources (7,

8). In this process, fish is homogenized with water (1:9). The homogenate is then adjusted to low

pH (2.5 to 3.5) or high pH (10.5- 11.5) to solubilize the muscle proteins. Soluble proteins are

then separated from unwanted fractions (14) using centrifugation at 10,000g. The solubilized

proteins are then subjected to isoelectric conditions (pH 5.5 to 7.0) and are centrifuged at

10,000g. The final protein isolate (138) can then be used for various seafood analog products

(surimi) and ingredient applications (8).

Efforts to access the protein potential of low value fish sources are currently underway.

Conventional processing of pelagic species presents various challenges due to their small size,

shape, bone structure, and high levels of pro-oxidants and unstable lipids (4-7). The novel

alkaline/aided process, versus conventional techniques for the isolation of fish proteins, offers

many advantages including improved yield, functionality, stability, and color (6, 13, 16, 18, 19).

Enhanced protein stability arises from the reduction in lipids and pro-oxidants. Triglyceride and

phospholipids content is reduced through the acid/alkali-aided process; however, it is the

reduction in phospholipids and heme proteins that are attributed to lower levels of lipid oxidation

and increased stability (5, 9, 10).

Throughout the acid/alkali aided process two major reactants in lipid oxidation are most

important: hemoglobin and lipids. Hemoglobin is deemed the primary catalyst of lipid oxidation

in fish (12, 42). Hemoglobin undergoes various changes at low pH that could increase its









oxidative capacities (43, 51). Minimal amounts of lipid ( less than 1%) in the presence of

hemoglobin, are sufficient for lipid oxidation to occur (58, 139). Separating all lipid and

hemoglobin from the fish protein isolate is not feasible; thus, lipid oxidation is inevitable during

the acid/alkali-aided process and measures to delay and/or prevent lipid oxidation are necessary.

The effects of the alkali/acid-aided treatment on lipid oxidation in fish homogenate has

received limited attention. Undeland et al. (67) investigated how low pH (2.7) adjusted herring

homogenate was affected by pH exposure, centrifugation, and antioxidant addition. Addition of

erythorbate (0.2%) by itself or in combination with STPP (sodium tripolyphosphate) (0.2%) and

EDTA (0.044%) as well as high speed centrifugation of the homogenate significantly reduced

lipid oxidation. Thus far, lipid oxidation in homogenates at extended exposure times (up to 12

hours) and expanded pH ranges (2.5 to 11.5) have not been studied.

The objectives of this study were to investigate: (1) the effects of low and high pH on the

development of primary and secondary products of lipid oxidation in fish muscle homogenates,

(2) the effect of preformed lipid oxidation products in raw material on lipid oxidation

development in the homogenates at high and low pH, (3) the effectiveness of different

antioxidant preparations on lipid oxidation in fish muscle homogenates at low and high pH and

(4) the relevance of these homogenate findings to lipid oxidation at subsequent steps in the

alkaline/acid aided process.

Materials and Methods

Raw Materials

The first batch of Spanish mackerel (Scomberomorous maculates) was obtained from

Savon Seafood in May 2004 (Tampa, FL). Postmortem age of the fish was approximately 3 days.

Due to limited supplies of Spanish mackerel and inability to control for batch variability, the

second batch of Spanish mackerel was freshly caught in September 2005 by lab mates and









myself on Captain Dennis Voyles charter, Reel Therapy (Cedar Key, Florida). Fish were

manually skinned, gutted, and filleted. One half of the fillets were minced using a large scale

grinder (The Hobart MFG. Co., Model 4532, Troy, Ohio). Fillets and ground fish (100 gram

quantities) were vacuum packed and stored at -800C until needed.

Chemicals

Tricholoracetic acid, sodium chloride, chloroform (stabilized with ethanol), methanol,

ammonium thiocyanate, iron sulfate, barium chloride dihydtrate, 12 N hydrochloric acid, sodium

hydroxide, tocopherol, L- ascorbic acid, alpha-tocopherol, and citric acid were obtained from

Fischer Scientific (Fair Lawn,New Jersey). Thiobarbituric acid and 1,1, 3,3 tetraethoxypropane

were obtained from Sigma Chemical Co. (St. Louis, MO). Propyl gallate and

(ethlenediaminetetraacetic acid 99%), EDTA were obtained from Acros Organics (New Jersey,

USA).

Storage Studies and Preparation of Homogenate

Pre-frozen fish (-80C) were thawed in plastic bags under cold running water and

immediately placed on ice for further preparation. For the aging studies, fillets and/or mince

were held in one gallon zip lock bags at 40C. The homogenate was prepared by diluting the

mince with cold deionized water (1:9) and mixing (45 second at 40C, speed 3) using a

conventional kitchen blender equipped with a rheostat (Staco Inc., Dayton, Ohio, 3p 1500B).

The homogenate was aliquoted equally into 250 mL glass beakers. Homogenates were adjusted

to appropriate pH (2.5, 3, 3.5, 6.5, 10.5, 11, and 11.5) through drop-wise additions of2M HCL or

NaOH. The homogenates were lightly stirred over the 12 hour duration at 40C. To ensure that pH

values were maintained, homogenates were periodically checked and readjusted with drop-wise

amounts of 2 M NaOH or HCL.









Antioxidant Implementation

Antioxidants (tocopherol, ascorbic acid, citric acid, and EDTA) were thoroughly mixed

into the homogenate at 0.2% prior to pH adjustment. When used in combination, antioxidants

were added at 0.1% or 0.2% each prior to pH adjustment.

Total Lipid Analysis

Total lipid was measured in the minced muscle according to Lee et al. (140) using a 1:1

methanol/chloroform ratio. Results are reported as percentages based on wet weight.

Lipid Hydroperoxides

For designated time points, 3 grams of homogenate per each treatment was sampled for

lipid hydroperoxides. Samples were analyzed immediately. Lipid hydroperoxides were

determined according to Shantha et al. (141) with modifications by Undeland et al. (57). Total

lipids were extracted using 9 mL of a chloroform/methanol (1:1) mixture with vortexing. Sodium

chloride, 3 mL (0.5%), was then added to the sample. The sample was then vortexed for 30

seconds. The samples were centrifuged (Ependorf, model 5702, Brinkman Instruments, Inc.,

Westbury, N.Y) at 4 oC for 10 minutes at 2,000 g. To alleviate phase separation problems due to

pH, homogenates were adjusted to pH 6.5 before extraction with experimentally determined

volumes of IN NaOH or HCL. Two milliters of the bottom layer was obtained using a 2 mL

glass syringe and needle (Micro-Mate Interchangeble glass syringe, stainless steel needle, 20G

x18", Popper and Sons, Inc., New Hyde Park, N.Y). Two milliters of the chloroform layer was

mixed with 1.33 mL of chloroform/methanol (1:1). Ammonium thiocyantate (50 uL) was added

to the sample, and then iron chloride (94), 50 uL, was added to the sample. Following each

reagent addition, the mixture was vortexed for 2-4 seconds. After an incubation period of 5

minutes at room temperature (22.5C), samples were read at 500 nm using a spectrophotometer

(Agilent 8453 UV-Visible, Agilent Technologies, Inc., Palo Alto, CA). A blank contained all









reagents but the sample. A standard curve was constructed using a solution of iron (III) chloride.

Samples were kept in subdued light and on ice during the assay.

Thiobarbituric Acid Reactive Substances (TBARS)

TBARS were assessed according to Beuge et al. (142) as modified by Richards et al. (35).

A TCA/TBA solution (50%/1.3%) was freshly prepared and heated to 650C on the day of

analysis. It is imperative that the temperature not exceed 65C, to ensure that the TCA is

completely dissolved and that the TBA is not heat abused. The TCA/TBA solution was added to

0.5 g of homogenate and was heated for one hour at 650C. The samples were centrifuged (4C)

for 10 minutes at 2,500 g and read at 532 nm using a spectrophotometer (Agilent 8453 UV-

Visible, Agilent Technologies, Inc., Palo Alto, CA). A standard curve was constructed using 1,1,

3,3 tetraethoxypropane.

Analysis of Moisture and pH

Moisture content was assessed using a moisture balance (CSI Scientific Company, Inc.,

Fairfax, VA). The pH was measured using an electrode ( Thermo Sure-Flow electrode (Electron

Corp., Waltham, MA) and pH meter (Denver Instrument, model 220, Denver, CO). The

homogenate samples were directly measured for pH with continuous mixing.

Statistical Analysis

The SAS system was used to evaluate storage, pH adjustment, and antioxidant

implementation studies over time. All analyses were performed in at least duplicate and all

experiments were replicated twice. Analytical variation was established through first fitting two

way ANOVA models. Then multiple pair-wise comparisons were performed on estimated

measurement means, using Tukey adjustment controlling for Type I error.

Data was reported as mean + standard deviation. The analysis of variance and least significant

difference tests were conducted to identify differences among means, while a Pearson correlation









test was conducted to determine the correlations among means. Statistical significance was

declared atp < 0.05.

Results

Changes in Spanish Mackerel Lipid Oxidation

Water and lipid Content

As it is a rarity to receive fresh and quality consistent raw material for processing, it is

important to understand how pre-formed lipid oxidation products in the raw material affect lipid

oxidation of mackerel muscle homogenates at low and high pH. Upon receiving and/or

harvesting of fish, baseline levels of oxidation were measured to assess similarities and

differences in lipid oxidation of fish freshly caught or at postharvest storage for a few days. To

determine the quality of the Spanish mackerel raw material, lipid oxidation products for fresh

and aged fillets and/or mince were measured. The average moisture content, pH, and lipid

content for Batch 1 and 2 fish muscle are summarized in Table 3-1.

The Spanish mackerel obtained in May (Batch 1) was significantly (p<0.05) higher in lipid

content (7.5 2.12 %) than the September catches (Batch 2), 2.65 0.07%. Average moisture

contents for May (Batchl) and September catch (Batch 2) were 790.251% and 771.76%,

respectively. Average pH content of the Spanish mackerel mince for May (Batch 1) and

September (Batch 2) were 6.490.045 and 6.490.015, which are in agreement with previous

research (67). Depending upon various environmental factors (20), the observed differences

among the batches are to be expected. Addressing compositional difference in catch due to

differences in seasonal variability and postmortem age is important to study the changes

observed in the acidic-alkaline aided process. Industrially, these compositional factors could

have significant bearing on processing parameters and the quality of protein isolates obtained on

a mass processing scale.









Lipid Oxidation in Whole and Minced Spanish Mackerel

It was evident that aging of the fish muscle at 40C led to an increase in oxidation products

(Figure 3-1). Significant (p<0.05) increases in lipid oxidation for Batch 1 mince and Batch 2

mince were observed for up to 4 days and between 2 and 4 days of storage, respectively. Batch 2

mince after 4 days of storage reached equal levels of oxidation (-30 umol MDA/kg of tissue) as

Batch 1 mince at day 1. To detect changes in lipid oxidation incurred through mincing of raw

material, a comparative analysis of fillets and mince in Batch 2 Spanish mackerel was conducted

(Figure 3-1). Lipid oxidation in the fillets did not significantly increase over time. Batch 2 mince

accumulated overall significantly (p<0.003) higher TBARS levels than its corresponding fillets.

The levels of lipid oxidation for same day storage were not significantly different between Batch

2 fillets and Batch 2 mince; except for day four, (p<0.05). Lipid oxidation levels were

significantly higher in Batch 2 for days 2 and 3 than 0 and 1 days of storage for Batch 2 fillets.

Maximum TBARS values of -70 and -34 [tmol of malondialdehyde (MDA /kg) for Batch 1

mince and Batch 2 mince were reached after 1 and 4 days, respectively. When comparing lipid

oxidation over specific days, Batch 1 mince was significantly (p<0.05) greater than Batch 2

mince from 2 and 4 days of refrigerated storage.

Increasing the age of the raw material, significantly affected the lag time (P<0.004)

between Batch 1 and 2 mince. The onset of rancidity according to Richards et al. (14) is marked

by a TBARS increase of 20 umol/kg of tissue or more. Rancidity occurred within 1 day and after

2 days of storage in Batch 1 mince and Batch 2 mince, respectively. TBARS were overall

significantly (p<0.001) higher in Batch 1 than in Batch 2 mince.

Lipid Oxidation in pH Treated Spanish Mackerel Homogenates

An initial interest was to understand how preformed oxidation products in raw material

affected lipid oxidation with storage. The next objective was to evaluate the effects of low and









high pH on lipid oxidation in fish homogenates with respect to initial fish quality, cold storage,

and pH holding time.

Changes in homogenate lipid oxidation with storage

Figure 3-2 (b,d) and 3-3(b,d) and 3-4 (b,d) show that Batch 1 and 2 homogenates formed

significantly greater levels of TBARS with one and two days of mince storage versus

homogenates made from freshly prepared mince (p<0.001). As shown by Figures 3-2 (b) and 3-4

(d), TBARS levels in Batch 2 homogenates prepared from mince stored for two days were

equivalent to Batch 1 homogenates made from fresh mince.

Initial levels of PV and TBARS (Figures 3-2, 3-3, and 3-4) were equal in homogenates

prepared from mince stored for 0 or 1 days, and were significantly (p<0.001) less than

homogenates from mince stored for two days. Initial Batch 1 PV values for day 0, 1, and 2 were:

0.54 0.017, 0.60.018 2.520.42 mmol LPH/kg homogenate. Initial Batch 2 values at day 0, 1,

and 2 were 0.750.01, 0.6300.09, 0.88.007 mmol LPH/kg tissue, respectively. Significant

(p<0.001) increases in homogenate PV values for 0 and 1 days of mince storage were observed

after twelve hours of pH exposure. Endpoint levels of primary oxidation were equivalent in

homogenates made from 1 and 2 day stored mince and were significantly greater than

homogenates made from freshly prepared mince.

Batch 1 and 2 homogenates made from either 0 or 1 days storage mince started at

approximately the same TBARS levels (Figure 3-3). Batch 1 initial values on day 0, 1, and 2

were: 3.93 1.077, 6.0253.33, and 13.623.38 umol MDA/kg of homogenate. Batch 2 initial

values at day 0, 1, and 2 were 3.22 1.26, 6.090.57, and 7.23 0.94 umol MDA/kg

homogenate. Figures 3-3 (b,d) and 3-4 (b,d) show that with holding times of four or more hours,

day one TBARS values reached equal or greater levels of oxidation compared to initial day 2









value (p<0.05). Holding times greater than 8 hours had no affect on TBARS formed in

homogenates made from mince aged for 1 or 2 days.

Changes in lipid oxidation with pH and storage

As Figure 3-2(b) shows, pH 3 and 3.5 homogenates made from freshly prepared Batch 1

mince had significantly greater TBARS levels (P<0.001) than all other pH treatments. The levels

of TBARS in the pH 3.5 homogenates were significantly greater that those at pH 3 (p<0.016).

Maximum levels of lipid oxidation in pH 3.5 and 3.0 homogenates made from freshly prepared

Batch 1 mince were 64.00 5.58 and 53.524.98 umol MDA/kg of homogenate, respectively

(Figure 3-3b). TBARS levels were significantly greater in homogenates made from 1 day stored

mince versus freshly prepared mince (Figure 3-3b). Overall day 1 levels of lipid oxidation at pH

3 and 3.5 were not significant from day 2 levels (Figure 3-4b). After one day of mince storage,

levels of lipid oxidation at pH 2.5 were significantly (p<0.001) greater than control and alkaline

treatments (Figure 3-4b), with the pH 2.5 homogenate still accruing greater levels of lipid

oxidation after two days of refrigerated storage (Figure 3-4b). After two days of mince storage,

pH 2.5 TBARS were significantly (p= 0.007) less than observed levels at pH 3 (Figure 3-4b). No

significant increases in lipid oxidation were observed among the control or alkaline treatments

with mince storage (Figure 3-2b,3-3b,3-4b).

Changes in lipid oxidation with initial quality and pH

Homogenates 3.5 and 3 had significantly greater TBARS than pH 2.5, while TBARS were

significantly higher in homogenate 2.5 than pH 6.5, 10.5, 11.0, and 11.5 homogenates. For

Batch 1, levels of oxidation at pH 3.5 and pH 3 were not significantly different. Control (pH 6.5)

homogenates in Batch 1 developed significantly greater TBARS than alkaline homogenates. For

Batch 2, pH 3.5 homogenates had significantly greater TBARS than alkaline pH homogenates

(p<0.04). Batch 1 mince acidic homogenates oxidized significantly more than Batch 2 acidic,









alkaline, and control homogenates. Batch 2 alkaline (pH 10.5, 11, 11.5) treatments incurred

significantly (p<0.001) less TBARS than control (6.5) homogenate in Batch 1 and equivalent

levels to the alkaline treatments in Batch 2.

Changes in lipid oxidation with holding time and pH

After 4 hrs and up to 12 hrs, Batch 1 incurred significantly greater PV levels than Batch 2

mince (Figures 3-2a,c). PV levels (mmol LPH/kg homogenate) started out significantly higher

(p<0.001) in Batch 1 compared to Batch 2; however, after 12 hours the Batch 1 and 2

homogenates had equivalent PV levels.

As shown in Figure 3-2(a), PV values increased significantly after four hours for

homogenates at pH 3.5. Homogenate 2.5, 3, and 6.5 significantly increased in PV levels after 12

hours. Homogenate 3 and 3.5 reached the same levels of primary oxidation after 12 hours and

were followed by homogenates 6.5, 2.5, and alkaline pH in descending order of primary

oxidation.

From Figures 3-2 (b), homogenate holding times of four or more hours at acidic pH were

sufficient for rancidity to occur in Batch 1 homogenates. Homogenates held for 4 hours at acidic

pH had significantly (p<0.001) greater TBARS than control and alkaline homogenates held for

the same time. A similar observance was found in unadjusted pH turkey homogenates with 2.5

hours of refrigerated storage (143). The pH 3.5 homogenates were not significantly different than

pH 3 homogenates after 4 hours. After eight hours, pH 3 and 3.5 homogenates were

significantly greater than pH 2.5 (p<0.05). After a 12 hour holding time, homogenates at 3.5 had

significantly (p =0.027) greater TBARS than pH 3.0 homogenates, and both pH treatments

proved to be significantly greater (p<0.001) than all other pH treatments at pH 2.5 and higher pH

values (6.5, 10.5, 11, 11.5). The control (pH 6.5) treatment had equal levels of oxidation

compared to pH 2.5 and alkaline values after 12 hours.









Overall pH effects

The extent and rate of lipid oxidation (TBARS) was significantly (p<0.001) higher in

homogenates at acidic pH (2.5, 3, 3.5) than at alkaline (10.5, 11,11.5) or control pH (6.5).

Figures 3-2(b), 3-3(b), 3-4(b) shows that there was no significant difference in TBARS

formation among alkaline homogenates. TBARS formation proceeded faster and to a greater

extent in the following homogenate treatment order: acid> control>alkaline. Levels of TBARS

in descending order relative to individual pH treatment were as follows: 3.5, 3, 2.5, 6.5, and

alkaline pH (10.5, 11.0, 11.5). Levels of lipid oxidation among alkaline homogenates were

insignificant.

Lipid Oxidation of Homogenates with Antioxidant Implementation

The effect of antioxidant implementation in homogenates at low and high pH was worth

investigating, as it may decrease overall levels of lipid oxidation in the early stages of the

alkali/acid aided process and improve the quality of the final protein isolate. A commonly used

combination of alpha tocopherol and ascorbic acid in muscle foods was employed.

Batch 1 mince was stored for one day and was subjected to pH 3 as per findings in the

previous section, in which rancidity occurred after one day of storage and the greatest overall

levels of homogenate lipid oxidation developed (3 and 3.5). The levels of lipid oxidation of

homogenates from day old mince at pH 3 and 3.5 were not significantly different. pH 3 was

chosen, as this is close to the acid-aided process pH values used for the isolation of proteins. To

ensure that Batch 2 was of comparable oxidative quality of Batch 1 after one day of storage,

Batch 2 mince was stored for 4 days (See Figure 3-1).

The antioxidants ascorbic acid and/or tocopherol, citric acid/and or EDTA were added at

two concentrations, 0.1% and 0.2% w/w, in numerous combinations. The effects of these









antioxidants were also tested under physiological pH conditions (pH 6.5) in the homogenate

system.

Levels of lipid oxidation between batches were comparable, as all antioxidant treatments

at pH 3 for Batch 1 homogenates, with exception of pH 3 (0.1% ascorbic acid + 0.1%

tocopherol) homogenates (Figure 3-5b), were not significantly different from pH 3 homogenates

from Batch 2 (Figure 3-5c). However, pH 6.5 from Batch 1 homogenate (Figure 3-5b) formed

less TBARS than pH 6.5 from Batch 2 (p=0.0002) (Figure 3-5 c). Batch 2 pH 3 (0.1% ascorbic

acid + 0.1% tocopherol) formed greater TBARS compared to the corresponding Batch 1

treatment. The difference between pH 3 (0.1% ascorbic acid + 0.1% tocopherol) samples from

Batch 1 and 2 is negligible regarding quality as both treatments formed total levels of lipid

oxidation below the point of rancidity (>20 umol MDA/kg tissue). Because pH 6.5 was not

comparable among batches, pH 6.5 treatments from Batch Iwere only compared to other

homogenate treatments prepared from Batch 1 mince. Homogenate values were compared

among corresponding pH 3 treatments in both batches.

In Batch 1 treatments, various differences were observed. Overall levels of TBARS formed

in control pH 3 homogenates were significantly greater than all Batch 1 and 2 pH 3 treatments.

Homogenate treatment at pH 6.5 (0.1% ascorbic acid + 0.1% tocopherol) and pH 6.5 (0.1%

ascorbic acid) formed the highest TBARS levels (p<0.05) among all other treatments in Batch 1.

The antioxidant treatments were all equally effective in delaying/preventing oxidation in the pH

3 in contrast to pH 6.5.

As shown in Figure 3-6, the addition of EDTA (0.1%), citric acid (0.1%), and ETDA

(0.1%) + citric acid (0.1%), at pH 3 were rather ineffective in reducing lipid oxidation (TBARS,

PV). Over a twelve hour holding time, PV levels at pH 3 were not significantly affected by citric









acid (0.1%) and were significantly (p<0.001) reduced by the addition of EDTA (0.1%) and

EDTA (0.1%) + citric acid (0.1%) (Figure 3-6b). As shown in Figure 3-6b, the use of these

antioxidants when evaluated separately or in combination did not significantly reduce overall

levels of TBARS. These results suggest that the sole incorporation of chelators does not protect

against lipid oxidation in Spanish mackerel fish homogenates (51). However, in combination

with other antioxidants, chelators are promising inhibitors of oxidation according to Undeland

(67).

Lipid Oxidation of Homogenates after Readjustment to pH 5.5 and 7

In the acid and alkali-aided processes, homogenate proteins that are solubilized at low or

high pH are recovered by readjusting the pH to values where they aggregate. Consistent with

previous finding, Batch 1 mince was stored for one day and was subjected to pH 3 as per

findings in (Lipid oxidation in whole and minced Spanish Mackerel, Chapter 3) in which

rancidity occurred after one day of storage and the greatest overall levels of homogenate lipid

oxidation with respect to pH (3 and 3.5) homogenates occurred. Batch 2 mince was stored for 4

days as per previously discussed findings (Lipid oxidation in whole and minced Spanish

Mackerel, Chapter 3) to ensure approximate oxidation levels as Batch 1 samples. Homogenates

at pH 3 and 11 were held for one hour and then were readjusted to pH 5.5 and 7. The former pH

is commonly used to recover fish muscle proteins. However, improper folding of heme proteins

when readjusted from low pH to pH 5.5 may cause significant lipid oxidation due to incomplete

refolding (51). More refolding and less pro-oxidative activity is expected of hemoglobin with

readjustment to pH 7 (51).

The extent of lipid oxidation increased over time (p<0.0001) for all samples. Lipid

oxidation was measured at 0, 3, and 12 hours. These time points were chosen as per section,

"Lipid oxidation in whole and minced Spanish Mackerel,"(Chapter 3), in which trends of lipid









oxidation were realized. Batch 1 and 2 homogenates proceeded from the same initial levels of

oxidation; however, greater levels of lipid oxidation formed in Batch 1 compared to Batch 2

(p<0.0001).

Batch and homogenate pH readjustment

In Batch 1, greater levels of oxidation (PV,TBARS) occurred with the implementation of

antioxidants and readjustment from pH 3 to either pH 5.5 or 7 (p<0.0002) (Figures 3-7 a,b)

compared to pH 3. When homogenates were readjusted from pH 3 to 5.5 or 7 in the presence of

the antioxidant combination that was effective at pH 3 (0.1% ascorbic acid + 0.1% tocopherol)

(w/w) (Figures 3-5 b,c), a significant (p<0.0001) pro-oxidative effect was noted (Figures 3-7

a,b). This could be explained by the pro-oxidative effect of this combination observed (Figure 3-

5a) at alkaline pH 6.5, which is intermediate to pH 5.5 and 7. Interestingly, the same antioxidant

treatments with readjustment from pH 11 to pH 5.5 or 7 did not exhibit the same pro-oxidative

effect on the homogenates in terms of PV values as it did for the system readjusted from low pH

(Figure 3-8a); however, the significant increase in TBARS with pH readjustment to pH 5.5 or 7

and antioxidant implementation at pH 11 (Figures 3-8b) were consistent with the same

treatments at pH 3 (Figure 3-7b). According to TBARS results, antioxidant treatments with

readjustment from pH 11 were not significantly different from the low levels of oxidation at

control pH 11 in Batch 2 (3-8c). The pro-oxidative effect of readjustment from pH 3 to 7 with

antioxidants was further emphasized, having greater lipid oxidation TBARS and PV values

(p<0.0035) compared to pH 7 alone as exhibited in Batch 1 and 2 (Figure 3-7 a,b,c,d). TBARS

and PV levels of samples with pH readjustment from pH 3 and 11 were less (p<0.02) than

corresponding treatments with antioxidants. Oxidation levels with readjustment to pH and

antioxidants proved to be higher at pH 7 than at pH 5.5 in Batch 1 (PV) p<0.0001.









Time and homogenate pH readjustment

After 12 hours, trends among measures of lipid oxidation (TBARS, PV) among pH 3 and

11 readjustment treatments were observed. TBARS levels proved to be greater (p<0.0001) in the

pH 11 homogenate adjusted to 5.5 (0.1% ascorbic acid + 0.1% tocopherol) compared to

homogenates at pH 3, 5.5, 7, 3-5.5, and 3-7. Readjustment from pH 3 to 7 (0.1% ascorbic acid +

0.1% tocopherol) proved to be more pro-oxidative than the same treatment readjusted from pH

11 and all other treatments at pH 3 and 11. Samples at pH 5.5 had greater TBARS than samples

at pH 11 (p < 0.02). The pH readjustment treatments were always surpassed in levels of lipid

oxidation (TBARS) by corresponding treatments with antioxidants in Batch 1 (p<0.02) (Figures

3-7b and 3-8b). In Batches 1 and 2, readjustments to pH 7 with antioxidants had greater TBARS

than treatments adjusted to pH 5.5 with and without antioxidants except for pH (p<0.0001),

except for one sample pH 11-5.5 (Figure 3-8 d). PV values seemed to follow the same trends as

TBARS; however, statistically little differences were observed.

Discussion

Lipid Oxidation and Storage Stability

Aged fish muscle should incur greater and more rapid lipid oxidation, and this was

observed in the fillets and ground muscle (Figure 3-1). This is in agreement with the literature. It

has been reported that if the fish muscle has surpassed the induction phase of lipid oxidation,

which can occur with storage, the rate and extent of lipid peroxidation will be greater (144). In

addition, the levels of lipid oxidation assessed in the ground muscle were significantly higher

than those observed in the fillets. This is evident in the fact that mince from freshly caught fish

(Batch 2) accumulated significantly higher levels of lipid oxidation than the corresponding fillets

(Figure 3-1). Regardless of postmortem age of the fish, there is a significant different in the

extent of lipid oxidation with mechanical processing. Aside from compositional muscle aspects,









the differences in lipid oxidation in the mince and fillets are well explained and are consistent

with observed changes in muscle due to extrinsic factors. Various external factors can contribute

to lipid oxidation in muscle food including and not limited too: temperature, oxygen exposure,

light, mechanical processing, packaging atmosphere, introduction of pro-oxidants, and

cooking/heating (145). The extrinsic factor most relative to the storage study is mechanical

processing of meat. Various levels of oxidation will arise through mincing, flaking,

emulsification, and other physical handling. Tissue disruption increases oxygen-lipid interaction,

and thus, increases the occurrence of lipid oxidation. Mincing also releases pro-oxidants such as

hemoglobin and myoglobin and brings them in closer contact with lipids (80, 146). The Spanish

mackerel muscle was maintained at 40C during the grinding process; however, a paintyy" aroma

was observed and is indicative of increased lipid oxidation due to tissue disruption (30, 147) .

With storage and physical manipulation of tissue, reduced quality is evident in flavor, odor,

color, and nutritional changes (29, 30). This storage study served as comparative assessment of

raw material quality, and thus, serves as an applicable baseline for changes in lipid oxidation

implicated in the alkali/acid-aided protein isolation process.

Lipid Oxidation in Acid/Alkali Adjusted Homogenates

The initially higher PV levels in homogenates from Batch 1 mince is to be expected

(Figures 3-2, 3, 4). Batch 1 was stored for several days prior to arrival. Lipid hydroperoxides

are initial products of lipid oxidation. Lipid hydroperoxides are transient lipid oxidation

products, degrading soon after formation (48). Free radicals more readily react with antioxidants

than lipid components, due to bond dissociation energies. Once antioxidant defenses are

depleted, the rates of lipid hydroperoxides and secondary product formation will proceed faster

(148). According to Petillo et al. (143), antioxidant sources found in the blood plasma of stored

mince Spanish mackerel were exhausted with storage. Ascorbate and glutathione were exhausted









first as they serve to reduce the activation of metal containing pro-oxidants (148). Increases in

volatile secondary lipid oxidation products were perceptually detected in the homogenate

headspace; however, further studies are needed to substantiate these findings. The greatest

intensities were observed in the acidic treatments and negligible differences were detected in the

alkaline treatments. These easily detected volatiles are most likely aldehydes and can contribute

to rancidity and alterations in proteins comprising DNA, enzymes, and phospholipids structure

(87, 149, 150). In studies conducted by Alasalvar (151), the most predominant oxidation

products formed from cold storage of mackerel were hexanal, heptanal, octanal, dodecanal, 2-

pentenal, methyl-2-pentenal E-2-hexanal, and Z-4- heptenal. Thus, these volatiles could

potentially contribute to oxidation of the homogenates.

Comparable TBARS levels in homogenates made from one and two day old Batch 1 mince

is most likely due to the depletion of hydroperoxides, thereby, slowing the formation of

secondary products (148). Greater levels of lipid oxidation were observed among acidified

homogenates (Figures 3-2,3,4). These high levels may be explained in part by the high moisture

content (97-98%) in the homogenates. Water activity is an influential factor in the rate of lipid

oxidation. It is known that with high water activity, the rate and extent of lipid oxidation is

greater (148). Thus, the homogenates would accumulate higher levels of oxidation, not only due

to acidification but due to enhanced incorporation of pro-oxidants and oxygen via the water

phase.

During the homogenization process and pH adjustment, various changes to the

predominant catalysts, heme proteins can occur and affect rates of lipid oxidation. According to

Shikama (60) and Kristinsson (152), heme proteins become oxidatively stable due to increased

heme group preservation at alkaline pH (60, 152). With lowering of pH, subunit dissociation,









unfolding, release of heme groups and/or iron, and loss of oxygen can occur with heme proteins

(43, 50, 52, 153-155). Such modifications are more likely when hemoglobin/ myoglobin are

subjected to low pH; thereby, converting hemoglobin to more pro-oxidative forms including

deoxy-, met-, perferryl-, and ferryl-hemoglobin/myoglobin (37). Hemoglobin/ myoglobin Bohr

effect is amplified with decreases in pH below neutrality and leads to deoxygenation (50).

According to Richards et al. (42), deoxyhemoglobin/myglobin are highly pro-oxidative.

Consequentially, activated heme proteins can give rise to the production of reactive oxygen

species and decomposition of lipids and/or lipid oxidation products as observed in the acidified

homogenates.

The effect of pH on hemoglobin is particularly relevant to the processing of migratory

pelagic species. Hemoglobin in dark muscle pelagic species has a lower affinity for oxygen. A

more flexible tertiary structure facilitates quick release of oxygen for metabolic respiration (156).

Richards and Hultin (157) reported significantly greater catalytic activity in mackerel versus

trout hemoglobin (157). Significant pH effects could be realized during the isolation of dark

muscle fish species, due to less stable hemoglobin. Further investigation is required to

understand the effects of pH shifts on heme proteins in a variety of fish species homogenates and

how these implications affect lipid oxidation during protein processing.

The effects of pH in muscle foods, or single or multi-component model systems could

explain the effect of pH in the lipid oxidation of homogenates. Higher levels of lipid oxidation in

the homogenates at low pH could be explained by heme protein changes in aqueous model

systems. A study conducted by Kristinsson (12) revealed that considerable changes in flounder

hemoglobin tertiary structure occurred with lowering of pH. The most significant unfolding

transitions occurred between pH 4 and 3.5 and the greatest unfolding occurred at pH 2.5 (12).









Hemoglobin unfolding occurred very quickly (seconds) at pH 1.5-2.5, while unfolding at higher

acidic pH (3-3.5) took over 30 minutes. Alkaline pH (10-11) brought about very little or

negligible changes in heme structure (12). The low levels of lipid oxidation in the alkaline

homogenates can be explained by heme group and oxidative stability (51). The proxidative

activity of hemoglobin/myoglobin at low pH and inactivity at high pH is greatly dependant on

distal and proximal histidine interactions with the heme group and oxygen (158). Under acidic

conditions, interactions with the heme and distal histidine and proximal histidine can be lost

through autoxidation and denaturation (12, 152). Under alkaline conditions, substantial heme

unfolding at low pH would explain the high levels of oxidation observed in the acidic

homogenates (Figures 3-2 ,3 ,4). In a study by Kristinsson and Hultin (51), conformational

changes were reported for trout hemoglobin (0.8 uM in water, 50C) at pH 3, 7, and 10.5 over

time. According to UV-Vis spectral (300-700), trout hemoglobin conformational changes

occurred in two minutes at pH 3; while changes at pH 7 and 10.5 were observed in 29 and 120

hours, respectively (159). The levels of oxidation in the Spanish mackerel homogenates at

neutral pH were less than acidic homogenates; this is most likely due to less structural flexibility

and hydrophobicity of hemoglobin at neutral pH (51). Additionally, homogenates at pH 6.5 were

more susceptible to autoxidation relative to alkaline pH and significantly less susceptible to auto-

oxidation at low pH. Richards and Hultin (43) found that with lowering of pH from 7.6 to 6.0,

hemoglobin was quickly converted to pro-oxidative forms: deoxyhemoglobin and metmyoglobin

(43). In the same study, Richards and Hultin (43) found that a hemoglobin mediated washed cod

system had shorter lag phases at lower pH (43). Thus, the differing levels of lipid oxidation in

homogenates at neutral pH could be caused by less oxidative stability and considerably more

stability of hemoglobin at alkaline and acidic pH, respectively.









Various conformation changes within hemoglobin/myglobin could explain the differences

in oxidation among the pH treated homogenates. Because the extent of unfolding is greater at

low pH versus neutral and alkaline pH, hemoglobin becomes more hydrophobic and thus is

better able to interact with lipid membranes (152). Additionally, Kristinsson (12) found that

flounder hemoglobin existed as a molten globule (partially unfolded) state at low pH. A molten

globule is both hydrophobic and hydrophilic in nature, providing greater accessibility to lipids.

Thus at low pH, hemoglobin would be very pro-oxidative in the presence of membrane lipids in

the homogenates. Access of the protein to lipids is not only enhanced at low pH, but the heme

group becomes significantly exposed at low pH. According to Kristinsson (12), increased

hydrophobicity in flounder hemoglobin was attributed to greater tryptophan residue exposure

with lowering of pH from neutrality (pH 7). Increased rates of met-myoglobin/hemoglobin

formation are known to occur at low pH (60). Autoxidation, unfolding, and dissociation of

myglobin/hemoglobin can give rise to various reactive species such as ferryl hemoglobin. Ferryl

hemoglobin is believed to be a potent catalyst of oxidation in muscle foods, particularly at low

pH (56). These oxidized forms of hemoglobin/myglobin can readily catalyze lipid oxidation at

low pH (160) and could explain greater lipid oxidation in the acidic homogenates.

Results from washed cod muscle systems could help explain the overall levels and rate of

lipid oxidation (TBARS) in Spanish mackerel homogenates. The pH 6.5 homogenates were

lower than observed acidic homogenates and higher than alkaline homogenates in Batch 1

(Figures 3-2b,3b,4b).

Homogenate pH exposure times significantly affected the lipid oxidation development in

homogenates. Acidic homogenates made from freshly prepared Batch 1 mince developed

rancidity after four hours while neutral pH developed rancidity levels after 12 hours and alkaline









pH homogenates did not reach rancidity levels. Pazos et al. (161) observed similar trends in a

hemoglobin mediated lipid oxidation system using washed cod. At pH 3, the onset of lipid

oxidation (-5 hours) was less than pH 6.8 (- 10 hours) and the endpoint oxidation levels were

comparable. Likewise, the pH 3.5 and 3.0 homogenates from freshly prepared Batch 1 mince

developed rancidity levels in approximately 6 hours, while pH 6.5 developed rancidity in

approximately 12 hours. Pazos et al. (161) observed similar endpoint levels of oxidation

(TBARS) in the pH 3 and pH 6.8 adjusted washed systems. The pH 3 and 6.8 homogenates, from

freshly prepared Batch 1 mince, did not reach equivalent levels of oxidation at 12 hours;

however, greater holding times or mince storage may have resulted in equal endpoint levels of

lipid oxidation (Figures 3-2(b), 3-4(b)). As observed in the homogenates, Kristinsson and Hultin

(51) reported similar lipid oxidation trends in hemoglobin mediated washed system samples.

Trout hemoglobin was adjusted to acidic (pH 2.5, 3, 3.5), neutral (pH 7), and alkaline (10.5, 11)

pH values. The greatest extent and rate of lipid oxidation occurred at low pH, followed by

neutral pH, and then alkaline pH. Increased rates and extents of lipid oxidation followed by

longer lag phases and decreased levels of lipid oxidation at neutral (6.5) was observed in the

Spanish mackerel homogenates. The development of TBARS was greatest in a washed cod

system at pH 2.5 and 3 than at pH 3.5 (51). The homogenate samples at pH 3.5 over time had

greater rates and levels of lipid oxidation than pH 2.5 and 3. As Kristinsson and Hultin (51)

observed at pH 10.5 and 11 in the washed cod system, minimal levels of lipid oxidation

developed in pH 10.5, 11, and 11.5 homogenates.

Although quantitative color differences among the homogenates were not reported, it was

evident that acidified homogenates were yellowish white while the alkaline and neutral

homogenates were pink in color. From these color differences, it is evident that significant









unfolding and denaturation occurred in the acidified homogenates, while the less unfolded and

native heme proteins in the other treatments retained their color. Richards et al. (162) and Lee et

al. (163) found a significant loss in redness, "a" value was due to conformational changes as

heme proteins were converted from their oxy-to met-forms. Hultin (5) reported that meat color

changes are byproducts of free radical induced oxidation, secondary to lipid hydroperoxide

formation.

Homogenate pH Readjustment

To investigate changes in lipid oxidation in acid/alkali-aid process upon pH readjustment,

Spanish mackerel homogenates were held for one hour at pH 3 and 11 and were then readjusted

to pH 5.5 and 7. The homogenates were held at pH 5.5 and 7 at 40C for 11 hours with constant

mixing. The readjustment from acidic and alkaline pH to pH 5.5 and 7 mimics the pH

readjustment step in the acid/alkali aid process in which proteins are precipitated and then

centrifuged to obtain a protein isolate.

To date, no studies have investigated changes in lipid oxidation over time in acid/alkali

homogenates upon readjustment, nor has the effectiveness of antioxidants in preventing lipid

oxidation during homogenate readjustment been investigated. Lipid oxidation levels (TBARS) in

pH 3 and 11 homogenates were not significantly affected by pH readjustment to pH 5.5 or 7

(Figures 3-7b, 3-8b). Trout hemoglobin refolding upon readjustment from acidic or alkaline pH

to neutral pH 7 was studied by Kristinsson and Hultin (51). Trout hemoglobin was adjusted to

pH 3 and 10.5 and was then readjusted to pH 7 and added to a washed cod system. The effect of

pH lowering and pH holding times on the pro-oxidative activity of refolded hemoglobin were

investigated. Less refolding occurred for more unfolded acidified hemoglobin (pH 2.5) compared

to hemoglobin at pH 3.5. Longer acidic pH holding times and lower pH values gave

hemoglobins that were incompletely unfolded and most likely more pro-oxidative relative to









native hemoglobin (pH 7). Alkaline readjusted hemoglobin did not incur significant changes in

pro-oxidative activity under refolding conditions (51). Similarly, pH readjustment of

homogenates from pH 11 did not significantly change and were minimal (Figures 3-8 b). In

contrast to Kristinsson and Hultin (51) findings in which increased levels of lipid oxidation were

observed upon readjustments from pH 3, levels of lipid oxidation in homogenates at pH 3 with

readjustment to pH 5.5 and 7 remained unchanged. Insignificant changes in homogenate lipid

oxidation levels could be attributed to changes in hemoglobin conformation and thus pro-

oxidative activity. Kristinsson and Hultin (51) observed significant increases in pro-oxidative

activity of hemoglobin at pH 3, when held at pH 3 for 20 minutes versus 90 seconds.

Conformational studies gave proof to greater refolding with shorter holding times at pH 3 (51).

The hemoglobin in homogenates at pH 3 were likely held for a duration long enough (1 hour) to

cause irreversible denaturation and thus inability for structure recovery. Similar to lowering of

pH, incomplete refolding can increase the exposure of hydrophobic sites and heme within the

tertiary structure of hemoglobin. The site exposure makes hemoglobin more pro-oxidative (164).

Due to longer holding times, acidified hemoglobin was unable to refold. Pro-oxidative activity

of the hemoglobin would then be very similar among the pH 3 homogenates with and without pH

readjustment.

Interestingly, the incorporation of tocopherol and ascorbic acid into pH 3 and pH 11

readjustment treatments proved to be pro-oxidative, compared to controls and readjustment

treatments (Figures 3-7b, 3-8b). The pH 3 and 11 homogenates with antioxidants formed greater

levels of lipid oxidation when readjusted to pH 7 compared to 5. These increases in oxidation

with readjustment to neutral pH is most likely impart due to the instability of ascorbic acid at

neutral pH as per discussion in the following section.









Antioxidant Implementation

As lipid oxidation could be problematic during acid processing of various dark muscled

fish species (13), antioxidants were added to the homogenates. Ascorbic acid and tocopherol,

either alone or in combination at 0.1% and 0.2%, were equally effective in preventing oxidation

in pH 3 homogenates (Figures 3-5 b,c). In a study conducted by Undeland et al. (67),

incorporation of antioxidants in the pre-wash and/or homogenization step significantly decreased

levels of lipid oxidation in the acid-aided isolates obtained from herring. Incorporation of

erythorbate (0.2% w/w) in combination with EDTA (0.2% w/w) or sodium tripolyphosphate

(STPP) (0.2% w/w) into herring homogenates at pH 2.5, have proven beneficial in maintaining

the quality of dark muscled fish protein isolate. Erythrobate, and iso-form of ascorbic acid were

found to be effective heme-iron reducing agents in herring homogenates. According to other

findings (68, 153), antioxidant implementation early in the modification of mackerel fish muscle

is most efficacious in reducing overall levels of oxidation in the final product. However,

antioxidant interventions prove to be most beneficial in light versus dark mince, most likely due

to significant levels of hemoglobin (68, 153). Antioxidants could not only be implemented in the

homogenates, but in the pre-wash of the Spanish mackerel. Additionally, various procedural

modifications in combination with antioxidants could reduce lipid oxidation in homogenates

made from dark muscle species: washing the mince, removal of dark muscle, and/or reducing the

holding time at extreme pH (67). The effects of how tocopherol and ascorbic acid incorporation

into Spanish mackerel homogenates affect lipid oxidation levels in the protein isolate requires

further investigation.

The incorporation of ascorbic acid (0.1%) alone or in combination with tocopherol (0.1%)

at pH 6.5 (Figure 3-5b) proved to be pro-oxidative in comparison to corresponding treatments at

pH 3 (Figures 3-5b). A decrease in lipid oxidation due to use of lower antioxidant concentration









increases was observed in the pH 3 homogenates (Figure 3-7b). Studies have shown a

concentration and pH effect on the ability of ascorbic acid and tocopherol to scavenge free

radicals. It was expected that lower levels of ascorbic acid (0.1%) rather than higher levels

(0.2%) in the homogenates would be pro-oxidative. Previous studies demonstrated a

concentration effect of ascorbic acid and tocopherol at physiological pH (64, 65). It has been

reported that at high and low concentrations, tocopherol and ascorbic acid impart pro-oxidant

and antioxidant effects, respectively. Cillard (165) found that high concentrations ( > 7,600

mg/kg, 0.7%) and low concentration (<3,800 mg/mL, 0.38%)oftocopherol induced pro-

oxidative effects in linoleic acid. Ascorbic acid could be prooxidative (>5,000 mg/kg, 0.5%) in a

hemoglobin containing solution (166). Watts (167) found that ascorbic acid was pro-oxidant at

176 mg/mL in a carotene-linoleate aqueous model, and it was concluded that chelators were

needed to sequester catalytic iron in the presence of these ascorbic acid concentrations.

Ultimately, there is a net prooxidative or antioxidative role played by these antioxidants.

Ascorbic acid is known to promote lipid oxidation by facilitated iron release from heme proteins

(168), on the other hand ascorbate can quench myoglobin ferryl radicals post mortem (169). The

roles that ascorbic acid and tocopherol play in lipid oxidation is determined by net antioxidant

activity. At pH 3, ascorbic acid and tocopherol either alone or in combination at 0.1% and 0.2%

were equally effective in preventing oxidation (Figures 3-5b,c), very similar to the study

conducted by Undeland et al. (67) in which erythorbate (0.2%) was added to herring

homogenates. Hultin and Kelleher (68) found that washing Atlantic mackerel mince with 0.2%

ascorbic solution and 1.5 mM EDTA, effectively prevented lipid oxidation. At high

concentrations (0.2%), ascorbic acid can quench ferryl radicals. In a study conducted by

Richards and Li (170), incorporation of high (2.2mM) and low levels (20-200 uM) of ascorbate









into hemoglobin-mediated lipid oxidation system resulted in inhibition and no pro-oxidative

effect on TBARS, respectively. In the same study conducted by Hultin and Kelleher (68),

ascorbic acid may have prevented lipid oxidation through reduction of heme iron from the ferric

to the ferrous form. The concentration effect of antioxidants could be very different among

studies; due to differences in pH and aqueous system compositions. The change in effectiveness

of these antioxidants with pH and concentration is to be expected.

Ascorbic acid's structural and oxidative instability at neutral pH may explain the high

degree of oxidation found in 6.5 pH treatments using ascorbic acid at 0.1% concentrations as

well as significant increase in oxidation after pH readjustment to pH 5.5 and 7 (0.1% tocopherol

+ 0.1% ascorbic acid). Because antioxidants have various pH dependant ionic forms, a non-linear

relationship between pH and antioxidant stability exists (171). Changes in pH are likely to cause

changes in antioxidant structural stability and thus function. It has been found that tocopherol

and ascorbic acid have different stabilities at different pH values (172). In general, ascorbic acid

is stable at acidic pH and unstable at neutral and alkaline pH. Tocopherol is stable at neutral,

acidic, and alkaline pH (172). Tocopherol's stability and thus ability to prevent oxidation exist

over a wide pH and thus it is most likely that ascorbic acid counteracts the positive effects of

tocopherol, when adjusted to neutral pH. It is likely that ascorbic acid is more antioxidative at

low pH compared to pH 6.5. Ascorbic acid is deprotonated with neutralization as it has a pKa

4.04 (173). Ascorbic acid could effectively reduce heme iron from ferric to the active ferrous

form. The ferrous form is known to breakdown lipid hydroperoxides and hydrogen peroxide to

form lipid radicals and ROS, respectively (169, 170, 174). Thus, it is likely that in the pH 6.5

homogenates, ascorbic acid is prooxidative through indirect propagation of lipid oxidation. At

low pH, ascorbic acid served as an effective antioxidant. Similarily, Undeland et al. (67) found









that ascorbic acid at 0.2% concentrations effectively reduced lipid oxidation in the protein

isolates made from acidified (pH 2.5) herring homogenates. Due to the fact that ascorbic acid

was not effective in reducing lipid oxidation in the homogenates at pH 6.5, a more stable analog,

erythorbate, may be a better choice with pH fluctuations. It is likely that both concentration

effect and pH influenced the effectiveness of ascorbic acid as an antioxidant.

Due to the key roles antioxidant roles exemplified in various studies, ascorbic acid and

tocopherol were used in the present studies. Post mortem antioxidants are preferentially

exhausted due to one-electron reduction potentials, also known as the "pecking order" (175) and

locations within muscle cells (5). The major antioxidant defenses against lipid oxidation in

muscle foods are the water-soluble inhibitors, ascorbate and glutathione, and the lipid-soluble

inhibitors, tocopherol and coenzyme Q. The tocopherol isomer of greatest abundance in skeletal

muscle is a- tocopherol and accumulates through dietary intake (176). In a paired fillet kinetics

study conducted by Petillo et al.(143), the loss of antioxidants in Atlantic mackerel was as

follows: glutathione and ascorbate, alpha tocopherol, and ubiquinone-10 (oxidized form of

coenzyme Q). The rate of antioxidant loss was significantly higher in the dark muscle than in the

light muscle (177). Similarly, the same sequence of antioxidant depletions was indicative of one-

electron transfers, according to Jia et al. (28) and Pazos et al. (178, 179).

Over time, it is evident that significant increases in TBARS levels occurred after four

hours in the pH 3 homogenate samples (Figure 3-5b). This is most likely due to losses in

tocopherol, which is highly correlated with increases in lipid oxidation and hydroperoxides in

muscle foods (177, 179, 180). The loss of alpha tocopherol is deemed the last defense against

lipid oxidation, in vitro. In turkey breast and thigh homogenates, the loss of a-tocopherol as it

relates to lipid oxidation was observed. Within 2.5 hours at 40C, there was significant increase in









lipid oxidation ( TBARS and PV) and a substantial loss of a-tocopherol (50%) (181). It is

evident that the loss of tocophereol was prevented by the implementation of ascorbic acid and

tocopherol. Other studies have reported preservation of tocopherol in muscle foods. Propyl

gallate and grape polyphenols were found to significantly preserve tocopherol in horse mackerel

during frozen storage (179). It is commonly known that ascorbate regenerates tocopherol and

serves an important function in preserving the free radical scavenging capacity of tocopherol in

membrane systems (182). Ascorbate is known for its ability to reduce hypervalent hemoglobin, a

highly pro-oxidative hemoglobin species. Ascorbate also possesses the ability to break down

hydroperoxides, which are activators of met-Mb/Hb to the ferryl form (183). It is widely

accepted that in membrane systems, a lipophilic antioxidant (tocopherol) and a hydrophilic

antioxidant ascorbatee) should be combined due to synergistic affects. In the homogenates, a

significant synergistic affect was not evident as all antioxidants treatments, alone or in

combination, had equivalent levels of lipid oxidation inhibition at pH 3 (Figure 3-5 b).

EDTA and Citric Acid Implementation at Low pH

EDTA and citric acid at 0.1% concentrations, together or in combination were ineffective

in inhibiting lipid oxidation (PV, TBARS) in pH 3 homogenates (Figures 3-6a,b). In contrast to

these finding, Undeland et al. (67) found that the implementation of EDTA (1.5 mM; 0.044%)

effectively reduced lipid oxidation in protein isolates obtained from acidified (pH 2.5) herring

homogenates. Similar to pH 3 homogenates results EDTA (2.2 mM) did not significantly affect

lipid oxidation levels in a hemoglobin mediated lipid oxidation system according to Richard and

Li (170). Due to the ineffectiveness of the chelators, it is evident that free iron or iron released

from heme may not have served as the predominant catalyst in the lipid oxidation of the

homogenates. To fully evaluate the effective of these chelators, it would be necessary to

evaluate the levels of TBARS formed in the final isolate. The appropriate implementation of









antioxidants in different types of muscle systems should be considered, and the effectiveness of

appropriate heme-related and/or low molecular appropriate antioxidants should be assessed. In a

model study conducted by Kanner et al (41), oxidation was generated through two catalytic

mechanisms: hydrogen peroxide activated met-myoglobin and iron chloride/ascorbate; however,

only the oxidation induced by the heme bound iron was inhibited by muscle cytosolic extracts in

washed fish systems (41). Obviously the chelators and or antioxidant mechanism differ for free

transition metals versus those that are heme bound proteins. Additionally, chelators in

combination with other antioxidants could be effective in reducing overall lipid oxidation in the

final product isolate as observed by Undeland et al. (67).

Conclusion

This storage study enables the understanding of how raw material of different quality may

affect lipid oxidation formation during the alkali/acid-aided protein isolation process, particularly

in the initial homogenization and pH adjustment step of pelagic specie protein extraction. Later

research (Chapter 4) will investigate whether the lipid oxidation incurred during the

homogenization steps affects the quality of the actual protein isolate. Lipid oxidation formation

proved to be greatest in Spanish mackerel homogenates at pH 3 and 3.5. The implications of

holding the proteins at an acidic pH followed by pH readjustment and isolation on protein and

lipid oxidation will be addressed in the following chapter.

The implementation of tocopherol and ascorbic acid at low pH significantly decreased

levels of lipid oxidation in homogenates; however, antioxidants proved to be pro-oxidative with

pH readjustment of homogenates from acidic and alkaline pH to neutral pH. Increases in lipid

oxidation in the readjusted homogenates with antioxidants were observed. Readjustment alone

had no affect on the levels of lipid oxidation in homogenates. Longer holding times significantly

increased the levels of lipid oxidation in the readjusted homogenates with antioxidants.









Successful implementation of antioxidants into the acid and alkali-aided protein isolation process

requires further investigation of appropriate antioxidants and their effectiveness at various pH

values and holding times.









Table 3-1. Initial average mince values for moisture content, pH, and lipid content. Assessments
were conducted on two separate batches of fish: Batch 1 (3 day old raw material) and
Batch 2 (freshly caught raw material). Lipid content means and standard deviations
were calculated from triplicate analysis. Different letters within each row indicate a
significant difference (p<0.05).


Moisture Content (%) 790.251a 771.76a

pH 6.490.045a 6.490.015a
Lipid Content (%) 7.52.12a 2.65 0.07b


0 1 2 3 4
Time (days)


Figure 3-1. Lipid oxidation in Spanish Mackerel mince as assessed by thiobarbituric acid
reactive substances (TBARS). Experiments were conducted on two separate batches
offish: Batch 1 (3 day old raw material received in May) and Batch 2 (freshly caught
raw material). Each sampling point is representative of a TBARS mean with standard
deviation bars from triplicate analysis.


Batch 2 (September)- Fresh


Batch 1 May- 3days old















-4-pH2.5 I) 90 -4-pH2.5 D)
10
1- -- pH 3.0 80 -pH3
-A- ppH 3.5 70 -A- H 3.5
8 2 70
P 6.560 X-pH6.5
S6 -0-pH10.5 -D-pH 10.5
5 50
S *I pH 11 0 OI pH 11
2 40



1 -I 0 30 -
















o 6
.0 -. 70 ..pH3... 105
0 2 4 6 8 10 12 0
0 2 4 6 8 10 12
Time (hours) Time (hours)









12 -100
-4- pH 2.5 90 --pH 2.5
10 -(CpH)3 H 8 p3 (D)
-U-pH3 .080C)U-pE

--pH 3.5 A--pH 3.5
--X-pH 6.5 60 pH6.5
60
S-0- pH 10.5
<50 0 pH11
...&--p H 11.0
S40 ---pH 11.5
4 pH 11.5
30
2 -20

0 -
0
0 2 4 6 8 10 12
0 2 4 6 8 10 12
Time (hours) Time (hours)




Figure 3-2. Development of lipid hydroperoxides and thiobarbituric reactive substances
(TBARS) in pH treated homogenates. Panels A and B represent lipid hydroperoxides
and TBARS in homogenates made from Batch 1 mince that was freshly prepared.
Panels C and D represent lipid hydroperoxides and TBARS in homogenates made
from Batch 2 mince that was freshly prepared. Each sampling point is representative
of TBARS and PV mean with standard deviation bars from triplicate analysis.































0 2 4 6
Time (hours)


10 12


0 2 4 6
Time (hours)


8 10 12


0 2 4 6
Time (hours)


8 10 12


4 6
Time (hours)


Figure 3-3. Development of lipid hydroperoxides and thiobarbituric reactive substances
(TBARS) in pH treated homogenates. Panels A and B represent lipid hydroperoxides
and TBARS in homogenates made from Batch 1 mince, after 1 day of storage. Panels
C and D represent lipid hydroperoxides and TBARS in homogenates made from
Batch 2 mince, after 1 day of storage. Each sampling point is representative of
TBARS and PV mean with standard deviation bars from triplicate analysis.


---pH 2.5
--pH 3.0 (D)
-A-pH 3.5
-X-pH 6.5
-0-pH 10.5
*--pH11

-+-pH 11.5


10 12




































0 2 4 6
Time (hours)


8 10 12


0 2 4 6
Time (hours)


8 10 12


0 2 4 6
Time (hours)


8 10 12


Figure 3-4. Development of lipid hydroperoxides (PV) and thiobarbituric reactive substances
(TBARS) in pH treated homogenates. Panels A and B represent lipid hydroperoxides
and TBARS in homogenates made from Batch 1 mince, after 2 days of storage.
Panels C and D represent lipid hydroperoxides and TBARS in homogenates made
from Batch 2 mince, after 2 days of storage. Each sampling point is representative of
TBARS and PV mean with standard deviation bars from triplicate analysis.


4 6
Time (hours)


10 12


-pH2.5 (D)
-- pH 3.0
-A-pH 3.5
-X-pH 6.5
-D-pH 10.5
S--pH 11

-pH 11.5





























0 2 4 6
Time (hours)


8 10 12


100
90 --pH 3.0 (Acid control)
80 -pH 6.5 (Control)
70
-A-pH 3 Vit C. (0.2%)
60
50 --pH 3 VitE.(0.2%)

40 --DpH 3 Vit C. (0.1%) + VitE. (0.1%
30
20
10
0
0 2 4 6
Time (hours)


(B)






)8 10 12






8 10 12


--pH 3.0
-0-pH 6.5
-- pH 3 0.1% C
- +-- pH 3 0.1% E
-- pH 3 Vit. C (0.1%)+ VitE (.1%)
-A-pH Vit. C (0.2%)
--pH 3 Vit. E (0.2%)
--pH 3 (0.2%E+0.2%C)


(C)


E 10P ... ..........

0 2 4 6 8 10 12
Time (hours)






Figure 3-5. Effect of ascorbic acid and tocopherol at 0.1% and 0.2% w/w on lipid oxidation in
homogenates. Panel A represents pH 3 homogenate from Batch 1, after 1 day of
storage. Panel B represents pH 6.5 homogenate from Batch 1, after 1 day of storage.
Panel C represents pH 3 homogenates equivalent to Batch 1 quality by using Batch 2,
after 4 days of storage. Each sampling point represents the mean and standard
deviation, from triplicate analysis, repeated twice.






























0 2 4 6 8 10 12
Time (hours)


0 2 4 6
Time (hours)


8 10 12


Figure 3-6. Effect of EDTA and citric acid at 0.1% w/w on lipid oxidation in pH 3 homogenates.
Panel A represents lipid hydroperoxides in Batch 1, after 1 day of cold storage.Panel
B represents TBARS in homogenates. Each sampling represents the mean and
standard deviation, from triplicate analysis from replicate experiments.
















200- +pH 3
p H3 (A) :180 H5 (B)
r -'-pH3to7 -A-pH3to 7
S15 -D-pH 3 to 5 5 (0 1% Vit C+ 0 1% Vit E) 160 --pH 3 to 5 5 (0 1% Vit C + 0 1%Vt E)
-A-pH 3 to 7 (0 1% Vit C +0 1% Vit E) --pH 3 to 7 0 (0 1% Vit C +0 1% Vit E)
S140
120
10 2 4 100
80
60
5
S40
20
0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Time (hours) Time (hours)






20 50
--pH 3.0 (C) --p3 (D)
-pH 3.0 to 5.5 pH to5.5
S -A-pH3to7 40 --pH3to
1i5 E -*-pH3to7
S-D pH 3 to 5.5 (0.1% Vit.CO..1% Vit. E) 0
-O-pH 3 to 5.5 (0.1% Vit.C+0.1% Vit. E) ---pH 3 to 5.5 (0.1% Vit. E+0.1% Vit.C)
p -A-pH 3 to 7 (0.1 Vit.C+0.1%0. Vit. E)
-- H 5.5 30 ---pH 3 to 7(0.1% Vit.C+ 0.1%Vit. E)
10 -. pH7 ---pH 5.5


























control homogenate at pH 3 is also shown. Homogenates were held at pH 3 for one
0.. ..pH 7
-220

E 5
10



0 2 4 6 8 10 12 0 2 4 6 8 10 12
Time (hours) Time (hours)








Figure 3-7. Effect of pH readjustment to neutral pH (5.5 and 7.0) antioxidants (0.1% ascorbic
acid +0 .1% tocopherol), on development of lipid hydroperoxides (PV) and
thiobarbituric reactive substances (TBARS) in pH treated homogenates. Panel A and
B represent lipid hydroperoxides and TBARS in homogenates made from Batch 1
mince, after 1 day of storage. Panels C and D represent lipid hydroperoxides and
TBARS in homogenates made from Batch 2 mince, after 1 day of storage (4'C). A
control homogenate at pH 3 is also shown. Homogenates were held at pH 3 for one
hour prior to readjustment. Each sampling point is representative of the mean with
standard deviations from triplicate analysis, repeated twice.














20 200
-*-pH11 (A) 180 -"pH ll(B
(A) 160 -U-pH11to55
15 --pH 11 to 55 Fi -A-pH 11to 7
140
-A-pH 11 to7 1 -D-pH 11 to 5 5 (0 1% Vt C+01% Vit E)
o 120 --pH 11 to 7(0 l%Vit C+0 1% Vit E)
1 -0-pH 1 to 5 5(0 1% Vt C+0 l%Vit E) 100
to 10[ 100
S --pH 11 to 7(0 1%Vlt C+0 1%Vit E) 80
5 60
E 40
<20
0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Time (hours) Time (hours)





20 50

--pH II 4pH 1 (D)
-)p~llfoI(C) -pHll (D
-U-pH 11 to 5.5 (C ) 40 -1-pH 11 to 5.5
15
--A--pH 11 to 7 -A-pH 11to 7
S -a-pH 11 to 5.5 (0.1% Vit. E+0.1% Vit. E) 0 30 -0-pH 11 to 5.5 (0.1% Vit. C+0.1% Vit. E)
pH 11 to 7 (0.1% Vit.C+0.1%Vit. E) h pH 11 to 7 (0.1%Vit.C+0.1% Vit.E)



10
m -pH 5.5 -tpH 5.5





0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Time (hours) Time (hours)








Figure 3-8. Effect of pH readjustment to neutral pH (5.5 and 7.0) antioxidants (0.1% ascorbic
acid +0 .1% tocopherol), on development of lipid hydroperoxides (PV) and
thiobarbituric reactive substances (TBARS) in pH treated homogenates. Panel A and
B represent lipid hydroperoxides and TBARS in homogenates made from Batch 1
mince, after 1 day of storage. Panels C and D represent lipid hydroperoxides and
TBARS in homogenates made from Batch 2 mince, after 1 day of storage (40C). A
control homogenate at pH 11 is also shown. Homogenates were held at pH 11 for one
hour prior to readjustment. Each sampling point is representative of the mean with
standard deviations from triplicate analysis, repeated twice.









CHAPTER 4
THE EFFECTS OF THE ALKALI/ACIDIC AIDED ISOLATION PROCESS ON LIPID AND
PROTEIN OXIDATION

Introduction

Oxidative processes inherently occur during handling and/or storage of muscle foods.

Muscle foods are particularly prone to oxidative degradation due to proportionally high levels of

unsaturated fatty acids, metals (heme bound or free), and multitudes of other catalysts and/ or

propagators (80). The actual mechanisms by which oxidative processes change muscle food

proteins are not well understood.

Increasing interests in the acid/alkali aided isolation process warrant study of quality

changes in muscle tissue during the process and the final isolated protein. According to various

studies, protein oxidation is implicated in various protein functionality changes: gelation,

emulsification, viscosity, solubility, and hydration (78, 114, 184-186). Various physical and

chemical protein changes are attributed to the reaction of protein with reactive oxygen species

(ROS) as well as lipid and protein oxidation products, resulting in polymerization, degradation,

and formation of other protein complexes (102, 124, 125, 130, 133, 187-191).

The main intent of the acid/alkali aided isolation process is to obtain quality isolates from

underutilized fish sources, namely pelagic species, by-catch, and fish byproducts. The majority

of these underutilized sources have significant levels of dark muscle (red muscle). Dark muscle

is particularity difficult to work with due to its metabolic nature. Dark muscle is heavily reliant

on oxidative metabolism, and relative to white muscle, depends on higher concentrations of

heme proteins for increased blood flow, mitochondrial function, and lipids for sustained energy

(192). Although, much of these problematic constituents are removed during the isolation

process, residuals can pose difficulties in obtaining a high quality isolate (5). To date, studies

addressing the simultaneous effects of extreme pH and/or readjustment to neutral pH on protein









and lipid oxidation have not been conducted. The objectives of this study, using Spanish

mackerel as a model pelagic species, were to 1) investigate how the extent of lipid oxidation in

low or high pH- adjusted homogenates affects lipid oxidation development in the final Spanish

mackerel protein isolate, and 2) investigate how lipid oxidation in acidic/alkali pH adjusted

homogenates and in the final protein isolate relate to protein oxidation and functionality.

Information regarding overall functionality and quality of isolates obtained from the acid/alkali

aided separation is important for improvement and successful implementation of this process

industrially.

Materials and Methods

Raw Materials

Spanish mackerel was freshly caught during the fall season (September) in Cedar Key,

Florida. The freshly caught Spanish mackerel corresponds to Batch 2 (Materials and Methods,

Chapter 3). Fish were manually skinned, degutted, and filleted. One half of the fillets were

minced using a large scale grinder (The Hobart MFG. Co., Model 4532, Troy, Ohio). Fillets and

ground fish (100 gram quantities) were vacuum packed and stored at -800C until needed. The

initial average moisture and lipid content and pH were 771.76 %, 6.490.015, and and 2.65

0.07, respectively (Table 3-1).

Chemicals

Tricholoracetic acid, sodium chloride, sodium tripoly phosphate chloroform (stabilized

with ethanol), methanol, ammonium thiocyanate, iron sulfate, barium chloride dihydtrate, 12 N

hydrochloric acid were obtained from Fischer Scientific (Fair Lawn,New Jersey). Thiobarbituric

acid, 1,1, 3,3 tetraethoxypropane, streptomycin sulfate SigmaMarker molecular weight

standard, and EZ-Blue stain solution were obtained from Sigma Chemical Co., (St. Louis, MO).

Propyl gallate and (ethlenediaminetetraacetic acid 99%), EDTA, TRIS, Potassium Chloride









(KCL) were obtained from Acros Organics (New Jersey, USA). Pre-cast linear 4-20% Tris-HCL

gradient gel, Laemmli buffer, and B-mercaptoethanol, Power Npac 300 power supply were

obtained from Bio-Rad laboratories (Hercules, CA). The Coomassie PlusTM- Better Bradford

Assay kit was obtained form Pierce (Rockford, II). The Zentech PC Test -Protein Carbonyl

Enzyme Immuno- Assay Kit was obtained from Biocell Corporation. LTD.(New Zealand).

Storage studies and Preparation of Homogenate and Isolate

Frozen, minced Spanish mackerel (-80C) was thawed in vacuum bags under cold running

water and was immediately placed in zip-lock bags at 40C for 4 days. From preliminary studies,

storing freshly caught fish for four days would mimic industrial fish quality for the isolate

process. In addition, raw material storage would ideally allow for detectable measurement

changes in both lipid and protein quality. Homogenates were prepared by diluting the mince with

cold deionized water (1:9) and mixing (45 second at 40C, speed 3) using a conventional kitchen

blender equipped with a rheostat (Staco Inc., Dayton, Ohio, 3p 1500B). The pH of the

homogenate was approximately 6.2; which is lower than homogenate prepared from fresh mince.

Homogenates were slowly adjusted to pH 3, pH 6.5 (control), or pH 11 through dropwise

addition of 2 N HCL or NaOH. After a specified holding time (1 to 8 hrs) at 40C, the

homogenates were adjusted to pH 5.5. The homogenates were centrifuged at 10,000 g for 20

minutes at 40C to collect the precipitated proteins. Protein isolates kept on ice during analysis

were assessed for moisture, pH, and protein content prior to analysis. Isolates were then held for

3 days in one gallon ZiplockTM polyethylene bags at 40C and were analyzed again.

Thiobarbituric Acid Reactive Substances (TBARS)

TBARS were assessed according to Lemon (193) as modified by Richards and Hultin

(43). Briefly, one gram of sample was homogenized with 6 mL of trichloracetic acid (TCA)

solution composed of 7.5% TCA, 0.1% propyl gallate, and 0.1% EDTA dissolved in deionized









water. The homogenate was filtered through Whatman No. 1 filter paper. The TCA extract (2

mL) was added to 2 mL of thiobarbituric acid (TBA) solution (0.23% dissolved in deionized

water). The TCA and TBA solutions were freshly prepared and heated if needed. The TCA/TBA

solution was heated for 40 minutes at 1000C. The samples were centrifuged for 10 minutes (4C)

at 2,500 g and were read at 532 nm using a spectrophotometer (Agilent 8453 UV-Visible,

Agilent Technologies, Inc., Palo Alto, CA). A standard curve was constructed using 1,1, 3,3

tetraethoxypropane (193).

Protein Quantification and Characterization

Protein isolate concentrations were assessed using a modified Biuret method (194). The

Bradford method was used to quantify proteins levels, either in low concentration (<1 mg/mL) or

limited quantity (195) using the Coomassie PlusTM- Better Bradford Assay kit. Protein

composition in isolates was analyzed with electrophoresis. SDS-PAGE (sodium dodecyl sulfate

polyacrylamide gel electrophoresis) (196). Protein concentrations were adjusted to 3 mg/ mL.

SDS-PAGE samples were prepared by combining 0.33 [L of protein sample, 0.66 [L of

Laemelli buffer, and 50 [iL of B- mercaptoethanol (Img/ mL protein concentration). Samples

were boiled for 5 minutes in 1.5 mL cryogenic tubes and were then cooled on ice. To elucidate

changes in disulfide bonds, electrophoresis samples were prepared without P-mercaptoethanol

using 50 [iL of Laemmli buffer instead. Pre-cast linear gradient 4-20% acrylamide gels were run

for 1 h at 200 V; mA 120 with a Power Npac 300 power supply. Samples were applied to the

gels in 10 and 15 [L aliquots. Molecular weight standards (205,000- 6,500 Mw) were run in 5

and 15 uL quantities. Gels were set with 12% TCA for 1 h and were stained with EZ-Blue stain

solution (minimum of 1 hr with agitation). Gels were then destined with deionized water for 18

hrs with agitation, changing out the water numerous times.









Rheology

To compare gel quality among isolates, oscillatory rheology was conducted using an AR

2000-Advanced Rheometer (TA Instruments, New Castle, DE). Viscoelastic behavior and

thermal stability of the protein isolates were evaluated and characterized according to Kristinsson

and Demir (13). Prior to analysis, protein isolates were adjusted to an average moisture content

of 85% through addition of deionized water. Sodium tripolyphoshate (0.3%) and sodium

chloride (600 mM) were added to the moisture adjusted isolates according to Liang and Hultin

(197). Isolate additives were incorporated using a mortar and pestle. Samples were adjusted to

pH 7.2 through dropwise addition of 1 N NaOH. Freshly prepared gels were analyzed in

duplicate using an oscillating acrylic plate (40 mm). Approximately 20 grams of sample were

place on the acrylic plate (50C). The gap between the head and acrylic plate was set to 500 itm.

After the gap was reached, access sample was carefully removed with a stainless steel spatula.

Two temperature ramps were used to measure gel forming ability. A heating ramp (5-800C) and

a cooling ramp (80-50C) at rates of 20C/ min. Oscillation procedures were conducted using 0.01

maximum strain and an angular frequency 0.6284 rad/s. Changes in the storage modulus (G') as

a function of temperature were analyzed. The storage modulus is indicative of elasticity (198).

Protein Oxidation

Protein carbonyl content was measured for the pH derived isolates using the Zentech PC

Test (Protein Carbonyl Enzyme Immuno-Assay Kit). Briefly, protein carbonyls are derivatized

with dinitrophylhyrazine (DNP). The DNP bound proteins bind to the Elisa plate. The plate

bound DNP is then reacted with an anti-DNP-biotin antiobody. A stretavidin-linked horseradish

peroxidase is then bound to the complex. The chromatin reagent (contains peroxide) is then

added. The addition of peroxide initiates 3,3',5,5'-tetramethylbenzidine (TMB) oxidation and

thus a color reaction. The application of the stopping reagent (an acid) terminates the reaction.









The yellow chromophore is measured at 450 nm in a microplate reader (Molecular Device

Corporation, Sunnyvale, CA). A standard curve is constructed using standards provided in the

kit, concentration (nmol/g) against absorbance (450 nm). The inter-assay variation was within kit

protocol recommendations (<15%). Prior to analysis, all samples (0.5 g) were homogenized 1:1

with 20 mM Tris-HCL, pH 7.5, 0.5 M KCL buffer and a glass homogenizer. This buffer rather

than water was used to maintain pH and increase isolate solubility according to Kristinsson and

Hultin (199). Connective tissue and/or insoluble components were centrifuged out at 2,000 g

prior to analysis. A glass homogenizer versus a motorized tissue tearer was used to reduce

foaming. All samples were normalized to 1 mg/mL and dilutions were performed when

necessary using the Enzyme Immuno Assay (EIA) kit buffer. All samples and standards were run

in at least triplicate.

Analysis of Moisture and pH

Moisture content was assessed using a moisture balance (CSI Scientific Company, Inc.,

Fairfax, VA). The pH was measured using an electrode ( Thermo Sure-Flow electrode (Electron

Corp., Waltham, MA) and pH meter (Denver Instrument, model 220, Denver, CO). The

homogenate samples were directly measured for pH. Isolates samples were diluted with

deionized water (1:10) to measure pH.

Statistical Analysis

The SAS system was used to evaluate data. A two-way ANOVA model was used to

evaluate statistical significance in lipid oxidation (TBARS, PV), and protein oxidation, with time

and among pH treatments as factors. All analyses were performed in at least duplicate and all

experiments were replicated twice. Least square means were used for post hoc multiple

comparisons with Tukey adjustment controlling for Type I error.









Data were reported as mean standard deviation. The analysis of variance and least

significant difference tests were conducted to identify differences among means, while a Pearson

correlation test was conducted to determine the correlations among means. Statistical

significance was declared atp < 0.05.

Results

Initially, to ensure that the quality offish was approximately equivalent to studies in

Chapter 3, Spanish mackerel batch 2 mince was stored at 40C for four days. Homogenates were

subjected to pH 3 as per findings in Chapter 3, in which the greatest overall levels of homogenate

lipid oxidation with respect to pH occurred at pH 3 and 3.5. Because the overall levels of lipid

oxidation with respect to homogenates 3 and 3.5 were insignificant from each other, and the fact

that pH 3 is an intermediate of pH values (2.5, 3, and 3.5) tested in Chapter 3, pH 3 was chosen

as the primary acid treatment. Likewise, no significant differences in lipid oxidation were

observed among homogenates at alkaline pH of 10.5, 11, or 11.5, so the intermediate pH 11 was

chosen as a treatment for this study. This pH is commonly used in the alkaline process. Prior to

experimentation, moisture content and pH of the starting homogenate were 97.60.707% and

6.2150.077.

Oxidative Stability of Acid/Alkali Derived Protein Isolates

Levels of lipid oxidation were assessed throughout the isolation process using one and

eight hour exposure times at either pH 3 or pH 11. The isolates were obtained through

centrifuation at pH 5.5. The isoelectric point of fish proteins is close to pH 5.5, enabling for

lowest repulsion and highest yield of extracted proteins (mainly of myofibrillar nature)(] 7, 200,

201). Isolates were assessed for oxidation right after isolation and after 3 days of 40C storage.

Average moisture and protein contents of acid, alkaline, and control (pH 6.5) isolates from one

and 8 hour hold times were similar and are summarized in Table 4-1.









The first centrifugation step (Figure 2-2) was omitted in this experiment to increase protein

isolate yield and retain more lipids (13). Kristinson and Demir (13) observed a gel-like sediment

in the soluble phase following the first centrifugation of acidified mullet. Similarly, a gel-like

sediment was observed after readjusting Spanish mackerel homogenates from extreme and

neutral pH to pH 5.5 (Figure 4-1). This phase was grayish white, soft, slimy, and easily

disrupted. Unlike the sediment observed by Kristinsson and Demir (13), the gel like sediment

was found above and adjacent to the collected proteins. Despite differences in processing, this

sediment was observed. This sediment, mainly neutral lipids, was removed with a plastic spatula.

When comparing the two processes, it is evident that more lipids are retained as exhibited by the

Spanish mackerel sediment layer when the first centrifugation step is excluded.

Holding times (1 or 8 h) did not significantly affect lipid oxidation in homogenates (See

Figures 4-2,3,4). The supernatant from the isolation of homogenate at pH 3 for eight hours had

the greatest level of TBARS (gtmol MDA/kg of tissue) compared to all other pH homogenates,

isolates, and supernatants at pH 3, 6.5, and 11 (Figure 4-3).The levels of TBARS in the

supernatant (Figure 4-3) held at pH 3 for 8 h (94) were significantly (p<0.001) greater compared

to the corresponding isolates, or 52.59 5.48 and 31.176.97 umol MDA/kg tissue, respectively.

The level of lipid oxidation in the pH 3 isolate after 1 hour storage was significantly (p <0.03)

less than the corresponding supernatant. The amount of oxidation measured in the pH 11 derived

isolate, after an 8 hour holding time, was significantly less than the same isolate stored for three

days (p<0.0291) (Figure 4-3). Levels of oxidation in pH 11 derived isolate (Figure 4-4), after 8

hours of holding, were significantly lower than the corresponding isolate at pH 6.5 (p<0.0289).

TBARS levels measured in the pH 6.5 supernatant, after an 8 hour hold, were significantly

greater than TBARS levels in isolate made after an 8 hour hold at pH 11 (p<0.0053).









Significantly (p<0.0068) higher TBARS formed in the homogenate exposed to pH 3 for one hour

compared to the pH 11 isolate, derived from an 8 hour homogenate holding time (Figure 4-3).

As shown in Figure 4-3 and 4-4, isolate from the pH 11 treatment obtained after 8 hours of

holding had significantly greater levels of oxidation after 3 days of storage than the

corresponding freshly prepared isolate (p <0.03).

Levels of protein oxidation were assessed among the isolates after 1 hour, 8 hours, and 8

hours followed by three days of storage (See Figure 4-5). Overall levels of oxidation were

significantly different (p<0.0001) with respect to pH treatment: pH 11> pH 6.5> pH 3 and

isolate: 1 hr> 8hr> 8hrs+ 3days. Specifically, isolate from the pH 11 treatment (1 h holding)

contained significantly higher levels of protein oxidation compared to all other combinations of

isolate and pH, with exception of these isolates from the 6.5 treatment (1 h holding) and the pH

11 treatment (8 hr holding) (Figure 4-5). The isolate produced from holding the 6.5 homogenate

for one hour, was significantly higher in carbonyls than the corresponding isolate at pH 11 (p<

0.0031). The isolate formed from a holding time of 8 hours at pH 11 had significantly

(p<0.0001) greater levels of protein oxidation than the corresponding pH 11 and pH 6.5 isolates

(p <0.0006) formed from a one hour holding time. The pH 11 isolate derived from an 8 hour

holding time was more oxidized than the corresponding pH 6.5 isolate (Figure 4-5). The pH 6.5

isolate carbonyl levels from the 8 h holding treatment were significantly less than carbonyls after

1 hour of holding. After the isolates from the 8 hr holding treatment were stored for three days, a

significant decrease (p<0.03) in protein carbonyls compared to the fresh 8 hour isolates was

observed among pH 1 land 6.5 treatments (Figure 4-5). Mean values were comparable to an

average carbonyl content of 1.5 nmol/mg of protein by Liu and Xiong (189) in red and white

muscle of chicken.









Gel Forming Ability of Isolates

In terms of theological differences, the storage modulus (G') of isolates was not

significantly affected by holding time or storage; however, the elasticity (G') of isolates as

significantly (p< 0.0002) affected by pH treatment with pH 11 isolate having the highest G' (Pa),

followed by pH 6.5, and then pH 3 (Figures 4-6).

Examples of rheograms generated during the oscillatory testing of isolate pastes derived

from 1 hour, and isolates derived from 8 hr pH holding times in addition to three days of

refrigerated storage are shown in Figures 4-7,8,9. Initial G' values were similar in the pH 11 and

6.5 isolates. These initial G' values were greater than G' for pH 3 isolates. Initial increase in the

storage modulus (G') on heating were observed at 300C for all samples at pH 3.0, 6.5 (control),

and pH 11 (Figures 4-7,8,9). During heating, one noticeable transition temperature (Tm) was

observed at 400C for the isolates at pH 6.5 (control) (Figure 4-9). The same Tm, although

significantly less noticeable, was also observed for pH 3.0 and pH 11 isolates (Figure 4-7,8).

Significant increases in G' were observed between 50 and 600C for isolates derived from pH 3

and 6.5 homogenates and between 40 and 55C for isolates derived from pH 11 homogenates

(Figures 4-7,8,9). G' values were stabilized between 70 an 800C. Significant increases in G'

during the cooling step were observed in all isolates. During cooling, the greatest increase in G'

occurred in the pH 11 isolates, followed by pH 6.5 (control), and pH 3.

Protein Composition of Homogenates and Recovered Isolate Fractions

SDS Page, with and without, mercaptoethanol showed differences in protein composition

among isolate fractions. The inclusion versus the exclusion of mercaptoethanol in samples was

applied to the gels to assess whether differences among protein fractions, if any, were due to

covalent (S-S) bonds or non-covalent cross-linking, possibly attributed to protein oxidation.

Cross-linking was not observed in any of the samples; however, hydrolysis was observed for all









pH (11,6.5, 3). Hydrolysis was evident in isolates derived from adjusting and holding

homogenates at acidic (3), alkaline (11) or control (6.5) pH for 1 hour, followed by readjustment

to pH 5.5 in Figure 4-10: Lanes 2-5 (pH 11); Lanes 10 and 11 (pH 6.5); Lanes 14 and 15 (pH 3).

Hydrolysis was also observed in isolates derived from adjusting and holding homogenates at

acidic (3), alkaline (11) or control (6.5) pH for 8 hour, followed by readjustment to pH 5.5 in

Figure 4-11: Lanes 4 and 5 (pH 6.5); Lanes 8 and 9 (pH 3); and Lanes 13 and 14 (pH 11).

Proteolysis was evident in isolates derived from adjusting and holding homogenates at acidic (3),

alkaline (11) or control (6.5) pH for 8 hours, followed by readjustment to pH 5.5, and finally

followed by 3 days of refrigerated storage in Figure 4-12: Lanes 4 and 5 (pH 6.5), Lanes 6 and 7

(pH 11); Lanes 14 and 15 (pH 3). Hydrolysis was also evident in pH 3, 11, and (6.5) control

homogenates at initial pH adjustment (Figure 4-13) and 1 and 8 h holding (Figures 4-14,15), at

pH 3, 6.5 (control), and 11.

The most prevalent proteins fractions were myosin heavy chain (MHC)(-205 Kda) and

actin (- 43kDa), and tropomyosin (35.5 kDa). Other identified proteins included desmin (-56

kDa), troponin T (41 kDa), and topomyosin-beta (-38 kDa).

Discussion

The isolation of fish proteins following pH readjustment from extreme pH is performed for

greater protein recovery and to recover proteins with modified functional properties. With

greater protein recovery, co-extraction of unwanted components (lipids and heme proteins) and

consequentially oxidative degradation could occur, negatively affecting the quality and

performance of the isolated proteins.

Changes in pH of Mince with Storage

The pH (6.2) of mince stored for 4 days was less than initial pH values of unstored mince

homogenate (6.45) as reported in Chapter 3, indicative of metabolite accumulation and cellular









degradation post mortem (31). Various postmortem changes can influence oxidative occurrence

within the fish muscle: decreases in ATP and increases in ATP byproducts, loss of reducing

compounds ascorbatee, NAD(P)H, glutathione, etc.), increased levels of low molecular weight

transition metals (copper, iron), oxidation of iron in heme proteins, disruption of cellular

structure, loss of antioxidants, and accumulation of calcium and other ions. With mincing,

postmortem muscle tissue can experience conditions very similar to ischemia-reperfusion.

Consequentially, production of superoxide and hydrogen peroxide can occur, possibly leading to

protein and lipid oxidation (5).

Lipid Oxidation During Extreme pH Adjustment and Readjustment to 5.5

It was observed that through separation of Spanish mackerel proteins by pH readjustment

from extreme pH (3, 6.5, 11) to neutral pH (6.5) that lipids were retained (Figure 4-1).

Kristinsson and Demir (13) reported greater lipid levels in acid mullet isolates when the first

centrifugation step was skipped. This could be explained by the co-precipitation of membrane

and neutral lipids with proteins at isolelectric pH. Co-precipitation is most likely a result of

increased protein hydrophobicity and surface activity with pH adjustment as observed by

Kristinsson and Hultin (18) with cod myosin. Increases in hydrophobicity result in enhanced

emulsification ability and interaction with hydrophobic residues of lipids and heme proteins.

Increases in hydrophobicity is most likely attributed to denaturation of myofibrillar and

sarcoplasmic proteins, as supported by low extractability of rockfish proteins during the

acid/alkali aided process (202). Kristinsson and Hultin (199) have shown that significant

structural changes occur in myosin upon incomplete refolding of cod myosin with the

readjustment to pH 7.5 from acid and alkaline treatment. Partial refolding of myosin's tertiary

structure could facilitate hydrophobic interactions. Furthermore, low levels of retained

hemoglobin upon aggregation can pose lipid oxidation problems. Undeland et al. (67) found that









extensive hemoglobin chelation by 0.2% STPP and 0.2% EDTA during acid isolation of herring

still gave rise to low TBARS levels. These results emphasize that despite significant removal of

hemoglobin in the pH cycle, hemoglobin could still be problematic regardless of concentration

found in the isolate.

The greatest level of TBARS were found in the supernatant from the isolation of pH 3

Spanish mackerel homogenate held for eight hours relative to all other pH homogenates, isolates,

and supernatants at pH 3, 6.5, and 11 (Figure 4-2 and 4-4). Additionally, the pH 3 supernatant,

after 1 h of holding, was greater than the corresponding isolate (Figure 4-2). According to

Undeland (16), lipids were present in the highest amounts in the emulsion supernatantt) layer

followed by the gel sediment in the acidic pH (2.7) extraction of herring. Lower lipid levels were

observed from alkaline extraction (10.8) in which the emulsion layer had 37% lipid and the

isolate had 18% lipid (16). Just as Undeland (67) reported, the Spanish mackerel supernatant

fraction at acidified pH (3) contained the highest TBARS levels (Figure 4-3). During the

conventional production of horse mackerel surimi performed by Eymard et al. (138), significant

TBARS were found in the wash water. Lipid oxidation products formed during acid processing

could be water soluble or at least suspendable, which partially explains why the supernatant

fractions (Figures 4-2 and 4-3) had such high TBARS.

Overall, lipid oxidation in the pH treated isolates were not significantly different (Figure

4-4). Lack of significant differentiation in TBARS between alkali and acid processing compared

to other published studies is attributed to skipping the first centrifugation step. Kristinsson and

Demir (13) observed significant differences among color values when incorporating the first

centrifugation in the separation of mullet and croaker proteins; however, in absence of

centrifugation both acid and alkaline isolates were negatively affected in color value, suggesting









more retention of heme proteins. With both centrifugation steps, the alkaline treatment was

whiter in color (high L value) and was less yellow (low b value) than acid isolate. This was

attributed to increased removal of heme proteins, found in the supernatant of the first

centrifugation step (13). Kristinsson and Hultin (51) concluded that alkaline treatment resulted in

greater hemoglobin solubility and reduced co-precipitation at pH 5.5 as per discussion of

homogenate pH adjustment in Chapter 3. From the lipid oxidation studies, it appears that acidic

and alkaline pH processing of Spanish mackerel with exclusion of the first centrifugation is

equally problematic.

Holding times did not significantly affect the amount of oxidation among Spanish

mackerel isolates within treatments (Figures 4-2,3,4). According to Undeland et al. (203), the

holding time at acidic pH (4, 30, or 70 minutes) prior to acidic isolation of herring isolates did

not affect overall levels of oxidation. In another study, Kristinsson (51) found that the

prooxidative activity of trout hemoglobin at pH 3.0 and 3.5 versus pH 2.5 was unaffected by

holding time prior to pH readjustment. Kristinsson et al. (204) reported no significance

difference in lipid oxidation among the alkaline or acidic pH isolation of catfish proteins;

however, when the first centrifugation was skipped, significantly more lipid oxidation resulted

relative to the other treatments. The lack of significance among the acid and alkali treatment

concurs with some previous findings. The significant difference between alkaline and control pH

isolates, after 8 hours of holding, is interesting (Figure 4-4). Kristinsson and Liang (205),

reported reduced levels of lipid oxidation in an alkaline isolate relative to a surimi control and

acid isolates from Atlantic croaker. As suggested by Kristinsson (12), this could be a result of

hemoglobin adopting a more stable tertiary structure upon alkaline treatment compared to acid

treatment (51). Additionally, oxidative products were most likely removed through









centrifugation at 5.5, impart explaining a reduction in lipid oxidation product in the alkaline

isolate made an after 8 h holding (Figure 4-3).

Lipid oxidation was observed among all isolate treatments (Figure 4-4). Similarly, the

greatest increase in lipid oxidation during 7 days of storage of beef heart surimi, washed with

buffer at 5.5 and 7.0, was observed in the 5.5 treatment by Srinivasan et al. (136). It would be

interesting to investigate the effects of readjustment to pH 7, as previous studies suggest it could

pose less lipid oxidation problems (190); however, reduced protein recovery would most likely

result. With the final protein isolate at pH 5.5, it is to be expected that lipid oxidation would

proceed significantly faster than at pH 7 as hemoglobin is more pro-oxidative at pH below

neutrality as observed in a washed cod model systems (43, 51, 206). Lipid oxidation in isolates,

after storage at low pH (5.5) was evident. This can be explained by rapid autoxidation of heme

proteins in the isolates at pH 5.5. According to Richards and Hultin (43), increases in TBARS

were due to decreases in pH from 7.6 to 6.0 in a hemoglobin mediated washed cod system.

Protein Oxidation

Overall levels of protein oxidation (Figure 4-5) were highest at pH 11, followed by pH 6.5,

and pH 3. The effect of pH on lipid and protein oxidation has been studied in various meat

systems. Similar to observed pH affects in this study, Xiong et al. (124) and Wan et al. (123)

have shown that myofibrils from bovine cardiac muscle improved in functionality with decreases

in lipid oxidation. In a study conducted by Srinivasan et al. (190), beef heart was washed with

various pH buffers, ranging from pH 7.0 to 5.5. After seven days of storage, the levels of lipid

and protein oxidation were significantly higher at pH 5.5 and 6.0 than at pH 7.0. Srinivasan (136)

reported a simultaneous increase in markers of lipid (TBARS, conjugated dienes) and protein

oxidation carbonylss) with lowering of pH in bovine cardiac surimi. Kenney et al. (207) found

that washed beef heart surimi had superior textural properties compared to fish surimi. Similar to









dark muscled fish, beef heart contains an abundance of heme proteins, iron, and polyunsaturated

fatty acids (208).

The mince had significantly lower levels of protein oxidation compared to all isolation

treatments at pH 3, 6.5, and 11 (See Figure 4-5). Hultin and Luo (209) reported that protein

modifications will occur with changes in muscle temperature, salt concentration, or pH, leading

to significant degradation and polymerization of proteins due to protein oxidation (209). This

phenomenon could explain why the mince had lower levels of oxidation, as the isolates were

obtained through pH processing. Interestingly, carbonyl levels were higher in the isolates made

after one hour holding at pH 6.5 and 11 compared to pH 3. According to Srinivasan et al. (136),

bovine heart surimi washed with antioxidants above pH 5.7 produced stronger gels than surimi

washed below pH 5.7. The differences in G' at various washing pH was attributed to oxidative

modification of protein residues and or protein-lipid oxidation interactions, affecting the

antioxidant efficacy. The observations by Srinivasan et al. (136) could explain the differences in

the protein oxidation in Spanish mackerel isolates with pH, as different protein interactions occur

in the presence of oxidized lipids or proteins, more so at high pH. Kristinsson and Hultin (199)

reported partial refolding of cod myosin with pH 7.0 readjustment from alkaline pH. Partial

refolding at high pH could make alkaline treated isolates more susceptible to protein oxidation.

In the pH 11 Spanish mackerel isolates, protein carbonyl content significantly increased

after 8 hours of holding compared to one hour of holding (Figure 4-5); while there was an

apparent decrease in TBARS after 8 hour hold compared to the 1 hour hold. Similar trends

between protein and lipid oxidation were observed by Srinivasan and Hultin (210). In this study,

significant increases in protein and lipid oxidation were observed when washed cod muscle was

subjected to free radical exposure at pH 6.8 for 2 hours. Interesting kinetics between protein and









lipid oxidation were observed in this system. With storage, protein oxidation carbonyll) levels

increased quickly, while there was a lag phase in lipid oxidation. Lipid oxidation began to

accelerate at a point where protein oxidation declined slightly. Protein oxidation levels then

experienced another rapid increase and lipid oxidation levels also increased. Hultin (210)

concluded that perhaps protein or lipid oxidation could propagate each other. Similarly, longer

pH holding times may afford greater production of ROS through lipid oxidation, providing

adequate initiation of protein oxidation (211).

Significant decreases in protein oxidation after 3 days of storage occurred among the pH

11 and 6.5 isolates compared to the freshly prepared isolate (Figures 4-5). This is in contrast to

other storage studies, in which, protein and lipid oxidation increased over a duration equivalent

to 7 days or more (63, 70, 136). It is possible that oxidized, aggregated proteins or insoluble

proteins could not be adequately assessed via the carbonyl assay, as solubilized proteins are

applied to the protein carbonyl kit. With storage, protein degradation was observed in the gels.

Protein aggregation may have possibly occurred in the isolates as protein bands were observed at

the top of the SDS-PAGE gels. It is also possible that protein oxidative products served as

intermediary reactive species during the oxidative process, becoming pro or antioxidive with

changes in pH and conformation environment (212, 213). Protein oxidation products could have

served as antioxidants for the alkaline pH and control pH homogenate treatments and as oxidants

in the pH 11 derived isolate, stored at pH 5.5. It has been proposed that protein radical species

are highly dynamic and even transient, depending upon interaction with other molecules and

environmental conditions (213). In a BSA hydrogen peroxide activated protein oxidation model

study conducted by Ostdal et al. (211), BSA successfully oxidized urate, linoleic acid and a









complex amino acid residue. Thus, protein radicals could play various roles in the oxidation of

muscle foods either as end-products or intermediates.

Carbonyl measures are readily performed and acceptable measures of complex-muscle

system samples (112, 210). Carbonyls are a global form of oxidation and can be formed as result

of various reactive species (76), unlike other protein oxidation products. Strict reliance on

carbonyl levels as an indicator of protein oxidation is tenuous, as various other compounds form

as result of protein oxidation. A more comprehensive attempt to measure protein oxidation could

include the measure of dinitrotyrosine, 3- dinitrotyrosine, and modified AA residues (214).

Protein carbonyl content in isolates would be best analyzed by a method requiring chemical

digestion. Chemical digestion would eliminate solubility problems and possible underestimation

of protein oxidation levels, upon analysis and/or plating. Kjaersgard et al. (215) have

successfully used two-dimensional immunoassaying followed by mass spectroscopy in the

identification and quantification of oxidized proteins and residues in fish species.

Gel Forming Ability

Noticeable differences in rheograms among the isolates derived from pH 3, 6.5 (control),

and pH 11 homogenates were observed. All isolate pastes showed initial increases in G' at 300C.

Yougsawatdigul and Park (200) also observed initial increases in surimi as well as alkaline and

acid isolates at 340C (200). Unlike the acid (pH 3) and alkaline (pH 11) isolates (Figures 4-7,8),

a significant decrease in G' was observed during the heating step in the pH 6.5 (control) isolates.

The lowest G' was observed at approximately 470C for isolates derived from pH 6.5

homogenates (Figure 4-9). Yongsawatdgul and Park (200), Kristinsson and Liang (205), and

Kristinsson and Hultin (18) observed similar rheology differences among surimi and pH treated

samples of rock fish, Atlantic croaker, and cod muscle proteins, respectively. The Tm (400C)

observed in the control (6.5) isolate was not observed in the pH treated isolates. Similarly, Hultin









and Kristinsson (18) reported absence of this peak in pH treated cod proteins compared to

untreated cod proteins, and this observation was attributed to different protein-protein

interactions among the control and pH treated proteins (18). Ingadottir (216) and Theordore

(217) observed significant theological differences among surimi and pH treated samples for

tilapia and catfish, respectively. These theological differences between controls and acid/alkali-

aided isolates may be due to protein denaturation with pH treatment (218). In agreement with

findings by Kristinsson and Hultin (18) and Kristinsson and Liang (205) for washed rockfish and

croakers, a large drop in storage modulus occurred in pH 6.5 (control) isolate samples.

According to Kristinsson and Liang (205), a large decrease in G' could be attributed to

actyomyosin gel formation. Non-covalent protein interactions between actin and myosin are due

to various conformational changes. The Tm observed at 400C in the pH 6.5 isolates marks the

beginning of gel network formation through various conformational changes including myosin

light chain dissociation from myosin heavy chains (MHC) (219, 220), exposure of hydrophobic

groups, namely tryptophan, and transformation of the heavy myosin rod (tail) from a a- helix to a

random coil. G' significantly increased on cooling among all isolate samples.

The pH treatment had a significant affect on gel strength, as alkaline pH gave stronger gels

than the control (6.5) and the acid (pH 3) prepared samples (Figure 4-6). This is in disagreement

with previous findings by Kristinsson and Liang, in which croaker alkaline and acid isolates had

greater G' than surimi. Kristinsson and Hultin (18) reported similar G' values among acid (pH

2.5), alkaline (ph 11), and untreated (pH 7.5) cod muscle proteins; however, different

mechanisms of gelation among the treatments were observed. Shikha et al (221) studied the

effect of acid pH shifting in walley pollack. Surmi that was acidified and then readjusted to

neutral pH had lower gel strength (g/cm2) than the control (pH 7.2 mince). Various other studies









reported difference in G' values among alkaline, acid, and untreated samples for catfish, tilapia,

mullet, and Spanish mackerel (13, 216, 218). Theodore (217) reported G' trends comparable to

the Spanish mackerel isolate samples, in which the alkali treated catfish samples were followed

by surimi and then the acid samples in G' values. In agreement with this study, Kristinsson and

Demir (13) found that alkaline processing of Spanish Mackerel and mullet gave greater G'

values than surimi processing, followed by acid processing. Higher G' in the alkaline samples

indicates greater protein-protein interactions on cooling, enabling better gel-network formation

upon setting. According to Kristinsson and Hultin (199), proteins can unfold through extreme

pH and upon readjustment only partially refold. Specific changes in protein structure with

unfolding and refolding could explain the differences in G' among the isolates and control (pH

6.5) treatment. At low and high pH, proteins unfold and denature. Kristinsson and Hultin (199)

reported dissociation of the cod myosin head group with unfolding at extreme pH. With

readjustment to pH 7.5, the alkaline and acid treated myosin partially refolded; however, alkaline

treated cod myosin remained more dissociated. Due to more flexibility than acid isolates upon

refolding, alkaline treated proteins could form a better gel network. Increased exposure of thiol

groups could also explain the difference in G' among the alkaline (pH 11), control (pH 6.5), and

acid isolates (pH 2.5). Kristinsson and Hultin (199) found an increases in thiol exposure when

cod myosin was adjusted to alkaline pH and readjusted to pH 7.5. This increase in sulfhydryl

group (-SH) reactivity could explain the higher G' in the Spanish mackerel alkaline isolates

versus the control (pH 6.5) and acid isolates (Figure 4-6). The acid isolates did show an increase

in sulfhydryl reactivity with refolding according to Kristinsson and Hultin (199), but this

increase was less compared to alkaline refolding. A lower G' in the acid versus alkaline

processing could be due to less thiol interaction among proteins.









Textural Changes and Protein and Lipid Oxidation

Greater overall levels of protein oxidation (Figure 4-5) in the alkaline treatments could

explain high G' (Figure 4-6). Protein oxidation could lead to textural changes in muscle food

products (86), and could explain the differences in gel strength among isolates. As a consequence

of oxidation, proteins unfold and exposes their hydrophobic residues. Srinivasan and Xiong

(130, 136) reported extensive sulfhydryl loss with protein oxidation. This increase in

hydrophobic affinity could promote protein interaction and aggregation, improving gel strength.

Protein oxidation proceeds in a manner comparable to lipid oxidation in the presence of

oxidizing lipids and/or other reactive oxygen species (ROS) (101, 222). Free radicals will react

with either amino acid (21) side chains and/or peptide backbones. Proteins will often fragment

and complex upon free radical attack (86). Hydroxyl groups are predominantly responsible for

protein oxidation in the absence of lipids (106). Certain AA groups are preferentially oxidized,

the most susceptible is cysteine. Myosin, the most abundant and important muscle food protein,

has 40 free thiol groups. (223). The abundance of myosin compared to sarcoplasmic proteins

(224) warrants investigation of how myosin is affected during the acid/alkali aided process (13).

Protein carbonyls are product of: AA residue oxidation, peptide backbone degradation, protein-

reducing sugar interaction, and interaction with non-protein carbonyl groups (86). Oxidized

proteins will nucleophilically attack proximal proteins, forming cross-linkages. According to

Xiong and Decker (87), protein cross-linkages result in protein polymerization, aggregation, and

insolubility. Cross-linking may cause protein aggregation and increased G' values (70). The

quality of muscle foods is greatly impacted by these physiochemical protein changes.

Intermolecular interactions may be weak with protein oxidation; however, oxidatively induced

polymerization may counteract these weak bonds and strengthen non-covalent interactions,

through increased rigidity and low mobility within the gel network.









The duration of storage for the Spanish mackerel isolates was short and the development of

carbonyls did not exceed 2 nmol/mg (Figure 4-5). According to Lee et al. (104) and Nishimura

et al. (225) there is a very high, positive correlation (r=0.99) in beef heart surimi, between

protein carbonyls and G' when there is sufficient protein oxidation and storage. It was observed

that increases in G' occurred with increases in protein carbonyls to at least (136, 190, 191).

According to other studies (136, 190, 191), protein oxidation improved gel forming ability in

muscle tissue through greater gel network formation. In contrast to these studies, the level of

oxidation was significantly less 15 nmol/mg protein. With increased storage, a greater correlation

between G' and protein oxidation as well as lipid oxidation may have been observed.

Lipid oxidation in the alkaline isolate after 8 h holding was significantly less than with 1

hour holding (Figure 4-4). Protein oxidation levels also significantly increased after 8 hr of

holding compared to 1 hr of holding (Figure 4-5). A possible correlation between protein

oxidation and lipid oxidation and G' in the isolates exists. As per discussion in previous "Protein

oxidation" section, protein oxidation could be propagated by lipid oxidation. According to Saeed

and Howell (70), the elastic modulus in Atlantic Mackerel fillets stored at -200C and -30C was

significantly decreased with the incorporation of vitamin E (500 ppm). Because the transfer of

free radicals or protein-lipid interactions could proceed from lipid oxidation (70), it was inferred

that vitamin E quenched lipid radicals and reduced protein oxidation formation. This study

elucidates that lipid oxidation is an initiator of protein oxidation.

Protein oxidation did not proportionally increase with lipid oxidation (Figure 4-4, 4-5) in

the isolates. This is in contrast to other studies in which increases in protein oxidation are

proportional to increases in lipid oxidation. According to Smith (122), decreases in myofibrillar

gel strength were observed in deboned turkey with increases in TBARS, and decreases in









solubility and myosin ATPase activity (122). The authors attributed the decrease in solubility and

ATPase activity to protein oxidation. The oxidative effects of the acid/alkali aided process on

lipid and proteins has received little attention and deserves further investigation.

SDS page

Electrophoretic patterns (Figures 4:7-12) revealed hydrolysis in alkaline and acid samples.

Hydrolysis is a concern as it could negatively affect gel strength due to gel softening (226, 227).

Several studies suggest that proteolysis during pH processing can result in poorer gels.

According to various studies, the alkaline/acid-aided process can initiate protease activity (16,

201, 228). Enzymes can be activated at low pH and at the isoelectric pH 5.5 (201, 229). Similar

to this study, hydrolysis was observed in acid/alkali aided samples by Kristinsson and Demir

(13), Choi and Park (16, 201), Undeland et al. (36) and Ingadottir (216) in catfish, spanish

mackerel, mullet, croaker, pacific whiting, herring, and tilapia. A band of approximately 150

KDA was observed in the gels (4-10,11,12, 13, 14, 15), similar to other studies analyzing the

affects of pH processing on fish proteins (205, 230-232). Myosin heavy chain (MHC)

degradation is indicated by this band (230). The observed hydrolysis among the Spanish

mackerel isolates is most likely attributed to protease activation versus alkaline or acid

adjustment as the control (pH 6.5) also contained hydrolysis. Because proteolytic hydrolysis was

observed among all treatments and it did not negatively affect the gel strength of either alkaline

or control treatments, hydrolysis could not explain the differences in gel strength in the alkaline,

acid, or neutral isolates (Figure 4-6). SDS page results do therefore suggest that the differences

in gel strength are impart due to changes in comformation with pH treatment, or these are

chemically induced changes linked to oxidation. With storage, protein aggregation or

degradation was not observed in the gels; however, it is possible that aggregates were too large

to enter the gel.









Conclusion

Protein oxidation and lipid oxidation were significantly affected by holding time and pH.

The pH 11 isolate contained the highest overall levels of protein oxidation, followed by pH 6.5,

and pH 3. Overall, levels of lipid oxidation were similar among the pH treatments, and

significant removal of lipid oxidation occurred with centrifugation at pH 3, after an 8 h holding.

A positive correlation between lipid oxidation and protein oxidation was not apparent, as

observed in other studies. G' was consistently higher in alkaline isolates compared to pH 6.5 and

pH 3. Increases in G' may be attributed to conformational changes at higher pH (6.5 and 11); and

or protein modifications due to protein oxidation.

As research currently exists, positive correlations between lipid and protein oxidation and

gel forming ability can only be tentatively made. The study of protein oxidation in muscle foods

has only just begun and further findings are necessary to substantiate the influence of lipid and

protein oxidation on protein physiochemistry and functionality.










Table 4-1. Moisture content (%) and protein content (%) of isolates obtained after readjustment
from pH 3, 11, or 6.5 to pH 5.5, and held for 1 and 8 hours. Moisture and protein
content means and standard deviations were calculated from triplicate analysis.


Isolate 1 (1 hour hold)


Control

(pH 6.5-5.5)


Acid

(pH 3-5.5)


Alkaline

(pH 11-5.5)


Isolate 2 (8 hour hold)


Control

(pH6.5-5.5)


Acid

(pH 3-5.5)


Alkaline

(pH 11-5.5)


Moisture Content 81.3 0.353 82.5 1.41 81.5 0.707 80.8 0.354 80.05 1.20 81.3 0.707
(%)
Protein Content
(%) 10.9 2.02 12.30.198 12.290.505 10.42.52 12.30.500 11.80.940


Supernatant: neutral lipids, saturated fatty
acids, and (sarcoplasmic) proteins







Sediment: Connective tissue,
membrane and neutral lipids







Protein Isolate: primarily myofibrillar
proteins, heme proteins, and other
precipitated/ aggregated proteins





Figure 4-1. Schematic of the observed centrifugation layers with direct readjustment from
extreme pH (3 and 11) and neutral pH (6.5) to pH 5.5. Centrifugation prior to
readjustment was not employed. This sediment layer was skimmed off with a plastic
spatula.














60
| pH 6.5
50
SSpH 11 a
40
abc abc abc
30 abc abc bc c

~ 20

10


initial homogenate homogenate after isolate supernatant
one hour
Isolation Treatments


Figure 4-2. Lipid oxidation (TBARS) during the production of acid (pH 3), alkaline (pH 11) and
"control" (pH 6.5) derived isolates. Homogenates were exposed to the given pH
values for one hour prior to obtaining an isolate and corresponding isolates. Each bar
in the graph is representative of TBARS (umol MDA/kg of tissue) with standard
deviations. Different small letters indicate a significant difference.


70
pH 3
60-
SpH 6.5
50 -50
4 MpH 11 b
b
40 -
abc b bc b b
b30 b bcb
be
~ 20
10



initial homogenate isolate supernatant 8 hr + 3 days
homogenate after 8 hours
Isolation Treatments

Figure 4-3. Lipid oxidation (TBARS) during the product of acid (pH 3), alkaline (pH 11) and
"control" (pH 6.5) derived isolates. Homogenates were exposed to the given pH
values for 8 hours prior to obtaining and an isolate and corresponding isolates. Each
bar in the graph is representative of TBARS (umol MDA/kg of tissue) with standard
deviations. Different small letters indicate a significant difference.




Full Text

PAGE 1

1 EFFECTS OF LOW AND HIGH pH ON LI PID OXIDATION IN SPANISH MACKEREL By HOLLY TASHA PETTY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Holly Tasha Petty

PAGE 3

3 To all with a passion for knowledge and the compassion to share it.

PAGE 4

4 ACKNOWLEDGMENTS I thank my supervising committee chair (Hordur G. Kristinsson) for the exceptional privilege of conquering a future world problem, for his mentorshi p, for his friendship, and for giving me the courage to realize my potential and dreams. I thank my supervisory committee (Steve Talcott, Marty Marshall, and Charlie Sims ) for their valuable expertise, instruction, and revolutionary contributions to th e academic field of food science. I thank the external member on my committee (Christian Leeuwenburgh) for he lping me realize my passion for applied physiology and kinesiology and for providing worldl y understanding of free radicals, aging, and life. I especially thank Wei Huo for his statisti cal expertise and generous help. Others in my graduate department who especially provided me with support include Maria Plaza, Dr. Ross Brown, Susan Hillier, Carmen Graham, Bridget Stokes, Maria Ralat, Marianne Mangone, and Walter Jones. I would like to thank a ll of my family and many influentia l friends for encouraging me to fulfill this dream. I specifically thank Andrew Piercy for his kindness and patience and Sean McCoy for his unconditional inspiration and contri bution. I thank all of my lab peers for lending me knowledge, support, tear s, and smiles. I thank Captain Denni s Voyles and his family for their friendship and the Spanish mackerel capture of a lifetime. I thank those I have not specifically acknowledged or who crossed my graduate path, fo r their positive influences and support. Lastly, I am thankful, for the true lessons I have le arned, and irreplaceable treasures I have earned.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES................................................................................................................ .......10 ABSTRACT....................................................................................................................... ............12 CHAPTER 1 INTRODUCTION..................................................................................................................14 2 LITERATURE REVEIW.......................................................................................................16 Obtaining a Fish Protein Isolate.............................................................................................16 Conventional Surimi Processing.....................................................................................16 Acid/Alkali-Aided Processing.........................................................................................17 Advantages of the Acid/Alkali-Aided process................................................................17 Lipids in Fish Muscle.......................................................................................................... ...19 Lipid Composition.............................................................................................................. ....19 Lipid Oxidation in Fish........................................................................................................ ...20 Autocatalytic Mechanism................................................................................................20 Transition Metals: Catalyst s of Lipid Oxidation.............................................................21 Heme vs. Non Heme Ir on and Lipid oxidation...............................................................21 Heme Proteins and Autoxidation............................................................................................22 Hemoglobin vs Myoglobin and Lipid Oxidation....................................................................23 Lipid Content and Lipid Oxidation.........................................................................................23 Effects of Extreme pH on Lipid Oxidation in Fish Muscle.............................................24 Lipid Oxidation and Antioxidant Implementation..........................................................25 Lipid oxidation and Proteins...........................................................................................26 Protein Oxidation.............................................................................................................. ......27 Mechanisms of Protein Oxidation..........................................................................................28 Protein Scission............................................................................................................... .......29 Protein Polymerization......................................................................................................... ..29 Protein Carbonylation.......................................................................................................... ...30 Lipid Mediated Protein Oxidation..........................................................................................31 Protein Oxidation Accumulation............................................................................................31 Protein Oxidation and Muscle Food Quality..........................................................................32 Protein oxidation a nd conformational changes in proteins.............................................33 Functional properties and protein oxidation....................................................................34 Significance................................................................................................................... .........36 3 EFFECTS OF ALKALI/ACIDIC ADJ USTMENT ON LIPID OXIDATION IN SPANISH MACKEREL HOMOGENATE...........................................................................37

PAGE 6

6 Introduction................................................................................................................... ..........37 Materials and Methods.......................................................................................................... .38 Raw Materials.................................................................................................................. .......38 Chemicals...................................................................................................................... .........39 Storage Studies and Prep aration of Homogenate...................................................................39 Antioxidant Implementation...................................................................................................40 Total Lipid Analysis........................................................................................................... ....40 Lipid Hydroperoxides........................................................................................................... ..40 Thiobarbituric Acid Reactive Substances (TBARS)..............................................................41 Analysis of Moisture and pH..................................................................................................41 Statistical Analysis........................................................................................................... .......41 Results........................................................................................................................ .............42 Changes in Spanish Mackerel Lipid Oxidation......................................................................42 Water and lipid Content........................................................................................................ ..42 Lipid Oxidation in Whole and Minced Spanish Mackerel.....................................................43 Lipid Oxidation in pH Treated Spanish Mackerel Homogenates...........................................43 Changes in homogenate lipid oxidation with storage......................................................44 Changes in lipid oxidation with pH and storage.............................................................45 Changes in lipid oxidation w ith initial quality and pH....................................................45 Changes in lipid oxidation with holding time and pH.....................................................46 Overall pH effects............................................................................................................47 Lipid Oxidation of Homogenates w ith Antioxidant Implementation.....................................47 Lipid Oxidation of Homogenates afte r Readjustment to pH 5.5 and 7..................................49 Batch and homogenate pH readjustment.........................................................................50 Time and homogenate pH readjustment..........................................................................51 Discussion..................................................................................................................... ..........51 Lipid Oxidation and Storage Stability....................................................................................51 Lipid Oxidation in Acid/Alkali Adjusted Homogenates........................................................52 Homogenate pH Readjustment...............................................................................................58 Antioxidant Implementation...................................................................................................60 EDTA and Citric Acid Implementation at Low pH................................................................64 Conclusion..................................................................................................................... .........65 4 THE EFFECTS OF THE ALKALI/ACI DIC AIDED ISOLATION PROCESS ON LIPID AND PROTEIN OXIDATION...................................................................................75 Introduction................................................................................................................... ..........75 Materials and Methods.......................................................................................................... .76 Raw Materials..................................................................................................................76 Chemicals...................................................................................................................... ..76 Storage studies and Preparation of Homogenate and Isolate..........................................77 Thiobarbituric Acid Reactive Substances (TBARS).......................................................77 Protein Quantification and Characterization...................................................................78 Rheology....................................................................................................................... ...79 Protein Oxidation.............................................................................................................79 Analysis of Moisture and pH...........................................................................................80 Statistical Analysis..........................................................................................................80

PAGE 7

7 Results........................................................................................................................ .............81 Oxidative Stability of Acid/Alk ali Derived Protein Isolates...........................................81 Gel Forming Ability of Isolates.......................................................................................84 Protein Composition of Homogenates and Recovered Isolate Fractions........................84 Discussion..................................................................................................................... ..........85 Changes in pH of Mince with Storage............................................................................85 Lipid Oxidation During Extreme pH Ad justment and Readjustment to 5.5...................86 Protein Oxidation.............................................................................................................89 Gel Formin g Ability........................................................................................................92 Textural Changes and Protein and Lipid Oxidation........................................................95 SDS page....................................................................................................................... ..97 Conclusion..................................................................................................................... .........98 5 EFFECT OF ACID PROCESSING ON LIPID OXIDATION IN A WASHED SPANISH MACKEREL MU SCLE MODEL SYSTEM.....................................................107 Introduction................................................................................................................... ........107 Materials and Methods.........................................................................................................108 Raw Materials................................................................................................................108 Chemicals......................................................................................................................108 Washed Fish..................................................................................................................108 Preparation of Spanish Mackerel Hemolysate..............................................................109 Hemoglobin Quantification...........................................................................................110 Spanish mackerel Pressed juice.....................................................................................110 Thiobarbituric Acid Reactive Substances (TBARS).....................................................110 Spanish mackerel Pressed Juice protei n Quantification and Characterization..............110 Preparation of Low Moisture Washed Spanish Mackerel Mince Oxidation Model System........................................................................................................................111 Preparation of Hemoglobin and Pres sed Juice Oxidation Model System.....................112 Oxygen Radical Absorbance Capacity (ORAC)...........................................................113 Total Antioxidant Potential...........................................................................................114 Moisture Content and pH..............................................................................................115 Statistical Analysis........................................................................................................115 Results........................................................................................................................ ...........116 One and Two Way adjustment......................................................................................117 Two and Three Component Adjustment.......................................................................118 Comparison of one way and multi-component pH adjustment in Washed System......118 Antioxidant mechanisms of non-substr ate components WFM model system..............119 ORAC Test.............................................................................................................119 TEAC Assay...........................................................................................................120 SDS Page.......................................................................................................................121 Discussion..................................................................................................................... ........121 No pressed juice treatment (con trol without pressed juice)..........................................122 Lipid oxidation in the control samples..........................................................................123 The Effect of pH Adjustment on Tissue Components...................................................125 The pH effect on hemoglobin.................................................................................125 Hemoglobin adjustment to low pH and readjustment to pH 6.8...................................127

PAGE 8

8 Pressed Juice adjustment...............................................................................................129 Lipid oxidation in the adjusted washed fish muscle......................................................131 Antioxidant Capacity as related to oxidation in washed muscle...................................135 Conclusion..................................................................................................................... .......139 6 CONCLUSION.....................................................................................................................145 LIST OF REFERENCES.............................................................................................................148

PAGE 9

9 LIST OF TABLES Table page 3-1 Initial average mince values for mois ture content, pH, and lipid content.........................67 4-1 Moisture content (%) and pr otein content (%) of isolates.................................................99

PAGE 10

10 LIST OF FIGURES Figure page 2-1 Conventional surimi processing schematic........................................................................16 2-2 Flow schematic of ac id and alkaline process.....................................................................18 3-1 Lipid oxidation in Spanish Mackerel mince......................................................................67 3-2 Development of lipid oxidation in homogenates, freshly prepared...................................68 3-3 Development of lipid oxidation in homogenates after 1 day of storage............................69 3-4 Development of lipid oxidation in homogenates after 2 days of storage..........................70 3-5 The effect of ascorbic acid and toco pherol on pH 3 and pH 6.5 homogenates.................71 3-6 The effect of EDTA and citric acid on acid homogenates.................................................72 3-7 The effect of pH readjustment in homogenates at pH 3....................................................73 3-8 The effect of pH readjustment in homogenates at pH 11..................................................74 4-1 Schematic of the obser ved centrifugation layers...............................................................99 4-2 Lipid oxidation of acid, alkaline, and control isolates after one hour..............................100 4-3 Lipid oxidation acid, al kaline, and control isol ates after 8 hours....................................100 4-4 Lipid oxidation of acid, al kaline, and neutral isolates, 1 and 8 hours and 3 days...........101 4-5 Levels of protein oxidation afte r 1 hour and 8 hrs pH, and 3 days..................................101 4-6 Storage Modulus (G) of isolate pastes derived from 1 and 8 hrs, and 3 days................102 4-7 Examples of rheograms from acid (pH 3) protein isolates..............................................102 4-8 Examples of rheograms from al kaline (pH 11) protein isolates......................................103 4-9 Examples of rheograms from c ontrol (pH 6.5) protein isolates.......................................103 4-10 Protein composition of isolat es from adjusting for 1 hour..............................................104 4-11 Protein composition of isolat es from adjusting for 8 hour..............................................104 4-12 Protein composition of isolates, 8 hours of holding followed by 3 days.........................105 4-13 Protein composition of initial homogenates....................................................................105

PAGE 11

11 4-14 Protein composition of ho mogenates held for one hour.................................................106 4-15 Protein composition of ho mogenates held for 8 hours...................................................106 5-1 The role of acidified (pH 3) and pH6.8 tissue components on lipid oxidation................141 5-2 The role of two and three acidified (pH 3) and pH (6.8) tissue components...................142 5-3 Antioxidant capacity of Spanish m ackerel hemoglobin and pressed juice......................143 5-4 Protein composition of untreated Spanish mackerel pressed juice..................................144

PAGE 12

12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECT OF LOW AND HI GH PH ON LIPID OXIDATION IN SPANISH MACKEREL By Holly Tasha Petty May 2007 Chair: Hordur Kristinsson Major Department: Food Science and Human Nutrition Because of over fishing, traditional sources ca nnot satisfy the demand for fish proteins. The production of protein isolate from underutili zed fish sources using a new high and low pH process is being investigated. Underutilized fish sources include pelagic species, by-catch, and fish byproducts. The objective of this study was to investigate how the acid /alkali-aided process influences lipid oxidation developm ent in homogenates, isolates, and washed fish prepared from Spanish mackerel muscle. Studies of homogenates at acid and alkaline pH showed that lipid oxidation increased with age of raw material, mince cold storage, pH exposure time, and acidifi cation. Lipid oxidation in the acid homogenates was eff ectively inhibited by ascorbic acid and tocopherol, but was unaffected by the addition of chel ators. The pH readjustment from acid to neutral pH did not significantly affect lipid oxidati on in Spanish mackerel homogena tes; however, the addition of antioxidants in combination with pH readjustment was problematic. Isolates prepared from acid and alkaline treatment had comparable levels of lipid oxidation, regardless of homogena te holding time. High levels of lipid oxidation were observed in the supernatant fractions from acid, alkaline, a nd control treatments. Overall, alkaline isolates contained greater levels of prot ein oxidation compared to acid and alkaline isolates. Protein

PAGE 13

13 degradation occurred during acid and alkali processing of Spanis h mackerel. Rheological testing showed that there was greater ge l strength in alkaline isolates, followed by the control and acid isolates. Protein oxidation coul d explain differences in gel qu ality between acid and alkaline isolates. The adjustment from pH 3 to pH 6.8 greatly affected all components in a hemoglobinmediated lipid oxidation washed Spanish macker el model system. The addition of pH treated hemoglobin to the washed Spanish mackerel system resulted in lipid oxidation levels comparable to the control. Increases in lipid oxidation compared to the control were observed when pressed juice was pH treated. The greatest levels of lipid oxidation in the Spanish mackerel model system occurred when washed fish muscle was pH treated. These results show that lipid oxidation incurred during acidic processing could be primarily due to phospholipid changes, secondary to changes in hemogl obin and/or aqueous extracts.

PAGE 14

14 CHAPTER 1 INTRODUCTION Because of the large demand for fish protein an d inability to satisfy the need by traditional means, the utilization of fatty pelagic species and by-catch is under current study. By-catch from shrimp vessels, pelagic species, and by-pr oducts are underutilized as human food ( 1 ). By-catch includes low value fish that are caught and discarded back into the sea or that are further used for fish meal ( 2 ). It is estimated that four pounds of by-catch result for every pound of shrimp caught. In addition, dark muscle fish species compose 40-50% of to tal catch worldwide ( 3, 4 ) Pelagic species present numerous obstacles if pr ocessed conventionally, due to small size, shape, bone structure, seasonal availabi lity, and most importantly the pr evalence of unstable lipids and pro-oxidants ( 4-6 ). Extracting functional and high quality fish prot eins from fish sources by solubilizing the proteins at low and high pH va lues and recovering solubilize d proteins with isoelectric precipitation has been conducted ( 7 ). From these findings, isolati on of functional proteins from underutilized, low value fish sources was developed ( 7, 8 ). These newly developed processes produce a protein isolate low in lipids. This is of major importance, due to the abundance of triacylglycerols (neutral storage lipids) and membrane phospho lipids in mitochondria of the dark fish muscle ( 5 ). Phospholipids are the prim ary substrates in lipid oxi dation due to their highly unsaturated nature ( 9, 10 ). Protein isolates made from white muscle using the alkali process have been found to possess greater oxidative and color stability over unprocessed or acid treated muscle ( 11 ). Studies have shown that high pH inactivates he moglobin, thereby preventi ng oxidation of lipids and formation of an undesired yellow-brown color as well as off odors and flavors. In contrast, acid treatment denatures hemoglobin, e nhancing its pro-oxidative activity ( 12 ). Pelagic fish

PAGE 15

15 species and fish byproducts are ri ch in blood and heme proteins. The blood and heme proteins are used for oxygen transport for production of long-term energy in the dark muscle via oxidative metabolism. In white muscle, heme pr oteins are used for more short-term energy ( 5 ). A number of other changes in homogenized fish muscle can occur at low and high pH values that can contribute to the acceleration or delay of lipid oxidation. Currently there is a limited understanding of how low and high pH infl uences oxidative processes in fish muscle. Further investigation is therefor e needed regarding the stability of fish muscle as affected by extreme pH. Due to the abundance of heme protei ns in pelagic fish species, it is imperative to study their oxidative stability and to control their prooxidative effects. These proposed studies could further the implementation of pelagic fish protein isolates as a food source and functional food ingredient, and an understanding of oxidation mechanisms in general.

PAGE 16

16 CHAPTER 2 LITERATURE REVEIW Obtaining a Fish Protein Isolate Conventional Surimi Processing The production of surimi has proved to be some what useful in utilizing fish proteins. Briefly, fish muscle is ground and washed at approximately 1 part fish to 2-5 parts water. The washing step can be repeated several times depending upon processing conditions (i.e. type of fish and plant location). During the washing ste p, 30% of the total proteins in the form of sarcoplasmic proteins are lost. The fish muscle is then dewatered using a screw press. Normally, cryoprotectants are added to prevent protein denaturation during freezi ng and thawing. This process (Figure 2-1). Unfortunate ly, high quality surimi from conventional processing has only resulted from white flesh fish. Attempts to create surimi from dark pelagi c species has resulted in products having poor color and lipid stability due to pigments, pro-oxidants, and an abundance of lipids ( 4, 6 ). Figure 2-1. Conventional surimi proc essing schematic. The final surimi is heated to form a gel, resulting in various s eafood analog products (13). Ground Fish Muscle Refine and Dewater Stabilize ( Add cr y o p rotectants and freeze ) Surimi Wash with 25 parts water ( 3 times with 0.3% NaCl in last ste p)

PAGE 17

17 Acid/Alkali-Aided Processing Solubilization and recovery of protein by pH rather than salt adjustment, led to an economical protein isolate from low value fish sources ( 7 ). At high and low pH values, the proteins become solubilized due to large net charges. It was also discovered that at extreme pH, the cellular membranes surrounding the m yofibrillar proteins are disrupted ( 13, 14 ). As a consequence of these changes in protein solubili ty and membrane integrity, the viscosity of a muscle protein homogenate at very low and high pH values was found to decrease significantly. This viscosity decrease enables so luble proteins to be separated from undesirable materials in the fish muscle via high speed centr ifugation (e.g.bones, scales, conn ective tissue, microorganisms, membrane lipids, neutral lipids etc.). The soluble proteins can then be recovered and precipitated by adjusting the pH in to the range of thei r isoelectric point (~pH 5.3 to 5.6) (Figure 2-2). This process uses a very low (2.5 to 3.0) or high pH (10.8 to 11.5), thus enabling for a simple and economical recovery of low value fish species ( 7, 8, 15 ). Studies have demonstrated that of the two processes, the alkali-aided process is substantially better than the acid-aided process since oxidation and color problems are re duced and protein functionality is superior ( 13, 16, 17 ). Advantages of the Acid/Alkali-Aided process Using the acid/alkali-aided process versus conve ntional processing is beneficial for various reasons. Protein recoveries can range from 90-95% using the acid/alkali-a ided process versus conventional surimi recoveries ranging from 5565% from fish fillets. Conventional processing only retains the myofibrillar proteins, in contrast this new process allows for the recovery of both myofibrillar and sarcoplasmic proteins Processing is more economical using the acid/alkali-aided process, re ducing labor, expenses, producti on complexity, and handling

PAGE 18

18 (homogenate compatibility in processing). Addi tionally, minimally processed and lower quality, less utilized Figure 2-2. Acid and alkaline process used to pr oduce protein isolate from lowand high-value fish protein sources. sources of fish protein can be implemented. Be fore processing, fish sources can be of lower quality and can be previously fr ozen. Lipids can be less of pr oblem due to the separation of triacylglycerols and phospsholipid s from the protein homogenate, enabling the production of a more stable product with better co lor and flavor; however, there is not absolute separation of heme proteins from the final protein isolate. Due to the abundance of mitochondria in dark muscle, the acid/alkali process is advantageous in the separation of phospholipids from dark muscle ( 5 ). Functional properties incl uding gelation, water holding cap acity, and solubility are all improved during the acid/alkal i-aided process due to a desira ble partial denaturation of the Ground Fish Material Homogenization 1 p arts fish : 5-9 p arts wate r pH reduction or increase p H 2.5 ( HCl ) / p H 11 ( NaOH ) Sediment Layer M embrane li p ids Centrifuge (10,000 xg) Solution Phase S oluble muscle Upper Layer Protein aggregation p H ad j usted to 5.5 (a) Sediment = Protein isolate Centrifuge (10,000 x g) Supernatant Mostl y water can be

PAGE 19

19 fish proteins. Overall, the prot ein isolate produced is safer due to elimination or reduction of toxins (PCBs and mercury) and microbial loads, and no pollution due to processing ( 6, 13, 16, 18, 19 ). Lipids in Fish Muscle Lipid Composition For dark muscle fish species, fat compos ition can vary in response to environment, seasonal variation, migratory behavi or, sexual maturation, and feed ( 20 ). Fish living in colder water will have more dark muscle. Dark muscle tends to vary in composition, especially with season and location; therefore, when sampling dark muscle or fatty fish species, it is important to consider the possible varia tions in lipid content ( 20 ). The two types of lipids contai ned within teleost (bony) fish muscle are phospholipids and triacyglycerides. Phospholipids or structural li pids are the major components of cell membranes within the fish tissue (myofibrillar protein); wh ile triaglycerides are found in adipocytes and serve as storage energy. Phospholipids are found in various membranes including those of the cell and organelles within the cell. In general, muscle from lean fish is composed of approximately 1% lipid, primarily phospholipid. The majority of lipids in lean fish are stored in the liver. Some lean species include the bottom-dwellers: cod, saithe, and hake. Major lipid deposits in fatty fish species are found in tissue other than the liver: subcutaneous tissue, in the belly flap, and throughout the muscle tissue. Fatty spec ies include the pelagic species such as herring, mackerel, and sprat. Most importantly, lipids are wide ly distributed within muscle tissue, primarily next to the connective tissue (myocommata) and be tween light and dark muscle ( 21 ) In pelagic species, the dark muscle uses triacylglycerides for conti nuous aerobic energy in migrations, whereas white

PAGE 20

20 muscle uses glycogen as the main source of ener gy and is quickly fatigable due to accumulation of lactic acid during an aerobic respiration ( 22 ). Lipid Oxidation in Fish Lipid oxidation serves as a mediator of importa nt processes in living biological systems ( 23 ). However, post-mortem lipid oxidation can onl y be controlled to a certain extent, before antioxidants and their systems are exhausted ( 24 ). Lipid oxidation is one of the primary causes of deterioration in cold stored fish muscle ( 25 ) and negatively affects color ( 26 ), odor and flavor ( 27 ), protein functionality and conformation ( 28 ), and overall nutritional c ontent of fish muscle ( 29, 30 ). This quality deterioration is due to th e high content of polyunsaturated fatty acids contained within fish muscle ( 9, 10, 25 ) along with highly act ive pro-oxidants ( 5 ). Major reactants in lipid oxidation are oxyge n and unsaturated fatty acids. Exposure of lipids to atmospheric oxygen results in unstable intermediates ( 31 ). Ultimately, these intermediates then degrade into unfavorable compounds attributi ng to off flavor and aroma. Lipid oxidation has many plausible mechanism of initiation, including non-enzymatic and enzymatic reactions. Non-enzymatic mechanisms include autoxidation (free-radical mechanism) and photogenic oxidation (si nglet oxygen mediated) ( 31 ). Enzymatic mechanisms include actions by lipoxygenase and cyl ooxygenase. Lipid oxidation most commonly occurs by a free radical mechanism, involving the forma tion of a reactive peroxyl radical ( 31 ). In addition, it has been reported that the detrimental effects due to heme (autoxidation mechanisms) are greater than effects due to lipoxygenase, mainly due to longer-lasting pr o-oxidative activity ( 32 ). Autocatalytic Mechanism The autocatalytic mechanism is deemed th e primary cause of lipid oxidation in post mortem fish ( 31 ) and is historically referre d to as lipid peroxidation ( 33 ). The free radical mechanism has three classic stag es: initiation, propagation, and te rmination. Initiation begins

PAGE 21

21 with hydrogen abstraction from an unsaturated fatty acid, particularly those having a pentadiene structure ( 31 ). The lipid radical (L ) then reacts with molecula r oxygen and a peroxyl radical results (LOO ). Propogation is the phase in which the peroxyl radical (LOO ) abstracts hydrogen from neighboring lipid molecules, ther eby forming more lipid radicals and lipid hydroperoxide (LOOH) molecules. Lipid hydroperoxides cannot be detected through sensory analysis, which is the reason why peroxides do no t correlate with off flavor. The degradation of these lipid hydroperoxides, due to metal catalysts, results in the unfavorable secondary intermediates of shorter chain length: ketones, aldehydes, alc ohols, and small alkanes. The secondary products give rise to off aromas, flavors, and yellow color in fish. Aldehydes are normally associated with off odor. Termination oc curs when a radical reaction is quenched by a reaction with another radical or antioxidant. It is important to note that these reactions are occurring simultaneously and that the mechanism of lipid oxidations can become quite complex. ( 22, 31 ). Transition Metals: Cataly sts of Lipid Oxidation Decomposition of lipid hydroperoxides at refrig erator and freezing temp eratures is due to electron transfer from metal ions Transition metals serve as pr imary oxidation catalysts in lipid oxidation, mainly iron. Iron content va ries among muscle food sources ( 31 ). The main catalysts in biological systems include heme and non-heme iron ( 5 ). Non-heme sources include released iron from heme proteins, protein complexes (tra nsferritin, ferritin, and metalloprotiens), and other low molecular weight complexes (diphosph onucleotides, triphosphonuc leotide, or citrate chelates) ( 34 ). Heme vs. Non Heme Iron and Lipid oxidation White muscle contains less iron than dark muscle, due to less myoglobin and hemoglobin ( 5 ). Through research, it has been shown that heme proteins serve as potent pro-oxidants ( 7, 35 ).

PAGE 22

22 Further more, it is believed that heme proteins serves as the main prooxidant in post mortem fish muscle ( 36, 37 ). It is believed that hemoglobin is the main initiator of oxidation in vitro systems ( 12, 38 ); whereas, hydrogen peroxide mediat ed metmyoglobin has been shown to initiate oxidation in vitro-model systems ( 39, 40 ). In addition, Kanner et al ( 41 ) found that lowmolecular weight iron also initiates lipid oxi dation. Nonetheless, studies have shown that hemoglobin is the leading pro-oxidant ( 16, 35, 42, 43 ). The basic mechanism of lipid oxidation is known; however, the predomin ant initiators(s) of oxida tion are still debated. Heme Proteins and Autoxidation If most of the hemoglobin is not removed through bleeding, the fish muscle will become more susceptible to oxidation. Mincing or mech anical mixing of fish muscle can promote oxidation through the incorpora tion of hemoglobin, hemoglobin subunits, free heme, and iron. ( 38 ). The presence of hemoglobin and/or myoglobin during the acid and alka li-aided process and in the end product (protein isolate) c ould lead to serious lipid oxidation ( 42, 44, 45 ). Whether oxidized or reduced, heme proteins can serve as potent pro-oxidants ( 46 ). Post mortem, heme proteins are in their ferrous state (Fe+2). In this state, oxygen can be bound or unbound to the heme thus giving oxyhemoglobi n/myoglobin and deoxyhemoglobin/myoglobin, respectively. The heme iron can become oxidized to the ferric state (Fe+3) and result in metmyoglobin/hemoglobin, with water binding to the 6th coordination site ( 47 ). This oxidation will also give rise to the supe roxide anion. Subsequent superoxi de dismutation can result in the highly reactive hydroperoxyl radical, hydroxyl radical, and hydrogen peroxide ( 6, 24 ) Autoxidation of hemoglobin or myoglobin by su peroxide can lead to the formation of the hypervalent ferryl-hemoglobin/myoglobin radical. Th e formation of this porphyrin cation radical is deemed the primary decomposer of lipid hydroperoxides, result ing in the secondary compounds, associated with off-flavors and odors ( 37, 48 ).

PAGE 23

23 Hemoglobin vs Myoglobin and Lipid Oxidation Myoglobin is located in the muscle cell itsel f while hemoglobin is found in the blood cells of the muscle tissue vascular system ( 49 ). Hemoglobins from va rious fish species are differentiated from how th ey react to changes in temperature, pH, etc ( 50 ). It has been found that with decreases below pH 7.6 ( 43 ) and as low as pH 1.5 ( 51 ), either through physiological changes post mortem and/or experimental mani pulation, a shift from oxy to deoxy-hemoglobin occurs, called the Bohr shift. The change to th e deoxy state is believed to increase hemoglobins pro-oxidative activity due to heme iron exposure ( 52 ). Physiologically, this phenomenon may be advantageous in the release of oxygen in the produ ction of pyruvate and consequently lactate, in pelagic species due low oxygen affinity ( 53 ). Recently, the comparison of myoglobin and hem oglobin from trout as catalyst for lipid oxidation was presented by Richards and Dettmann( 54 ) The researches found that hemoglobin has a greater tendency to indu ce lipid oxidation due to the mechanism of heme dissociation hversus myoglobin at neutral pH ( 55 ). The oxidative nature of myoglobin versus hemoglobin at low pH is being investigated. Lipid Content and Lipid Oxidation According to Richards and Hultin ( 56 ), only low levels of phospholipid (0.01%) are required for hemoglobin mediated lipid oxidation and for off flavors to develop. According to Undeland et al. ( 57 ), it is the presence a nd amount of hemoglobin that affects the extent of oxidation and not the amount of lipid. In a study conducted by Richards and Hultin ( 58 ), even a six-fold increase in membrane phospholipids did not affect the extent of lipid oxidation. Despite bleeding, hemoglobin still remains within the fish muscle and can still pose problems ( 59 ).

PAGE 24

24 Effects of Extreme pH on Lipi d Oxidation in Fish Muscle At extreme acidic pH, heme proteins are dena tured, increasing the propensity of lipid-prooxidant interactions. Lipid membranes become more susceptible to oxidation when in the presence of heme proteins adjusted to low pH or readjusted from low pH (3.0) to neutral pH (7) ( 18 ). This is due to unfolding of the heme protein and exposure of the heme iron upon denaturation, or lack of refolding with pH readjustment ( 51 ). In addition, transition metals and free heme become active at low pH ( 56 ). Autoxidation and ferryl-Hb/Mb production are enhanced at low pH ( 56, 60 ). Furthermore, the heme crevice of hemoglobin is unfolded at low pH, increasing reactions with lipid substrates, and thus oxidation ( 12 ). With acidic pH, there is also a greater likelihood of subun it dissociation of the hemoglobin and thus loss of the heme ( 51, 61 ). In contrast, the pro-oxidative activity and au toxidation of hemoglobin is reduced at alkaline pH due to increased stabili zation of the heme group ( 51, 60 ). Studies of the effect of extreme pH on he moglobin and oxidation have been conducted in model systems by Kristinsson and Hultin ( 51 ). Trout hemoglobin was added to a washed cod model system. This model muscle system c ontain virtually all me mbrane phospholipids and proteins, but is free of soluble proand antio xidants through the washi ng step. The ability of hemoglobin to oxidize cod at pH values 2.5, 3.5, 10.5 and 11 was studied. It was found that there was considerably more oxidation at low pH values compared to higher pH values. Undeland et al. ( 62 ) studied the minimization of oxida tion of herring muscle homogenate at low pH (pH 2.7) by reducing exposure ti mes, using high speed centrifugation, and implementing antioxidants or chelators. Storage stability of the protein isolates was only improved by antioxidant and chelator implementation, erythrobate and (ethylenediaminetetraacetic acid) EDTA, respectively ( 62 ).

PAGE 25

25 Various techniques to reduce lipid oxidation exist and in clude heme separation, lipid membrane centrifugation, pH exposure time, an d antioxidant implementation. However, the quantification of lipid oxidation in the homogenate made from fish of varying quality in certain conditions requires attention such as extended holding times ( 12 hours) and an expanded pH range (2.5 to 11.5). Lipid Oxidation and Anti oxidant Implementation The greatest defense against lipid oxidation is the use of antioxidants. Inhibition of lipid oxidation can occur directly or i ndirectly. Inhibitors can preven t initiation and propagation steps in the free radical mechanism ( 31 ). Antioxidants occur exoand endogenously. Antioxidants donate hydrogen to the peroxyl radical and elimin ate it as a reactive subs tance in the radical mechanism ( 31 ). Past research evaluated the effects of antioxi dants on lipid oxidation in mackerel fillets ( 63 ). Exogenous antioxidants includ ed: ascorbic acid, tocopherol, and butylated-hydroxytoluene (BHT). The antioxidants were applied separate ly and in combination. At frozen storage temperatures (-20C and C), the BHT and BHT co mbined with ascorbic acid proved to be the most inhibitive. This implies that there is a synergistic effect between lipid and water-soluble antioxidants in delaying lipid oxid ation in whole fish muscle ( 63 ). Concentration effects of antioxidants are we ll known in vivo. At low ascorbic acid concentrations, there is a reduction of transi tion metals (iron), pr omoting hydroxyl radical formation, and at high concentration, ascorbic acid scavenges hydroxyl and superoxide radicals in vivo ( 49 ). According to Witting ( 64 ) and Chow ( 65 ), tocopherol serves as a pro-oxidant at high concentration and an antioxi dant at low concentrations. At low pH versus high pH, ascorbic acid and tocopherol can i nhibit lipid oxidation. Together, ascorbic acid, in high concentrations, and tocopherol e ffectively prevent oxidation in

PAGE 26

26 vitro. Tocopherol by itself is especially e ffective in reducing li pid oxidation in vivo ( 66 ). In herring fish homogenate (pH 3), there was a sign ificant reduction in oxid ation as measured by malondialdehyde when implementing ascorbic acid and tocopherol ( 62 ). In addition, when incorporating antioxidants erythr obate (iso-ascorbic acid) and STPP (sodium tripolyphosphate), there was a significant reducti on in lipid oxidation in th e final protein isolate ( 67 ). According to previous studies, it is essentia l that antioxidants be added early in the acid aided process to obtain a good lipid stable final product ( 67-69 ). According to Undeland et al. ( 67 ), low molecular weight aq ueous extracts, also know as press juice, from fish muscle are effective in inhibiting lipid oxidation in cod lipid membranes caused by hemoglobin. Possible inhibitors in the press juice could include nucleotides and other reducing agents. The role of these comp ounds at low or high pH is not well understood. Lipid oxidation and Proteins The functionality of fish protein isolate prep arations relate mostly to the behavior and interaction between myofibrillar proteins (mos t notably actin and myosin). Sarcoplasmic and myofibrillar proteins contribute to varying portio ns of the overall protein isolate, depending on whether the alkaline or acidic process is used. Myofibrillar proteins are the primary components of protein isolates produced from the alkaline process, and sarc oplasmic proteins are the primary proteins in isolates prepar ed from acid processing ( 14 ). The effect of lipid oxidation and reactive oxyg en species (ROS), on protein stability and function in muscle foods is poor ly understood. Howell and Saeed ( 70 ) reported texture changes in frozen Atlantic mackerel ( Scomber scombrus) stored at 20 and 30C for up to 2 years, which could be related to an increase in lipid oxidati on. Increases in G (ela stic) and G (viscous) moduli were observed. With the implementation of antioxidants in the muscle, a correlation between the kinetics of lipid oxidation and rheological behavior was observed. In another study

PAGE 27

27 by Saeed and Howell ( 63 ), it was reported that lipid oxidation with froz en storage caused a reduction in ATPase activity, solubility, and my osin heavy chains. Th e incorporation of antioxidants in Atlantic mackerel prevented/ delayed the deleterious effects caused by lipid oxidation. Hitherto, studies have not addressed the effects of acid and alkali processing on protein oxidation development. Protein Oxidation The relationships between muscle food quality and protein oxidation ar e of great interest and have yet to be fully understood. Consequen ce of oxidative stress, pr oteins undergo covalent modification and incur structural changes. Repor ted protein modifications include and are not limited to back bone disruption, cross-linking, unfolding, side-chain oxidation, alteration in hydrophilic/hydrophobic character and conformati on, altered response to proteases, and introduction of newly formed reactiv e groups, or amino acid conversion ( 71-74 ). Although protein functionality and quality changes are commonly observed with increases in lipid oxidation, the actual mechanisms by which proteins are oxidized in muscle foods are not well understood.. Extensive research shows a positive correlation between pr otein oxidation and diverse physiologically deteriorative even ts, such as aging and disease. Biological events are highly dependant upon radicals a nd oxidative processes ( 75 ). Oxidative stress arises when antioxidant mechanisms are unable to counteract oxidant loads ( 76 ). Oxidative stress is secondary to partial and/or complete exhaustion or degr adation of antioxidant systems ( 24 ). In the event of oxidative stress, generated radicals and non -radical species can oxidize proteins directly or indirectly. Due to oxidatively reliant metabolism in vivo, direct pr otein oxidative attack occurs most frequently in the presence of reactive oxygen species (ROS): hydrogen peroxide, superoxide, nitric oxide, peroxynitrite, hydroxyl radical, peroxyl and/or alkoxyl radicals, etc. ( 77 ). Similarly, these ROS

PAGE 28

28 are produced in muscle foods during mechanical processing and storage ( 78-80 ). Indirect protein oxidation can occur in the pres ence of secondary lipid and protein oxidation products ( 81-83 ). Oxidizing agents common to living and post mortem muscle could include and are not limited to transition metals, protein-bound transition metals (Heme), U.V., X, irradiation, ozone, lipid and protein oxidation products, and oxidative enzymes ( 84, 85 ). Neutrophils, macrophages, myloperoxidase (hypochlorous acid) and oxireductase enzymes are just a few possible mediators specific to in vivo protein oxidation ( 84, 85 ). Rational for oxidative mechanisms post mortem are extrapolated fr om those which occur physiologically. The introduction of oxygen during muscle food proce ssing is analogous to reperfusion in vivo ( 86 ). Pro-oxidative environments present themselves in the aging and restructuring of meat products. With storage, endogenous antioxidant levels decrease in whole muscle; whereas, an increased exposure to oxygen and endogenous oxidants as well as decreasing endogenous antioxidants occur in processed muscle. Thus, radical species and derivatives induce greater levels of oxidative stress with d ecreases in antioxidant capacity ( 87, 88 ). Mechanisms of Protein Oxidation A multitude of mechanisms give rise to prot ein oxidation. Oxidatively susceptible protein sites in the presence of ir radiation and/or metal genera ted ROS have been studied ( 89-92 ). Structural protein units most prone to attack include the -carbon and side chains of amino acid residues as well as the peptide backbone. Amino acid residues most susceptible to oxidation in vivo and in vitro are cysteine a nd methionine due to the presence of a readily oxidizable sulfur atom and vulnerability to all oxidizing sources ( 76 ). Oxidation of cysteine can lead to disulfides (S-S), glutathiolation, and thyl radicals ( 93, 94 ), while the oxidation of methionine predominately leads to methionine sulfoxide ( 95 ). Oxidative modifications do not always result

PAGE 29

29 in deleterious effects. In fact negligible effects or antioxida tive capabilities may arise with protein oxidation ( 96-98 ). Reactive oxygen species (ROS) can directly oxidize protei ns by abstracting a hydrogen atom from the carbon atom of an amino acid, produc ing a carbon centered radical. Further reaction of these radicals with oxygen gives a peroxyl radical. Per oxyl radicals can then abstract hydrogens or react with hydroperoxyl radicals (p rotonated superoxide), forming the alkoxyl radical. A hydroxyl derivative can be produced fr om the reaction of the hydroperoxyl radical and the alkoxyl radical ( 89-92 ). In conjunction with protein radical formati on, oxidation of proteins, peptides, and amino acids can give rise to various protein derivatives. Pr otein oxidative derivatives common to both physiological and post mortem tissue result fr om scission, polymeriza tion, carbonylation, and protein-lipid interact ion. Specific oxidation modifications more common in vivo include hydroxylation, nitrosylation, nitration, sulfoxidati on, and/or chlorination of protein residues ( 99 ). Protein Scission Scission of peptide backbones and/or amino residues can occur with protein oxidation. Alkoxyl radicals and alkoxyl pero xides can be cleaved through two major mechanisms: the amidation pathway and the diamide pathway. Carbon radicals or derivatives are cleaved to form amide and -keto-acyl derivatives in the amidation pathway; while diamide and isocyante derivatives form in the diamide cleavage pathway ( 89 ). According to Garrison ( 89 ) and Uchida et al. ( 100 ) hydroxyl radical mediated cleavage/fra gmentation of individual residues such as glutamate and aspartate ( 89 ) and proline ( 100 ) can occur. Protein Polymerization Along with peptide cleavage, protein and/or amino radicals and derivatives can undergo polymerization, forming cro ss-linked protein residues ( 82-85) Protein cross-linkage can occur

PAGE 30

30 through covalent and non-covalent interactions. Non-covalent pr otein cross-linking is due to hydrophilic and hydrophobic bonding. Oxid ation of proteins can resu lt in greater exposure of hydrophobic groups, and thus, greater hydrophobic interactions among re sidues. Increased interaction of hydrophobic residue s can lead to aggregation. Likewise, hydrogen bonding can facilitate protein and pr otein-lipid aggregation ( 101 ). Covalent intraand inter-protein polymerization are due to radical mediated m echanisms. ROS induced protein cross-linking in vivo and in muscle foods can occur under seve ral primary conditions: (1) homolysis of two carbon, protein, or tyrosine radicals ( 89 ), (2) thio-disulfide exch ange of oxidized cysteine sulfhydryl groups ( 89, 102 ), (3) interaction of oxidatively de rived protein carbonyl groups with lysine/arginine residues ( 87, 103 ), (4) cross-linking of di aldehydes (malondialdehyde or dehydroascorbate) with two lysine residues ( 104, 105 ) Protein Carbonylation Metal catalyzed oxidation (MCO) is a primary mechanism by which proteins are directly oxidized in vivo to form carbonyl derivatives. Likewise, considerab le protein oxidation is due to MCO in muscle foods ( 106 ). In the vicinity of proteins a nd or amino acid residues, protein bound transition metals illicit the production of RO S and oxidation of amino acid side-chains. Hydrogen peroxide is reduced to ROS species (hydroxyl radical and fe rryl radical) in the presence of transition metals ( 107-110 ). MCO can lead to the formation of 2-oxo-histidine from histidine and carbonyl derivatives from lysi ne, arginine, proline, and threonine ( 107, 111 ). Carbonyl derivatives are frequen tly investigated in physiologi cal and post mortem systems, due to convenience and elucidation of overall protein oxidation. Car bonyl groups (aldehydes, ketones) are indicative of prot ein oxidization induced by various sources of ROS and oxidative stress. Carbonyls can be detected in various wa ys and the assay involves the derivitization of carbonyl groups with 2,4-dinitropheylhyrazin e (DNP), producing a 2,4-dinitrophynl (DNP)

PAGE 31

31 hydrazone ( 112 ). Following covalent derivatization, va rious quantitative techniques can be utilized including and not limited to 2-D pa ge, spectrophotometric methods, and high performance liquid chromatography (HPLC) ( 112 ). Additionally, carbonyl measurement is applicable to mixed samples: homogenate tissue, plasma, and proteins. Other methods require purification, but allow for differentiati on of the oxidative source (Shacter, 2000). Lipid Mediated Protein Oxidation Lipid mediated protein oxidation is be lieved to occur by various mechanisms ( 113 ). One proposed mechanism describes a protein or am ino residue radical a nd lipid oxidation product (malondialdhyde) condensation reaction, forming a lipid-protein complex. Proteins could react with lipid oxidation pr oducts (alkyl radicals, hyderoperoxi des, and secondary oxidation products) giving rise to a prot ein-lipid radicals ( 63 ). Another mechanism for lipid-protein oxidation is described: (1) free radicals react with proteins side chains pr oducing protein free radicals, (2) protein radicals could then react with oxyge n and produce a peroxy radical, and (3) hydrogen peroxide radicals formation could thus follow and indu ce formation of carbonyl products. The extent of protein oxidation coul d theoretically develop faster than lipid oxidation, due to the cytosolic associated location of proteins in musc le foods and possible exposure to free radicals (Srinivasan and Hultin, 1995). Specific amino acids ar e highly susceptible to lipid induced oxidation due to the presence of reactive si de chains: sulfhydryl, thioether, amino group, imidazole ring, and indole ring. Corresponding ami no acids include cysteine, methionine, lysine, arginine, histidine, and trypotphan ( 101, 114 ). Protein Oxidation Accumulation Physiologically, oxidatively modified protei ns can accrue over time. Reduced activities and/or levels of specific prot eases, low rates of cell turnover, depletion of antioxidant mechanisms, or inherently undetected oxidative modifications can lead to protein oxidation

PAGE 32

32 accumulation in vivo ( 115 ). In instances where an amino acid conversion occurs, the cell is unable to discriminate between oxidized a nd un-oxidized residues. MCO can lead to modifications that go unnoticed including the co nversion of proline to hydroxyproline, or the conversion of glutamate to asparagine or aspartate ( 76 ). The cell predominantly relies upon the protease known as the proteosome, to de grade oxidatively modified proteins ( 116 ). Methionine sulfoxide reductase is one of the few repair ing proteases, enabling for the conversion of methionine sulfoxide to me thionine (MET) in vivo ( 117 ). A decrease in proteosome activity occurs with aging and is believed to be pr edominantly responsible for the accumulation of oxidized proteins ( 106, 118, 119 ). Accumulation of protein oxid ation products increases the likelihood of cross-linking. Furthermore, polymeriz ed proteins can not be degraded by the proteosome and can prevent degradat ion of other oxidi zed proteins ( 73 ). Unlike living systems, reparation measures for protein oxidation in post mortem tissue do not exist. Thus an understanding of protein oxidation mechanisms in muscle foods and how functionality is affected by protein oxida tion is well warranted. Protein Oxidation and Muscle Food Quality Protein oxidation can induce phys iochemical changes in proteins. Oxidative modifications can lead to decreases and or loss in protein ac tivity and stability, negati vely impacting cellular processes in vivo and f unctionality in vitro ( 120 ). In post mortem muscle foods, the intrinsic property primarily affected by oxidative modifica tion is conformation and/ or structure. The functional property in muscle f oods most affected by this in trinsic change is hydration. Alteration in hydration capacity can ultimately affect other functionally properties: gelation, emulsification, viscosity, solubility, di gestibility, and water holding capacity ( 121-125 ). Due to the significant contributions to muscle functiona lity, myosin is extensiv ely studied in protein oxidation studies ( 86 ). Numerous muscle food st udies conclude that cha nges in protein structure

PAGE 33

33 and functionality, under oxidizing conditions, are due to protein oxidation. Understanding the cause of structural or functional protein changes as they relate to protein oxidation is important to muscle food quality. Protein oxidation and conformational changes in proteins Changes in conformational stability in muscle fo od proteins is attributed to changes in free energy. As a result of oxidative modification, intraand intermolecular interactions are altered within a proteins structure; thereby, decreasing the thermal en ergy and conformational stability. Liu and Xiong ( 126 ) studied thermal stability changes in the Fe (II)/ascorbate/H202 induced oxidation of myosin, derived from chicken p ectoralis muscle. Using differential scanning calorimetry (DSC), the protein oxidation by hydroxyl radicals and effects on myosin conformational stability were measured. Significa nt reduction in myosin thermal stability was observed as indicated by decreases in enthalpy ( H). Additionally, decreases in transition temperature (Tm) were observed ( 126 ). The un-oxidized myosin transition temperature (55C) designates the myosin conformational change to a randomly coiled, broken net, structure upon heating. Thus, decreases in Tm are indicative of decreased conformation stability ( 127 ). Relative to the native myosin, a new transition (6080C) temperature was observed among oxidized samples upon heating. This newly formed heati ng transition was attri buted to cross-linking among oxidized cysteine residues as the Tm was not observed among the oxidized samples containing N-ethyl-maleimide a thiol blocking reagent( 126 ). According to Srinivasan and Xiong ( 128 ), decreases in conformation stability were obse rved in bovine cardia c muscle myofibrillar proteins when subjected to hydroxyl radical attack ( 128 ). Similarly, Saeed and Howell ( 70 ) observed decreases in enthalpy ( H) and Tm in Atlantic mackerel fillet s stored at -20C versus those stored at -30C for 2 years. These change s were indicative of aggregation and denaturation within the mackerel fillets ( 70 ). In another study conduc ted by Saeed and Howell ( 63 ),

PAGE 34

34 denaturation of myofibrillar proteins and lipid oxidati on of docosahexaenoic and eicosapenaenoic acid were observed in frozen storage of Atlantic Mackerel and were significantly inhibited by incor porations of antioxidants: name ly BHT and ascorbic acid in a minced system ( 63 ). Increases in hydrophobic character and ultra violet absorption (U .V.) serve as markers of protein destabilization and are pr esumed indicators of protein oxi dation. With structural change and denaturation, hydrophobic residues located within the interior of a protein become exposed. Li and King ( 129 ) observed an increase in chicken myosin hydrophobicity in an iron (II) /ascorbate mediated oxidation system ( 129 ). Wang et al.( 130 ) observed an increase in hydrophobicity in cold stored (-15 or -29C) beef heart surimi along with increases in lipid oxidation and loss of sulfhydr yl groups. Wang et al.( 130 ) elucidated that increases in lipid oxidation may cause or be mechanis tically related to protein oxida tion development. Past studies have concluded that lipid oxidation induces pr otein oxidation; however the mechanisms of protein oxidation in the presence of lipids remain inconclusive. Functional properties and protein oxidation Increased hydrophobicity and cross-linking as a result of protein oxidation give rise to decreased solubility. According to Buttkus ( 131 ) and Dillard and Tappel ( 132 ), significant solubility decreases were attri buted to malonaldehyde protein a dduct products in muscle foods. Gel electrophoretic patterns re vealed protein aggregation with free-radical induced myosin oxidation. These patterns were analogous to electrophoretic patterns for processed meat ( 88, 121, 133 ). Jarenback and Liljemark ( 134 ) observed extensive decreas es (90%) in cod myosin solubility in a linoleic hydrope roxide system. Significant decrea ses in solubility have been observed in various other tran sition/metal mediated myosin (Turkey, beef) oxidation systems ( 123, 133 ). According to Decker et al ( 133 ), transition metal/ascorba te induced protein oxidation

PAGE 35

35 of turkey myosin resulted in decreased gelling ability, solu bility, and hydration. Decreases in functionality were attributed to increased leve ls of carbonyls and most likely cross-linking of oxidized proteins ( 133 ). Both a loss ( 135 ) and increase in gelling ability ( 121, 136 ) have been observed in the oxidation of myosin. In a storage study conducted by Saeed and Howell ( 70 ), viscoelastic and elastic modulus were measured in myosin prep ared from matched Atlantic Mackerel fillets stored at -20 C and -30C, for up to two year s. The storage/elastic modulus (G) and viscous modulus (G) were higher in the -20C versus the -30C derived myosin paste and were higher with increased storage. This is most likely due to covalent and/or other interactions in actomyosin ( 70 ). In this study, protein denaturation and a ggregation correlated well with the loss of ATPase activity and protein solubility show n in previous studies conducted by Saeed and Howell ( 63 ). The elastic modulus was significantly decreased with the incorporation of tocopherol (500 ppm) this was attributed to effective lipid oxidation pr evention. Transfer of free radicals or protein-lipi d interactions could proc eed from lipid oxidation ( 70 ). It was proposed that protein G and G were reduced due to less lipid oxid ation, induced aggregation. The samples with greater lipid oxidation had greater levels of protein aggregation.. It was inferred that the use of tocopherol w ould quench the radicals and redu ce the formation of lipid and protein oxidation product and thus reduce the ge lling ability. Cross-linking can occur through protein free radicals in musc le systems possibly causing aggr egation and may improve gel formation and strength ( 63 ). Liu and Xiong ( 137 ) found that in an ir on/ascorbate/H202 mediat ed myosin oxidation system, gel elasticity was signifi cantly reduced (40%) compared to the non-oxidized myosin. This decrease in elasticity was attributed to protein oxidation. Myosin cross-linking and scission

PAGE 36

36 were exibited by the presence of protein fragments and aggregates ( 137 ). In contrast to this study, Srinivasan and Hultin ( 121 ) found that the iron/ascorbate i nduced oxidation of washed and minced cod muscle resulted in increased gelli ng ability and carbonyl content. Hitherto, few studies have correlated measures of protein oxidation with protei n functionality and structure. Significance The development of lipid oxidation at low and high pH in fish homogenates and in protein isolates obtained from the alka li-acid aid process has only been i nvestigated to a limited extent. This research could be very helpful in impr oving the stability of th e homogenates and fish protein isolates in processing a nd utilization of pela gic species, which othe rwise are highly prone to oxidation. In addition, the im pact of lipid oxidation on protei ns during the acid and alkaliaided process has not been fully investigated. Contradictions exist as to what are the key catalysts of lipid oxidation and very limited knowle dge is available on the role of different key catalysts or mediators of oxidation at low and hi gh pH in muscle homogenates. This work will further strengthen the argument th at heme proteins are the prime mediators of lipid oxidation in fish muscle homogenates and prot ein products thereof. Basic work in this area is expected to greatly further our knowledge on oxidation processe s at low and high pH and allow for effective controls to minimize oxidation.

PAGE 37

37 CHAPTER 3 EFFECTS OF ALKALI/ACIDIC ADJUSTM ENT ON LIPID OXIDATION IN SPANISH MACKEREL HOMOGENATE Introduction Current fish protein sources throughout the world dwindle a nd methods to efficiently use existing fish supplies are well warranted. A nove l acid/alkali-aided process enables for the isolation of fish proteins from traditional and unconventional ( by-catch and pelagic) sources ( 7, 8 ). In this process, fish is homogenized with wate r (1:9). The homogenate is then adjusted to low pH (2.5 to 3.5) or high pH (10.511.5) to solubili ze the muscle proteins. Soluble proteins are then separated from unwanted fractions ( 14 ) using centrifugation at 10,000g. The solubilized proteins are then subjected to isoelectric conditions (pH 5.5 to 7.0) and are centrifuged at 10,000g. The final protein isolate ( 138 ) can then be used for va rious seafood analog products (surimi) and ingredient applications ( 8 ). Efforts to access the protein potential of low value fish sources are currently underway. Conventional processing of pelagic species presents various challe nges due to their small size, shape, bone structure, and high levels of pro-oxidants and unstable lipids ( 4-7 ). The novel alkaline/aided process, versus conventional techni ques for the isolation of fish proteins, offers many advantages including improved yiel d, functionality, stab ility, and color ( 6, 13, 16, 18, 19 ). Enhanced protein stability arises from the reduc tion in lipids and pro-oxi dants. Triglyceride and phospholipids content is reduced through the acid/alkali-aided pr ocess; however, it is the reduction in phospholipids and heme proteins that ar e attributed to lower le vels of lipid oxidation and increased stability ( 5, 9, 10 ). Throughout the acid/alkali aided process two major reactants in lipid oxidation are most important: hemoglobin and lipids. Hemoglobin is deemed the primary catalyst of lipid oxidation in fish ( 12, 42 ). Hemoglobin undergoes various change s at low pH that could increase its

PAGE 38

38 oxidative capacities ( 43, 51 ). Minimal amounts of lipid ( less than 1%) in the presence of hemoglobin, are sufficient for lipid oxidation to occur ( 58, 139 ) Separating all lipid and hemoglobin from the fish protein isolate is not feasible; thus, lipid oxidation is inevitable during the acid/alkali-aided process and measures to delay and/or preven t lipid oxidation are necessary. The effects of the alkali/aci d-aided treatment on lipid oxida tion in fish homogenate has received limited atten tion. Undeland et al. ( 67 ) investigated how low pH (2.7) adjusted herring homogenate was affected by pH exposure, centr ifugation, and antioxidant addition. Addition of erythorbate (0.2%) by itself or in combination with STPP (sod ium tripolyphosphate) (0.2%) and EDTA (0.044%) as well as high speed centrifuga tion of the homogenate significantly reduced lipid oxidation. Thus far, lipid oxidation in homogenates at extended exposure times (up to 12 hours) and expanded pH ranges (2.5 to 11.5) have not been studied. The objectives of this study were to investigate: (1) the e ffects of low and high pH on the development of primary and secondary products of lipid oxidation in fish muscle homogenates, (2) the effect of preformed lipid oxidation products in ra w material on lipid oxidation development in the homogenates at high and lo w pH, (3) the effectiveness of different antioxidant preparations on lipid oxidation in fi sh muscle homogenates at low and high pH and (4) the relevance of these homogenate findings to lipid oxidation at subsequent steps in the alkaline/acid aided process. Materials and Methods Raw Materials The first batch of Spanish mackerel ( Scomberomorous maculates ) was obtained from Savon Seafood in May 2004 (Tampa, FL). Postmortem age of the fish was approximately 3 days. Due to limited supplies of Spanish mackerel and inability to control fo r batch variability, the second batch of Spanish mackerel was fres hly caught in September 2005 by lab mates and

PAGE 39

39 myself on Captain Denn is Voyles charter, Reel Therapy (Cedar Key, Florida). Fish were manually skinned, gutted, and filleted. One half of the fillets were minced using a large scale grinder (The Hobart MFG. Co., Model 4532, Tro y, Ohio). Fillets and ground fish (100 gram quantities) were vacuum packed a nd stored at -80C until needed. Chemicals Tricholoracetic acid, sodium chloride, chloro form (stabilized with ethanol), methanol, ammonium thiocyanate, iron sulfate, barium ch loride dihydtrate, 12 N hydrochloric acid, sodium hydroxide, tocopherol, Lascorbic acid, alpha-toc opherol, and citric acid were obtained from Fischer Scientific (Fair Lawn,N ew Jersey). Thiobarbituric ac id and 1,1, 3,3 tetraethoxypropane were obtained from Sigma Chemical Co. (St. Louis, MO). Propyl gallate and (ethlenediaminetetraacetic acid 99%), EDTA we re obtained from Acros Organics (New Jersey, USA). Storage Studies and Preparation of Homogenate Pre-frozen fish (-80C) were thawed in plastic bags under cold running water and immediately placed on ice for furt her preparation. For the aging studies, fillets and/or mince were held in one gallon zip lo ck bags at 4C. The homogenate was prepared by diluting the mince with cold deionized water (1:9) and mi xing (45 second at 4C, speed 3) using a conventional kitchen blender equipped with a rheostat (Staco Inc., Dayton, Ohio, 3p 1500B). The homogenate was aliquoted equally into 250 mL glass beakers. Homogenates were adjusted to appropriate pH (2.5, 3, 3.5, 6.5, 10.5, 11, and 11.5) through drop-wise additions of 2M HCL or NaOH. The homogenates were lightly stirred over th e 12 hour duration at 4C. To ensure that pH values were maintained, homogena tes were periodically checked and readjusted with drop-wise amounts of 2 M NaOH or HCL.

PAGE 40

40 Antioxidant Implementation Antioxidants (tocopherol, ascorbic acid, ci tric acid, and EDTA) were thoroughly mixed into the homogenate at 0.2% pr ior to pH adjustment. When us ed in combination, antioxidants were added at 0.1% or 0.2% each prior to pH adjustment. Total Lipid Analysis Total lipid was measured in the minced muscle according to Lee et al. ( 140 ) using a 1:1 methanol/chloroform ratio. Results are repo rted as percentages based on wet weight. Lipid Hydroperoxides For designated time points, 3 grams of homoge nate per each treatment was sampled for lipid hydroperoxides. Samples were analy zed immediately. Lipi d hydroperoxides were determined according to Shantha et al. ( 141 ) with modifications by Undeland et al. ( 57 ). Total lipids were extracted using 9 mL of a chloroform/methanol (1:1 ) mixture with vortexing. Sodium chloride, 3 mL (0.5%), was then added to th e sample. The sample was then vortexed for 30 seconds. The samples were centrifuged (Ependo rf, model 5702, Brinkman Instruments, Inc., Westbury, N.Y) at 4 C for 10 mi nutes at 2,000 g. To alleviate pha se separation problems due to pH, homogenates were adjusted to pH 6.5 before extraction with experimentally determined volumes of 1N NaOH or HCL. Two milliters of the bottom layer was obtained using a 2 mL glass syringe and needle (Micro-Mate Intercha ngeble glass syringe, stainless steel needle, 20G x18, Popper and Sons, Inc., New Hyde Park, N.Y) Two milliters of the chloroform layer was mixed with 1.33 mL of chloroform/methanol (1:1 ). Ammonium thiocyantate (50 uL) was added to the sample, and then iron chloride ( 94 ), 50 uL, was added to the sample. Following each reagent addition, the mixture was vortexed for 2-4 seconds. After an incubation period of 5 minutes at room temperature (22.5C), samples were read at 500 nm using a spectrophotometer (Agilent 8453 UV-Visible, Agilent Technologies, Inc., Palo Alto, CA). A blank contained all

PAGE 41

41 reagents but the sample. A standard curve was co nstructed using a solution of iron (III) chloride. Samples were kept in subdued lig ht and on ice during the assay. Thiobarbituric Acid Reactive Substances (TBARS) TBARS were assessed acco rding to Beuge et al. ( 142 ) as modified by Richards et al. ( 35 ). A TCA/TBA solution (50%/1.3%) was freshly prepared and h eated to 65C on the day of analysis. It is imperative that the temperatur e not exceed 65C, to ensure that the TCA is completely dissolved and that the TBA is not heat abused. The TCA/TBA solution was added to 0.5 g of homogenate and was heated for one hour at 65C. The samples were centrifuged (4C) for 10 minutes at 2,500 g and read at 532 nm using a spectrophotometer (Agilent 8453 UVVisible, Agilent Technologies, Inc., Palo Alto, CA). A standard curve was constructed using 1,1, 3,3 tetraethoxypropane. Analysis of Moisture and pH Moisture content was assessed using a moistu re balance (CSI Scientific Company, Inc., Fairfax, VA). The pH was measured using an elec trode ( Thermo Sure-Flow electrode (Electron Corp., Waltham, MA) and pH meter (Denver Instrument, model 220, Denver, CO). The homogenate samples were directly m easured for pH with continuous mixing. Statistical Analysis The SAS system was used to evaluate st orage, pH adjustment, and antioxidant implementation studies over time. All analyses were performed in at least duplicate and all experiments were replicated twi ce. Analytical variation was es tablished through first fitting two way ANOVA models. Then multiple pair-wise comparisons were performed on estimated measurement means, using Tukey adjust ment controlling for Type I error. Data was reported as mean standard deviation. The analysis of varian ce and least significant difference tests were conducted to identify differences among mean s, while a Pearson correlation

PAGE 42

42 test was conducted to determine the correlati ons among means. Statistical significance was declared at p < 0.05. Results Changes in Spanish Mackerel Lipid Oxidation Water and lipid Content As it is a rarity to receive fresh and quality consistent raw material for processing, it is important to understand how pre-formed lipid oxidati on products in the raw material affect lipid oxidation of mackerel muscle homogenates at low and high pH. Upon receiving and/or harvesting of fish, baseline levels of oxidati on were measured to assess similarities and differences in lipid oxidation of fish freshly cau ght or at postharvest st orage for a few days. To determine the quality of the Spanish mackerel raw material, lipid oxida tion products for fresh and aged fillets and/or mince were measured. The average moisture content, pH, and lipid content for Batch 1 and 2 fish musc le are summarized in Table 3-1. The Spanish mackerel obtained in May (Batch 1) was significantly (p <0.05) higher in lipid content (7.5 2.12 %) than the September catches (Batch 2) 2.65 0.07%. Average moisture contents for May (Batch1) and September catch (Batch 2) were 79.251% and 77.76%, respectively. Average pH content of the Sp anish mackerel mince for May (Batch 1) and September (Batch 2) were 6.49.045 and 6.49.015, which are in agreement with previous research ( 67 ). Depending upon various environmental factors ( 20 ), the observed differences among the batches are to be expected. Addressi ng compositional difference in catch due to differences in seasonal variability and postm ortem age is important to study the changes observed in the acidic-alkaline aided process. Industrially, these compositional factors could have significant bearing on processing parameters and the quality of prot ein isolates obtained on a mass processing scale.

PAGE 43

43 Lipid Oxidation in Whole a nd Minced Spanish Mackerel It was evident that aging of the fish muscle at 4C led to an increase in oxidation products (Figure 3-1). Significant (p<0.05) increases in lipid oxidation for Batch 1 mince and Batch 2 mince were observed for up to 4 days and between 2 and 4 days of storage, respectively. Batch 2 mince after 4 days of storage reached equal leve ls of oxidation (~30 umol MDA/kg of tissue) as Batch 1 mince at day 1. To det ect changes in lipid oxidation incurred through mincing of raw material, a comparative analysis of fillets and mince in Batch 2 Spanish mackerel was conducted (Figure 3-1). Lipid oxidati on in the fillets did not significantly increase over time. Batch 2 mince accumulated overall significantly (p<0.003) higher TBARS levels than its corresponding fillets. The levels of lipid oxidation for same day storag e were not significantly different between Batch 2 fillets and Batch 2 mince; except for da y four, (p<0.05). Lipid oxidation levels were significantly higher in Batch 2 for days 2 and 3 th an 0 and 1 days of storage for Batch 2 fillets. Maximum TBARS values of ~70 and ~34 mol of malondialdehyde (MDA /kg) for Batch 1 mince and Batch 2 mince were reached after 1 an d 4 days, respectively. When comparing lipid oxidation over specific days, Batch 1 mince wa s significantly (p<0.05) greater than Batch 2 mince from 2 and 4 days of refrigerated storage. Increasing the age of the raw material, si gnificantly affected the lag time (P<0.004) between Batch 1 and 2 mince. The onset of rancidity according to Richards et al. ( 14) is marked by a TBARS increase of 20 umol/kg of tissue or mo re. Rancidity occurred within 1 day and after 2 days of storage in Batch 1 mince and Batc h 2 mince, respectively. TBARS were overall significantly (p<0.001) higher in Batc h 1 than in Batch 2 mince. Lipid Oxidation in pH Treated Spanish Mackerel Homogenates An initial interest was to understand how pr eformed oxidation products in raw material affected lipid oxidation with storage. The next ob jective was to evaluate the effects of low and

PAGE 44

44 high pH on lipid oxidation in fish homogenates with respect to initial fish quality, cold storage, and pH holding time. Changes in homogenate lipid oxidation with storage Figure 3-2 (b,d) and 3-3(b,d) a nd 3-4 (b,d) show that Batch 1 and 2 homogenates formed significantly greater levels of TBARS with one and two days of mince storages versus homogenates made from freshly prepared min ce (p<0.001). As shown by Figures 3-2 (b) and 3-4 (d), TBARS levels in Batch 2 homogenates prep ared from mince stored for two days were equivalent to Batch 1 homogenates made from fresh mince. Initial levels of PV and TB ARS (Figures 3-2, 3-3, and 3-4) were equal in homogenates prepared from mince stored for 0 or 1 days and were significantly (p<0.001) less than homogenates from mince stored for two days. Ini tial Batch 1 PV values for day 0, 1, and 2 were: 0.54 0.017, 0.60.018 2.52.42 mmol LPH/kg homogenate. Initial Batch 2 values at day 0, 1, and 2 were 0.75.01, 0.630.09, 0.88.007 mmol LPH/kg tissue, respectively. Significant (p<0.001) increases in homogenate PV values fo r 0 and 1 days of mince storage were observed after twelve hours of pH exposure. Endpoint leve ls of primary oxidation were equivalent in homogenates made from 1 and 2 day stored mince and were significantly greater than homogenates made from freshly prepared mince. Batch 1 and 2 homogenates made from either 0 or 1 days storage mince started at approximately the same TBARS levels (Figure 3-3). Batch 1 initial values on day 0, 1, and 2 were: 3.93 .077, 6.025.33, and 13.62.38 umol MDA/ kg of homogenate. Batch 2 initial values at day 0, 1, and 2 were 3.22 1.26, 6.09.57, and 7.23 0.94 umol MDA/kg homogenate. Figures 3-3 (b,d) and 3-4 (b,d) show that with holdi ng times of four or more hours, day one TBARS values reached equal or greater levels of oxidation compared to initial day 2

PAGE 45

45 value (p<0.05). Holding times greater than 8 hours had no affect on TBARS formed in homogenates made from mince aged for 1 or 2 days. Changes in lipid oxidation with pH and storage As Figure 3-2(b) shows, pH 3 and 3.5 homoge nates made from freshly prepared Batch 1 mince had significantly greater TBARS levels (P< 0.001) than all other pH treatments. The levels of TBARS in the pH 3.5 homogenates were signif icantly greater that th ose at pH 3 (p<0.016). Maximum levels of lipid oxidation in pH 3.5 a nd 3.0 homogenates made from freshly prepared Batch 1 mince were 64.00 .58 and 53.52.98 um ol MDA/kg of homogenate, respectively (Figure 3-3b). TBARS levels were significantly greater in homogenates ma de from 1 day stored mince versus freshly prepared mince (Figure 3-3b). Overall day 1 levels of lipid oxidation at pH 3 and 3.5 were not significant from day 2 levels (F igure 3-4b). After one day of mince storage, levels of lipid oxidation at pH 2.5 were significan tly (p<0.001) greater than control and alkaline treatments (Figure 3-4b), with the pH 2.5 homo genate still accruing greater levels of lipid oxidation after two days of refrig erated storage (Figure 3-4b). Af ter two days of mince storage, pH 2.5 TBARS were significantly (p= 0.007) less than observed levels at pH 3 (Figure 3-4b). No significant increases in lipid oxi dation were observed among the control or alkaline treatments with mince storage (Figure 3-2b,3-3b,3-4b). Changes in lipid oxidation with initial quality and pH Homogenates 3.5 and 3 had significantly great er TBARS than pH 2.5, while TBARS were significantly higher in homogenate 2.5 than pH 6.5, 10.5, 11.0, and 11.5 homogenates. For Batch 1, levels of oxidation at pH 3.5 and pH 3 were not significantly different. Control (pH 6.5) homogenates in Batch 1 devel oped significantly greater TBARS than alkaline homogenates. For Batch 2, pH 3.5 homogenates had significantly greater TBARS than alkaline pH homogenates (p<0.04). Batch 1 mince acidic homogenates oxidi zed significantly more than Batch 2 acidic,

PAGE 46

46 alkaline, and control homogenates. Batch 2 alkaline (pH 10.5, 11, 11.5) treatments incurred significantly (p<0.001) less TBARS than control (6.5) homogenate in Batch 1 and equivalent levels to the alkaline treatments in Batch 2. Changes in lipid oxidation with holding time and pH After 4 hrs and up to 12 hrs, Batch 1 incurred significantly greater PV levels than Batch 2 mince (Figures 3-2a,c). PV levels (mmol LPH/ kg homogenate) started out significantly higher (p<0.001) in Batch 1 compared to Batch 2; however, after 12 hours the Batch 1 and 2 homogenates had equivalent PV levels. As shown in Figure 3-2(a), PV values in creased significantly af ter four hours for homogenates at pH 3.5. Homogenate 2.5, 3, and 6.5 si gnificantly increased in PV levels after 12 hours. Homogenate 3 and 3.5 reached the same levels of primary oxidation after 12 hours and were followed by homogenates 6.5, 2.5, and alkaline pH in descending order of primary oxidation. From Figures 3-2 (b), homogenate holding time s of four or more hours at acidic pH were sufficient for rancidity to occur in Batch 1 homog enates. Homogenates held for 4 hours at acidic pH had significantly (p<0.001) gr eater TBARS than control and alkaline homogenates held for the same time. A similar observance was found in unadjusted pH turkey homogenates with 2.5 hours of refrigerated storage ( 143 ). The pH 3.5 homogenates were not significantly different than pH 3 homogenates after 4 hours. After ei ght hours, pH 3 and 3.5 homogenates were significantly greater th an pH 2.5 (p<0.05). After a 12 hour ho lding time, homogenates at 3.5 had significantly (p =0.027) greater TBARS than pH 3.0 homogenates, and both pH treatments proved to be significantly greater (p<0.001) than all other pH treatments at pH 2.5 and higher pH values (6.5, 10.5, 11, 11.5). The control (pH 6.5) treatment had equal levels of oxidation compared to pH 2.5 and alkaline values after 12 hours.

PAGE 47

47 Overall pH effects The extent and rate of lipid oxidation (TBARS) was signifi cantly (p<0.001) higher in homogenates at acidic pH (2.5, 3, 3.5) than at alkaline (10.5, 11,11.5) or control pH (6.5). Figures 3-2(b), 3-3(b), 3-4(b) shows that there was no si gnificant difference in TBARS formation among alkaline homogenates. TBARS fo rmation proceeded faster and to a greater extent in the following homogenate treatment order: acid> cont rol>alkaline. Levels of TBARS in descending order relative to individual pH treatment were as follows: 3.5, 3, 2.5, 6.5, and alkaline pH (10.5, 11.0, 11.5). Levels of lipid oxidation among alkaline homogenates were insignificant. Lipid Oxidation of Homogenates wi th Antioxidant Implementation The effect of antioxidant implementation in homogenates at low and high pH was worth investigating, as it may decrease overall levels of lipid oxidation in the early stages of the alkali/acid aided process and improve the quality of the final protein isolate. A commonly used combination of alpha tocopherol and ascorb ic acid in muscle foods was employed. Batch 1 mince was stored for one day and was subjected to pH 3 as per findings in the previous section, in which rancid ity occurred after one day of st orage and the greatest overall levels of homogenate lipid oxidation developed (3 and 3.5). The levels of lipid oxidation of homogenates from day old mince at pH 3 and 3.5 were not significantly different. pH 3 was chosen, as this is close to the acid-aided process pH values used for the isolation of proteins. To ensure that Batch 2 was of comp arable oxidative quality of Batc h 1 after one day of storage, Batch 2 mince was stored for 4 days (See Figure 3-1). The antioxidants ascorbic acid and/or tocophero l, citric acid/and or EDTA were added at two concentrations, 0.1% and 0.2% w/w, in nu merous combinations. The effects of these

PAGE 48

48 antioxidants were also tested under physiologi cal pH conditions (pH 6.5) in the homogenate system. Levels of lipid oxidation between batches were comparable, as all antioxidant treatments at pH 3 for Batch 1 homogenates, with ex ception of pH 3 (0.1% ascorbic acid + 0.1% tocopherol) homogenates (Figure 3-5b), were not significantly diffe rent from pH 3 homogenates from Batch 2 (Figure 3-5c). However, pH 6.5 fr om Batch 1 homogenate (Figure 3-5b) formed less TBARS than pH 6.5 from Batch 2 (p=0.0002) (F igure 3-5 c). Batch 2 pH 3 (0.1% ascorbic acid + 0.1% tocopherol) formed greater TBAR S compared to the corresponding Batch 1 treatment. The difference between pH 3 (0.1% ascorbic acid + 0.1% tocopherol) samples from Batch 1 and 2 is negligible rega rding quality as both treatments formed total levels of lipid oxidation below the point of rancidity (>20 umol MDA/kg tissue). Because pH 6.5 was not comparable among batches, pH 6.5 treatments from Batch 1were only compared to other homogenate treatments prepared from Batch 1 mince. Homogenate values were compared among corresponding pH 3 treatments in both batches. In Batch 1 treatments, various differences were observed. Overall levels of TBARS formed in control pH 3 homogenates were significantly greater than all Batch 1 and 2 pH 3 treatments. Homogenate treatment at pH 6.5 (0.1% ascorb ic acid + 0.1% tocopherol) and pH 6.5 (0.1% ascorbic acid) formed the highest TBARS levels (p<0.05) among all other treatments in Batch 1. The antioxidant treatments were all equally eff ective in delaying/preven ting oxidation in the pH 3 in contrast to pH 6.5. As shown in Figure 3-6, the addition of ED TA (0.1%), citric acid (0.1%), and ETDA (0.1%) + citric acid (0.1%), at pH 3 were rather ineffective in reducing lipid oxidation (TBARS, PV). Over a twelve hour holding time PV levels at pH 3 were not significantly affected by citric

PAGE 49

49 acid (0.1%) and were significantly (p<0.001) reduced by the addition of EDTA (0.1%) and EDTA (0.1%) + citric acid (0.1% ) (Figure 3-6b). As shown in Figure 3-6b, the use of these antioxidants when evaluated sepa rately or in combination did not significantly reduce overall levels of TBARS. These results suggest that the sole incorpora tion of chelators does not protect against lipid oxidation in Spanish mackerel fish homogenates ( 51 ). However, in combination with other antioxidants, chelat ors are promising inhibitors of oxidation according to Undeland ( 67 ). Lipid Oxidation of Homogenates a fter Readjustment to pH 5.5 and 7 In the acid and alkali-aided processes, homoge nate proteins that are solubilized at low or high pH are recovered by readjus ting the pH to values where they aggregate. Consistent with previous finding, Batch 1 mince was stored for one day and was subjected to pH 3 as per findings in (Lipid oxidation in whole and min ced Spanish Mackerel, Chapter 3) in which rancidity occurred after one day of storage and the greatest ov erall levels of homogenate lipid oxidation with respect to pH (3 and 3.5) homogena tes occurred. Batch 2 mince was stored for 4 days as per previously discussed findings (Lipid oxidation in whole and minced Spanish Mackerel, Chapter 3) to ensure approximate oxi dation levels as Batch 1 samples. Homogenates at pH 3 and 11 were held for one hour and then we re readjusted to pH 5.5 and 7. The former pH is commonly used to recover fish muscle protei ns. However, improper folding of heme proteins when readjusted from low pH to pH 5.5 may cau se significant lipid oxida tion due to incomplete refolding ( 51 ). More refolding and less pro-oxidative activity is expected of hemoglobin with readjustment to pH 7 ( 51 ). The extent of lipid oxidati on increased over time (p<0.0001) for all samples. Lipid oxidation was measured at 0, 3, and 12 hours. Thes e time points were chosen as per section, Lipid oxidation in whole and minced Spanish Mack erel,(Chapter 3), in which trends of lipid

PAGE 50

50 oxidation were realized. Batch 1 and 2 homogenate s proceeded from the same initial levels of oxidation; however, greater levels of lipid oxi dation formed in Batch 1 compared to Batch 2 (p<0.0001). Batch and homogenate pH readjustment In Batch 1, greater levels of oxidation (PV, TBARS) occurred with the implementation of antioxidants and readjustment from pH 3 to e ither pH 5.5 or 7 (p<0.0002) (Figures 3-7 a,b) compared to pH 3. When homogenates were readjust ed from pH 3 to 5.5 or 7 in the presence of the antioxidant combination that was effective at pH 3 (0.1% ascorbic acid + 0.1% tocopherol) (w/w) (Figures 3-5 b,c), a significant (p<0.0001) pro-oxidative effect was noted (Figures 3-7 a,b). This could be explained by the pro-oxidative effect of this combination observed (Figure 35a) at alkaline pH 6.5, which is intermediate to pH 5.5 and 7. In terestingly, the same antioxidant treatments with readjustment fr om pH 11 to pH 5.5 or 7 did not exhibit the same pro-oxidative effect on the homogenates in terms of PV values as it did for the system readjusted from low pH (Figure 3-8a); however, the signi ficant increase in TBARS with pH readjustment to pH 5.5 or 7 and antioxidant implementation at pH 11 (Figur es 3-8b) were consistent with the same treatments at pH 3 (Figure 3-7b). According to TBARS results, antioxidant treatments with readjustment from pH 11 were not significantly different from the low levels of oxidation at control pH 11 in Batch 2 (3-8c). The pro-oxidative effect of readjustment from pH 3 to 7 with antioxidants was further emphasi zed, having greater lipid oxid ation TBARS and PV values (p<0.0035) compared to pH 7 alone as exhibite d in Batch 1 and 2 (Fi gure 3-7 a,b,c,d). TBARS and PV levels of samples with pH readjustme nt from pH 3 and 11 were less (p<0.02) than corresponding treatments with antioxidants. Oxid ation levels with read justment to pH and antioxidants proved to be higher at pH 7 than at pH 5.5 in Batch 1 (PV) p<0.0001.

PAGE 51

51 Time and homogenate pH readjustment After 12 hours, trends among measures of lip id oxidation (TBARS, PV) among pH 3 and 11 readjustment treatments were observed. TBARS levels proved to be gr eater (p<0.0001) in the pH 11 homogenate adjusted to 5.5 (0.1% asco rbic acid + 0.1% tocophe rol) compared to homogenates at pH 3, 5.5, 7, 3-5.5, and 3-7. Readjustme nt from pH 3 to 7 (0.1% ascorbic acid + 0.1% tocopherol) proved to be more pro-oxidative than the same treatment readjusted from pH 11 and all other treatments at pH 3 and 11. Samp les at pH 5.5 had greater TBARS than samples at pH 11 (p < 0.02). The pH readjustment treatment s were always surpassed in levels of lipid oxidation (TBARS) by corresponding treatments with antioxidants in Batch 1 (p<0.02) (Figures 3-7b and 3-8b). In Batches 1 and 2, readjustment s to pH 7 with antioxidants had greater TBARS than treatments adjusted to pH 5.5 with and without antioxidants except for pH (p<0.0001), except for one sample pH 11-5.5 (Figure 3-8 d). PV values seemed to follow the same trends as TBARS; however, statistically li ttle differences were observed. Discussion Lipid Oxidation and Storage Stability Aged fish muscle should in cur greater and more rapid li pid oxidation, and this was observed in the fillets and ground musc le (Figure 3-1). This is in ag reement with the literature. It has been reported that if the fish muscle ha s surpassed the induction pha se of lipid oxidation, which can occur with storage, the rate and ex tent of lipid peroxidation will be greater ( 144 ). In addition, the levels of lipid oxidation assessed in the ground muscle were significantly higher than those observed in the fillets. This is eviden t in the fact that mince from freshly caught fish (Batch 2) accumulated significantly higher levels of lipid oxidation than the corresponding fillets (Figure 3-1). Regardless of postmortem age of the fish, there is a significant different in the extent of lipid oxidation with mechanical pro cessing. Aside from compositional muscle aspects,

PAGE 52

52 the differences in lipid oxidation in the mince and fillets are well explained and are consistent with observed changes in muscle due to extrinsic factors. Various external factors can contribute to lipid oxidation in muscle food including and not limited too: temperature, oxygen exposure, light, mechanical processing, packaging atmo sphere, introduction of pro-oxidants, and cooking/heating ( 145 ). The extrinsic factor most relati ve to the storage study is mechanical processing of meat. Various levels of oxi dation will arise through mincing, flaking, emulsification, and other physical handling. Tissu e disruption increases oxygen-lipid interaction, and thus, increases the occurrence of lipid oxidation. Mincing also releases pro-oxidants such as hemoglobin and myoglobin and brings th em in closer contact with lipids ( 80, 146 ). The Spanish mackerel muscle was maintained at 4C during the grin ding process; however, a painty aroma was observed and is indicative of increased lipid oxidation due to tissue disruption ( 30, 147 ) With storage and physical manipulation of tissue, reduced quality is evident in flavor, odor, color, and nutritional changes ( 29, 30 ). This storage study served as comparative assessment of raw material quality, and thus, serves as an a pplicable baseline for ch anges in lipid oxidation implicated in the alkali/acid-ai ded protein isolation process. Lipid Oxidation in Acid/Alkali Adjusted Homogenates The initially higher PV levels in homogenate s from Batch 1 mince is to be expected (Figures 3-2, 3, 4). Batch 1 was stored for seve ral days prior to arriva l. Lipid hydroperoxides are initial products of lipid oxidation. Lipid hydroperoxides are transient lipid oxidation products, degrading soon after formation ( 48 ). Free radicals more read ily react with antioxidants than lipid components, due to bond dissociati on energies. Once antioxidant defenses are depleted, the rates of lipid hydroperoxides and se condary product formation will proceed faster ( 148 ). According to Petillo et al. ( 143 ), antioxidant sources found in the blood plasma of stored mince Spanish mackerel were exhausted with stor age. Ascorbate and glut athione were exhausted

PAGE 53

53 first as they serve to reduce the activa tion of metal containing pro-oxidants ( 148 ). Increases in volatile secondary lipid oxidation products we re perceptually detected in the homogenate headspace; however, further studies are needed to substantiate these findings. The greatest intensities were observed in the acidic treatments and negligible differences were detected in the alkaline treatments. These easily detected volatil es are most likely alde hydes and can contribute to rancidity and alterations in proteins comprising DNA, enzymes, and phospholipids structure ( 87, 149, 150 ). In studies conducted by Alasalvar ( 151 ), the most predominant oxidation products formed from cold storage of mackerel were hexanal, heptan al, octanal, dodecanal, 2pentenal, methyl-2-pentenal E-2-hexanal, a nd Z-4heptenal. Thus these volatiles could potentially contribute to oxi dation of the homogenates. Comparable TBARS levels in homogenates made from one and two day old Batch 1 mince is most likely due to the depletion of hydrope roxides, thereby, slowing the formation of secondary products ( 148 ). Greater levels of lipid oxida tion were observed among acidified homogenates (Figures 3-2,3,4). Thes e high levels may be explained in part by the high moisture content (97-98%) in the homogenates. Water activity is an influentia l factor in the rate of lipid oxidation. It is known that with high water activity, the rate and extent of lipid oxidation is greater ( 148 ). Thus, the homogenates would accumulate higher levels of oxidation, not only due to acidification but due to enhanced incorpor ation of pro-oxidants a nd oxygen via the water phase. During the homogenization process and pH adjustment, various changes to the predominant catalysts, heme proteins can occur a nd affect rates of lipid oxidation. According to Shikama ( 60 ) and Kristinsson ( 152 ), heme proteins become oxidat ively stable due to increased heme group preservation at alkaline pH ( 60, 152 ). With lowering of pH, subunit dissociation,

PAGE 54

54 unfolding, release of heme groups and/or iron, a nd loss of oxygen can occur with heme proteins ( 43, 50, 52, 153-155 ). Such modifications are more li kely when hemoglobin/ myoglobin are subjected to low pH; thereby, converting hem oglobin to more pro-oxidative forms including deoxy-, met-, perferryl-, and ferryl-hemoglobin/myoglobin ( 37 ). Hemoglobin/ myoglobin Bohr effect is amplified with decreases in pH below neutrality and leads to deoxygenation ( 50 ). According to Richards et al. ( 42 ), deoxyhemoglobin/myglobin ar e highly pro-oxidative. Consequentially, activated heme proteins can give rise to the production of reactive oxygen species and decomposition of lipids and/or lipid oxidation products as observed in the acidified homogenates. The effect of pH on hemoglobin is particularly relevant to the pro cessing of migratory pelagic species. Hemoglobin in dark muscle pe lagic species has a lower affinity for oxygen. A more flexible tertiary structure facilitates qu ick release of oxygen for metabolic respiration ( 156 ). Richards and Hultin ( 157 ) reported significantly greater catalytic activity in mackerel versus trout hemoglobin ( 157 ). Significant pH effects could be re alized during the isolation of dark muscle fish species, due to less stable hem oglobin. Further investigation is required to understand the effects of pH shifts on heme protei ns in a variety of fish species homogenates and how these implications affect lipid oxidation during protein processing. The effects of pH in muscle foods, or si ngle or multi-component model systems could explain the effect of pH in the lipid oxidation of homogenates. Hi gher levels of lipid oxidation in the homogenates at low pH could be explained by heme protein changes in aqueous model systems. A study conducted by Kristinsson ( 12 ) revealed that consider able changes in flounder hemoglobin tertiary structure occurred with lowering of pH. The most significant unfolding transitions occurred between pH 4 and 3.5 and the greatest unfolding occurred at pH 2.5 ( 12 ).

PAGE 55

55 Hemoglobin unfolding occurred very quickly (s econds) at pH 1.5-2.5, whil e unfolding at higher acidic pH (3-3.5) took over 30 minutes. Alkali ne pH (10-11) brought about very little or negligible changes in heme structure ( 12 ). The low levels of lipid oxidation in the alkaline homogenates can be explained by he me group and oxidative stability ( 51 ). The proxidative activity of hemoglobin/myoglobin at low pH and inactivity at high pH is greatly dependant on distal and proximal histidine interact ions with the heme group and oxygen ( 158 ). Under acidic conditions, interactions with the heme and distal histidine and proximal histidine can be lost through autoxidation and denaturation ( 12, 152 ). Under alkaline conditions, substantial heme unfolding at low pH would explain the high le vels of oxidation observed in the acidic homogenates (Figures 3-2 ,3 ,4). In a study by Kristinsson and Hultin ( 51 ), conformational changes were reported for trout hemoglobin (0.8 uM in water, 5C) at pH 3, 7, and 10.5 over time. According to UV-Vis spectral (300-700) trout hemoglobin c onformational changes occurred in two minutes at pH 3; while cha nges at pH 7 and 10.5 were observed in 29 and 120 hours, respectively ( 159 ). The levels of oxidation in the Spanish mackerel homogenates at neutral pH were less than acidic ho mogenates; this is most likely due to less structural flexibility and hydrophobicity of hem oglobin at neutral pH ( 51 ). Additionally, homogenates at pH 6.5 were more susceptible to autoxidation relative to alkali ne pH and significantly less susceptible to autooxidation at low pH. Richards and Hultin ( 43 ) found that with lowering of pH from 7.6 to 6.0, hemoglobin was quickly converted to pro-oxida tive forms: deoxyhemoglobin and metmyoglobin ( 43 ). In the same study, Richards and Hultin ( 43 ) found that a hemoglobin mediated washed cod system had shorter lag phases at lower pH ( 43 ). Thus, the differing leve ls of lipid oxidation in homogenates at neutral pH could be caused by less oxidative stability an d considerably more stability of hemoglobin at alkalin e and acidic pH, respectively.

PAGE 56

56 Various conformation changes within hemogl obin/myglobin could explain the differences in oxidation among the pH treated homogenates. Because the extent of unfolding is greater at low pH versus neutral and alkaline pH, he moglobin becomes more hydrophobic and thus is better able to interact with lipid membranes ( 152 ). Additionally, Kristinsson ( 12 ) found that flounder hemoglobin existed as a molten globule (partially unfolded) state at low pH. A molten globule is both hydrophobic and hydrophilic in na ture, providing greater accessibility to lipids. Thus at low pH, hemoglobin would be very pro-oxi dative in the presences of membrane lipids in the homogenates. Access of the pr otein to lipids is not only enha nced at low pH, but the heme group becomes significantly exposed at lo w pH. According to Kristinsson ( 12 ), increased hydrophobicity in flounder hemoglobin was attribut ed to greater trypt ophan residue exposure with lowering of pH from neutrality (pH 7). Increased rates of met-myoglobin/hemoglobin formation are known to occur at low pH ( 60 ). Autoxidation, unfoldi ng, and dissociation of myglobin/hemoglobin can give rise to various reactive species su ch as ferryl hemoglobin. Ferryl hemoglobin is believed to be a potent catalyst of oxidation in muscle foods, particularly at low pH ( 56 ). These oxidized forms of hemoglobin/myglobi n can readily catalyze lipid oxidation at low pH ( 160 ) and could explain greater lipid ox idation in the acidic homogenates. Results from washed cod muscle systems could help explain the overall levels and rate of lipid oxidation (TBARS) in Spanish mackerel homogenates. The pH 6.5 homogenates were lower than observed acidic homogenates and higher than alkaline homogenates in Batch 1 (Figures 3-2b,3b,4b). Homogenate pH exposure times significantly a ffected the lipid oxidation development in homogenates. Acidic homogena tes made from freshly prepared Batch 1 mince developed rancidity after four hours while ne utral pH developed rancidity le vels after 12 hours and alkaline

PAGE 57

57 pH homogenates did not reach ra ncidity levels. Pazos et al. ( 161 ) observed similar trends in a hemoglobin mediated lipid oxidation system usi ng washed cod. At pH 3, the onset of lipid oxidation (~5 hours) was less than pH 6.8 (~ 10 hours) and the endpoint oxidation levels were comparable. Likewise, the pH 3.5 and 3.0 homoge nates from freshly prepared Batch 1 mince developed rancidity levels in approximately 6 hours, while pH 6.5 developed rancidity in approximately 12 hours. Pazos et al. ( 161 ) observed similar endpoint levels of oxidation (TBARS) in the pH 3 and pH 6.8 adjusted washed systems. The pH 3 and 6.8 homogenates, from freshly prepared Batch 1 mince, did not reach equivalent leve ls of oxidation at 12 hours; however, greater holding times or mince storage may have resulted in equal endpoint levels of lipid oxidation (Figures 3-2(b), 3-4(b)). As obs erved in the homogenates, Kristinsson and Hultin ( 51 ) reported similar lipid oxidation trends in hemoglobin mediated washed system samples. Trout hemoglobin was adjusted to acidic (pH 2.5, 3, 3.5), neutral (pH 7), and alkaline (10.5, 11) pH values. The greatest extent and rate of lipid oxidation occurred at low pH, followed by neutral pH, and then alkaline pH. Increased ra tes and extents of lipid oxidation followed by longer lag phases and decreased le vels of lipid oxidation at ne utral (6.5) was observed in the Spanish mackerel homogenates. The developmen t of TBARS was greatest in a washed cod system at pH 2.5 and 3 than at pH 3.5 ( 51 ). The homogenate samples at pH 3.5 over time had greater rates and levels of lipid oxidation than pH 2.5 and 3. As Kristinsson and Hultin ( 51 ) observed at pH 10.5 and 11 in the washed cod system, minimal levels of lipid oxidation developed in pH 10.5, 11, and 11.5 homogenates. Although quantitative color differences among the homogenates were not reported, it was evident that acidified homogenates were ye llowish white while the alkaline and neutral homogenates were pink in color. From these co lor differences, it is evident that significant

PAGE 58

58 unfolding and denaturation occurred in the acid ified homogenates, while the less unfolded and native heme proteins in the other treatments retained thei r color. Richards et al. ( 162 ) and Lee et al. ( 163 ) found a significant loss in redness, a value was due to conformational changes as heme proteins were converted from their oxy-to met-forms. Hultin ( 5 ) reported that meat color changes are byproducts of free ra dical induced oxidation, sec ondary to lipid hydroperoxide formation. Homogenate pH Readjustment To investigate changes in lipid oxidation in acid/alkali-aid proce ss upon pH readjustment, Spanish mackerel homogenates were held for one hour at pH 3 and 11 and were then readjusted to pH 5.5 and 7. The homogenates were held at pH 5.5 and 7 at 4C for 11 hours with constant mixing. The readjustment from acidic and al kaline pH to pH 5.5 and 7 mimics the pH readjustment step in the acid/alkali aid proces s in which proteins are precipitated and then centrifuged to obtain a protein isolate. To date, no studies have investigated change s in lipid oxidation over time in acid/alkali homogenates upon readjustment, nor has the effec tiveness of antioxidants in preventing lipid oxidation during homogenate readjustment been i nvestigated. Lipid oxidati on levels (TBARS) in pH 3 and 11 homogenates were not significantly affected by pH readjustment to pH 5.5 or 7 (Figures 3-7b, 3-8b). Trout hemoglobin refoldi ng upon readjustment from acidic or alkaline pH to neutral pH 7 was studied by Kristinsson and Hultin ( 51 ). Trout hemoglobin was adjusted to pH 3 and 10.5 and was then readjusted to pH 7 a nd added to a washed cod system. The effect of pH lowering and pH holding times on the pro-oxid ative activity of refolded hemoglobin were investigated. Less refolding occurred for more unfolded acidified hemoglobin (pH 2.5) compared to hemoglobin at pH 3.5. Longer acidic pH holding times and lower pH values gave hemoglobins that were incompletely unfolded a nd most likely more pro-oxidative relative to

PAGE 59

59 native hemoglobin (pH 7). Alkaline readjusted hemoglobin did not incur significant changes in pro-oxidative activity unde r refolding conditions ( 51 ). Similarly, pH readjustment of homogenates from pH 11 did not significantly change and were minimal (Figures 3-8 b). In contrast to Kristinsson and Hultin ( 51 ) findings in which increased le vels of lipid oxidation were observed upon readjustments from pH 3, levels of lipid oxidation in homogenates at pH 3 with readjustment to pH 5.5 and 7 remained unchange d. Insignificant changes in homogenate lipid oxidation levels could be attributed to cha nges in hemoglobin conformation and thus prooxidative activity. Kristinsson and Hultin ( 51 ) observed significant increases in pro-oxidative activity of hemoglobin at pH 3, when held at pH 3 for 20 minutes versus 90 seconds. Conformational studies gave proof to greater re folding with shorter ho lding times at pH 3 ( 51 ). The hemoglobin in homogenates at pH 3 were li kely held for a duration long enough (1 hour) to cause irreversible denaturation and thus inabilit y for structure recovery. Similar to lowering of pH, incomplete refolding can increase the e xposure of hydrophobic sites and heme within the tertiary structure of hemogl obin. The site exposure makes he moglobin more pro-oxidative ( 164 ). Due to longer holding times, acidified hemoglobi n was unable to refold. Pro-oxidative activity of the hemoglobin would then be very similar among the pH 3 homogenates with and without pH readjustment. Interestingly, the incorporation of tocophero l and ascorbic acid into pH 3 and pH 11 readjustment treatments proved to be pro-oxidati ve, compared to controls and readjustment treatments (Figures 3-7b, 3-8b). The pH 3 and 11 homogenates with antio xidants formed greater levels of lipid oxidation when readjusted to pH 7 compared to 5. These increases in oxidation with readjustment to neutral pH is most likely impart due to the instability of ascorbic acid at neutral pH as per discussion in the following section.

PAGE 60

60 Antioxidant Implementation As lipid oxidation could be problematic duri ng acid processing of various dark muscled fish species ( 13 ), antioxidants were added to the hom ogenates. Ascorbic acid and tocopherol, either alone or in combination at 0.1% and 0.2%, were equally effective in preventing oxidation in pH 3 homogenates (Figures 3-5 b,c). In a study conducted by Undeland et al. ( 67 ), incorporation of antioxidants in the pre-wash an d/or homogenization step significantly decreased levels of lipid oxidation in the acid-aided is olates obtained from he rring. Incorporation of erythorbate (0.2% w/w) in comb ination with EDTA (0.2% w/w) or sodium tripolyphosphate (STPP) (0.2% w/w) into herring homogenates at pH 2.5, have proven beneficial in maintaining the quality of dark muscled fish protein isolate. Erythrobate, and iso-form of ascorbic acid were found to be effective heme-iron reducing agents in herring homogenates. According to other findings ( 68, 153 ), antioxidant implementation early in the modification of mackerel fish muscle is most efficacious in reduci ng overall levels of oxidation in the final product. However, antioxidant interventions prove to be most beneficial in light versus dark mince, most likely due to significant levels of hemoglobin ( 68, 153 ). Antioxidants could not only be implemented in the homogenates, but in the pre-wash of the Span ish mackerel. Additionally, various procedural modifications in combination with antioxidant s could reduce lipid oxida tion in homogenates made from dark muscle species: washing the min ce, removal of dark muscle, and/or reducing the holding time at extreme pH ( 67 ). The effects of how tocopherol and ascorbic acid incorporation into Spanish mackerel homogenates affect lipid ox idation levels in the protein isolate requires further investigation. The incorporation of ascorbic acid (0.1%) al one or in combination with tocopherol (0.1%) at pH 6.5 (Figure 3-5b) proved to be pro-oxidativ e in comparison to corresponding treatments at pH 3 (Figures 3-5b). A decrease in lipid oxidation due to use of lower antioxidant concentration

PAGE 61

61 increases was observed in the pH 3 homogena tes (Figure 3-7b). Studies have shown a concentration and pH effect on the ability of ascorbic acid and tocopherol to scavenge free radicals. It was expected that lower levels of ascorbic acid (0.1%) rather than higher levels (0.2%) in the homogenates would be pro-oxida tive. Previous studies demonstrated a concentration effect of ascorbic aci d and tocopherol at physiological pH ( 64, 65 ). It has been reported that at high and low concentrations, to copherol and ascorbic acid impart pro-oxidant and antioxidant effects, respectively. Cillard ( 165 ) found that high con centrations ( > 7,600 mg/kg, 0.7%) and low concentration (<3,800 mg /mL, 0.38%)of tocopherol induced prooxidative effects in linoleic aci d. Ascorbic acid could be proo xidative (>5,000 mg/kg, 0.5%) in a hemoglobin containing solution ( 166 ). Watts ( 167 ) found that ascorbic acid was pro-oxidant at 176 mg/mL in a carotene-linoleate aqueous model, and it was c oncluded that chelators were needed to sequester catalytic iron in the presence of these as corbic acid concentrations. Ultimately, there is a net prooxidative or anti oxidative role played by these antioxidants. Ascorbic acid is known to promote lipid oxidation by facilitated iron release from heme proteins ( 168 ), on the other hand ascorbate can quench myoglobin ferryl radicals post mortem ( 169 ). The roles that ascorbic acid and tocopherol play in lipid oxidation is determined by net antioxidant activity. At pH 3, ascorbic acid and tocopherol either alone or in combination at 0.1% and 0.2% were equally effective in preventing oxidation (Figures 3-5b,c), very similar to the study conducted by Undeland et al. ( 67 ) in which erythorbate ( 0.2%) was added to herring homogenates. Hultin and Kelleher ( 68 ) found that washing Atlantic mackerel mince with 0.2% ascorbic solution and 1.5 mM EDTA, affec tively prevented lipid oxidation. At high concentrations (0.2%), ascorbic acid can que nch ferryl radicals. In a study conducted by Richards and Li ( 170 ) incorporation of high (2.2mM) and low levels (20-200 uM) of ascorbate

PAGE 62

62 into hemoglobin-mediated lipid oxidation system resulted in inhibition and no pro-oxidative effect on TBARS, respectively. In the same study conducted by Hultin and Kelleher ( 68 ), ascorbic acid may have prevented lipid oxidatio n through reduction of heme iron from the ferric to the ferrous form. The concentration effect of antioxidants could be very different among studies; due to differences in pH and aqueous sy stem compositions. The change in effectiveness of these antioxidants with pH and c oncentration is to be expected. Ascorbic acids structural and oxidative inst ability at neutral pH may explain the high degree of oxidation found in 6.5 pH treatments us ing ascorbic acid at 0.1% concentrations as well as significant increase in oxidation after pH readjustment to pH 5.5 and 7 (0.1% tocopherol + 0.1% ascorbic acid). Because antioxidants have various pH dependant ioni c forms, a non-linear relationship between pH and an tioxidant stability exists ( 171 ). Changes in pH are likely to cause changes in antioxidant structural stability and thus function. It has been found that tocopherol and ascorbic acid have different stab ilities at different pH values ( 172 ). In general, ascorbic acid is stable at acidic pH and unstable at neutral a nd alkaline pH. Tocopherol is stable at neutral, acidic, and alkaline pH ( 172 ). Tocopherols stability and thus ability to prevent oxidation exist over a wide pH and thus it is most likely that ascorbic acid counteracts the positive effects of tocopherol, when adjusted to neutral pH. It is li kely that ascorbic acid is more antioxidative at low pH compared to pH 6.5. Ascorbic acid is deprotonated with neutrali zation as it has a pKa 4.04 ( 173 ). Ascorbic acid could effectively reduce heme iron from ferric to the active ferrous form. The ferrous form is known to breakdown lipid hydroperoxides and hydrogen peroxide to form lipid radicals and ROS, respectively ( 169, 170, 174 ). Thus, it is likely that in the pH 6.5 homogenates, ascorbic acid is prooxidative thro ugh indirect propagation of lipid oxidation. At low pH, ascorbic acid served as an effec tive antioxidant. Similarily, Undeland et al. ( 67 ) found

PAGE 63

63 that ascorbic acid at 0.2% con centrations effectively reduced lipid oxidation in the protein isolates made from acidified (pH 2.5) herring hom ogenates. Due to the f act that ascorbic acid was not effective in reducing li pid oxidation in the homogenates at pH 6.5, a more stable analog, erythorbate, may be a better choice with pH fluctu ations. It is likely th at both concentration effect and pH influenced the effectiven ess of ascorbic acid as an antioxidant. Due to the key roles antioxidant roles exemp lified in various studies ascorbic acid and tocopherol were used in the pr esent studies. Post mortem antioxidants are preferentially exhausted due to one-electron reduction potenti als, also known as the pecking order ( 175 ) and locations within muscle cells ( 5 ). The major antioxidant defens es against lipid oxidation in muscle foods are the water-solubl e inhibitors, ascorbate and glut athione, and the lipid-soluble inhibitors, tocopherol and coenzy me Q. The tocopherol isomer of greatest abundance in skeletal muscle is tocopherol and accumulates through dietary intake ( 176 ). In a paired fillet kinetics study conducted by Petillo et al.( 143 ), the loss of antioxidants in Atlantic mackerel was as follows: glutathione and ascorbate, alpha toc opherol, and ubiquinone10 (oxidized form of coenzyme Q). The rate of antioxidant loss was sign ificantly higher in the dark muscle than in the light muscle ( 177 ). Similarly, the same sequence of antioxi dant depletions was indicative of oneelectron transfers, according to Jia et al. ( 28) and Pazos et al. ( 178, 179 ). Over time, it is evident that significant in creases in TBARS levels occurred after four hours in the pH 3 homogenate samples (Figure 35b). This is most likely due to losses in tocopherol, which is highly correlated with increases in lipid oxida tion and hydroperoxides in muscle foods ( 177, 179, 180 ). The loss of alpha tocopherol is deemed the last defense against lipid oxidation, in vitro. In turkey br east and thigh homoge nates, the loss of -tocopherol as it relates to lipid oxidation was observed. Within 2.5 hours at 4C, there was significant increase in

PAGE 64

64 lipid oxidation ( TBARS and PV) and a substantial loss of -tocopherol (50%) ( 181 ). It is evident that the loss of tocophereol was preven ted by the implementation of ascorbic acid and tocopherol. Other studies have reported preservation of tocoph erol in muscle foods. Propyl gallate and grape polyphenols were found to signifi cantly preserve tocopher ol in horse mackerel during frozen storage ( 179 ). It is commonly known that asco rbate regenerates tocopherol and serves an important function in preserving the free radical scav enging capacity of tocopherol in membrane systems ( 182 ). Ascorbate is known for its ability to reduce hypervalent hemoglobin, a highly pro-oxidative hemoglobin species. Ascorb ate also possesses the ability to break down hydroperoxides, which are activators of met-Mb/Hb to the ferryl form ( 183 ). It is widely accepted that in membrane systems, a lipophili c antioxidant (tocopherol) and a hydrophilic antioxidant (ascorbate) should be combined due to synergistic affects. In the homogenates, a significant synergistic affect was not evident as all antioxidants treatments, alone or in combination, had equivalent levels of lipid oxidation inhibition at pH 3 (Figure 3-5 b). EDTA and Citric Acid Im plementation at Low pH EDTA and citric acid at 0.1% concentrations, together or in combination were ineffective in inhibiting lipid oxidation (PV, TBARS) in pH 3 homogenates (Figures 3-6a,b). In contrast to these finding, Undeland et al. ( 67 ) found that the implementa tion of EDTA (1.5 mM; 0.044%) effectively reduced lipid oxidation in protein isolates obtained from acidified (pH 2.5) herring homogenates. Similar to pH 3 ho mogenates results EDTA (2.2 mM ) did not significantly affect lipid oxidation levels in a hemoglobin mediated lipid oxidation system according to Richard and Li ( 170 ). Due to the ineffectiveness of the chelators, it is evident that free iron or iron released from heme may not have served as the predom inant catalyst in the lipid oxidation of the homogenates. To fully evaluate the effectives of these chelators, it would be necessary to evaluate the levels of TBARS formed in the fi nal isolate. The appropr iate implementation of

PAGE 65

65 antioxidants in different types of muscle systems should be consid ered, and the effectiveness of appropriate heme-related and/or low molecular a ppropriate antioxidants should be assessed. In a model study conducted by Kanner et al ( 41 ), oxidation was genera ted through two catalytic mechanisms: hydrogen peroxide activated met-m yoglobin and iron chloride/ascorbate; however, only the oxidation induced by the heme bound iron wa s inhibited by muscle cy tosolic extracts in washed fish systems ( 41 ). Obviously the chelators and or antioxidant mechanism differ for free transition metals versus those that are he me bound proteins. Add itionally, chelators in combination with other antioxidants could be eff ective in reducing overall lipid oxidation in the final product isolate as obse rved by Undeland et al. ( 67 ). Conclusion This storage study enables the understanding of how raw material of different quality may affect lipid oxidation formation during the alkali/ acid-aided protein isolati on process, particularly in the initial homogenization and pH adjustment step of pelagi c specie protein extraction. Later research (Chapter 4) will investigate whet her the lipid oxidati on incurred during the homogenization steps affects the quality of the ac tual protein isolate. Lipid oxidation formation proved to be greatest in Spanish mackerel hom ogenates at pH 3 and 3.5. The implications of holding the proteins at an acidic pH followed by pH readjustment and isolation on protein and lipid oxidation will be addressed in the following chapter. The implementation of tocopherol and ascorbic acid at low pH signi ficantly decreased levels of lipid oxidation in homogenates; howev er, antioxidants proved to be pro-oxidative with pH readjustment of homogenates from acidic and alkaline pH to neutral pH. Increases in lipid oxidation in the readjusted hom ogenates with antioxidants were observed. Readjustment alone had no affect on the levels of lip id oxidation in homogenates. Longer holding times significantly increased the levels of lipid oxidation in the readjusted hom ogenates with antioxidants.

PAGE 66

66 Successful implementation of antioxi dants into the acid a nd alkali-aided protein isolation process requires further investigation of appropriate antioxidants and thei r effectiveness at various pH values and holding times.

PAGE 67

67 Table 3-1. Initial average mince values for moistu re content, pH, and lipid content. Assessments were conducted on two separate batches of fi sh: Batch 1 (3 day old raw material) and Batch 2 (freshly caught raw ma terial). Lipid content means and standard deviations were calculated from triplicate analysis. Di fferent letters within each row indicate a significant difference (p<0.05). Batch 1 May3days old Batch 2 (September)Fresh Moisture Content (%) 79.251a 77.76a pH 6.49.045a6.49.015a Lipid Content (%) 7.5.12a2.65 0.07b 0 10 20 30 40 50 60 70 80 90 100 01234 Time (days)TBARS (mol MDA/ kg tissue) Batch 1 Mince Batch 2 Mince Batch 2 Fillet Figure 3-1. Lipid oxidation in Spanish Mackerel mince as assessed by thiobarbituric acid reactive substances (TBARS). Experiment s were conducted on two separate batches of fish: Batch 1 (3 day old raw material re ceived in May) and Ba tch 2 (freshly caught raw material). Each sampling point is repres entative of a TBARS mean with standard deviation bars from triplicate analysis.

PAGE 68

68 0 2 4 6 8 10 12 024681012 Time (hours)PV (mmol LHP/kg tissue) pH 2.5 pH 3.0 pH 3.5 pH 6.5 pH 10.5 pH 11 pH 11.5 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 2.5 pH 3 pH 3.5 pH 6.5 pH 10.5 pH 11 pH 11.5 0 2 4 6 8 10 12 024681012 Time (hours)PV (mmol/kg homogenate) pH 2.5 pH3.0 pH 3.5 pH 6.5 pH 10.5 pH 11.0 pH 11.5 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 2.5 pH 3 pH 3.5 pH 6.5 pH 10.5 pH 11 pH 11.5 Figure 3-2. Development of lipid hydroperoxi des and thiobarbituric reactive substances (TBARS) in pH treated homogenates. Pane ls A and B represent lipid hydroperoxides and TBARS in homogenates made from Batch 1 mince that was freshly prepared. Panels C and D represent lipid hydroper oxides and TBARS in homogenates made from Batch 2 mince that was freshly prepar ed. Each sampling point is representative of TBARS and PV mean with standard de viation bars from triplicate analysis. (B) (A) (D) (C)

PAGE 69

69 0 2 4 6 8 10 12 024681012 Time (hours)PV (mmol LPH/kg homogenate) pH 2.5 pH 3.0 pH 3.5 pH 6.5 pH 10.5 pH 11 pH 11.5 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS ( umol MDA/ kg of homogenate) pH 2.5 pH 3.0 pH 3.5 pH 6.5 pH 10.5 pH 11 pH 11.5 0 2 4 6 8 10 12 024681012 Time (hours)PV (mmol LHP/kg of homogenate) pH 2.5 pH 3.0 pH 3.5 pH 6.5 pH 10.5 pH 11 pH 11.5 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 2.5 pH 3.0 pH 3.5 pH 6.5 pH 10.5 pH 11 pH 11.5 Figure 3-3. Development of lipid hydroperoxi des and thiobarbituric reactive substances (TBARS) in pH treated homogenates. Pane ls A and B represent lipid hydroperoxides and TBARS in homogenates made from Batch 1 mince, after 1 day of storage. Panels C and D represent lipid hydroperoxides a nd TBARS in homogenates made from Batch 2 mince, after 1 day of storage. E ach sampling point is representative of TBARS and PV mean with standard devi ation bars from tr iplicate analysis. (A) (C) (D) (B)

PAGE 70

70 0 2 4 6 8 10 12 024681012 Time (hours)PV (mmol LHP/kg homogenate) pH 2.5 pH 3.0 pH 3.5 pH 6.5 pH 10.5 pH 11.0 pH 11.5 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 2.5 pH 3.0 pH 3.5 pH 6.5 pH 10.5 pH 11.0 pH 11.5 0 2 4 6 8 10 12 024681012 Time (hours)PV (mmol LHP/kg homogenate) pH 2.5 pH 3.0 pH 3.5 pH 6.5 pH 10.5 pH 11 pH 11.5 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 2.5 pH 3.0 pH 3.5 pH 6.5 pH 10.5 pH 11 pH 11.5 Figure 3-4. Development of lipid hydroperoxides (PV) and thi obarbituric reactive substances (TBARS) in pH treated homogenates. Pane ls A and B represent lipid hydroperoxides and TBARS in homogenates made from Batch 1 mince, after 2 days of storage. Panels C and D represent lipid hydroper oxides and TBARS in homogenates made from Batch 2 mince, after 2 days of storag e. Each sampling point is representative of TBARS and PV mean with standard devi ation bars from tr iplicate analysis. (A) (B) (D) (C)

PAGE 71

71 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 3 pH 6.5 (control) pH 6.5 Vit. C (0.1%) pH 6.5 Vit. E (0.1%) pH 6.5 Vit. C (0.1%) + Vit. E (0.1%) pH 6.5 Vit. C (0.2%) pH 6.5 Vit. E (0.2%) pH 6.5 Vit.C (0.2%) + Vit. E (0.2%) 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 3.0 (Acid control) pH 6.5 (Control) pH 3 Vit C. (0.2%) pH 3 Vit E. (0.2%) pH 3 Vit C. (0.1%) + Vit E. (0.1%) 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS (umol MDA/kg homogenate) pH 3.0 pH 6.5 pH 3 0.1% C pH 3 0.1% E pH 3 Vit. C (0.1%)+ Vit E (.1%) pH Vit. C (0.2%) pH 3 Vit. E (0.2%) pH 3 (0.2%E+0.2%C) Figure 3-5. Effect of ascorbic acid and tocophe rol at 0.1% and 0.2% w/w on lipid oxidation in homogenates. Panel A represents pH 3 ho mogenate from Batch 1, after 1 day of storage. Panel B represents pH 6.5 homogena te from Batch 1, after 1 day of storage. Panel C represents pH 3 homogenates equiva lent to Batch 1 quality by using Batch 2, after 4 days of storage. Each sampling point represents the mean and standard deviation, from triplicate analysis, repeat ed twice. (A) (C) (B)

PAGE 72

72 0 1 2 3 4 5 6 7 8 9 10 11 12 024681012 Time (hours)PV (umol LPH/ kg of homogenate) pH 3 pH 3 (0.1% EDTA + 0.1% Citric acid) 0.1% Citric acid pH 3 EDTA 0 10 20 30 40 50 60 70 80 90 100 024681012 Time (hours)TBARS (MDA/ kg of homogenate) pH 3 pH 3 (0.1% EDTA+ 0.1% Citric acid) pH 3 Citric acid pH 3 EDTA (0.2%) Figure 3-6. Effect of EDTA and citric acid at 0.1% w/w on lipid oxidation in pH 3 homogenates. Panel A represents lipid hydroperoxides in Ba tch 1, after 1 day of cold storage.Panel B represents TBARS in homogenates. E ach sampling represents the mean and standard deviation, from triplicate analysis from replicate experiments. (A) (B)

PAGE 73

73 0 5 10 15 20 024681012 Time (hours)PV(mmol LPH/ kg of homogenate) pH 3 pH 3 to 5.5 pH 3 to 7 pH 3 to 5.5 (0.1% Vit. C+ 0.1% Vit.E) pH 3 to 7 (0.1% Vit.C +0.1% Vit.E) 0 20 40 60 80 100 120 140 160 180 200 220 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 3 pH 3 to 5.5 pH 3 to 7 pH 3 to 5.5 (0.1% Vit. C + 0.1% Vit. E) pH 3 to 7.0 (0.1% Vit.. C +0.1% Vit. E) 0 5 10 15 20 024681012 Time (hours)PV (mmol LPH/ kg of homogenate) pH 3.0 pH 3.0 to 5.5 pH 3 to 7 pH 3 to 5.5 (0.1% Vit.C+0.1% Vit. E) pH 3 to 7 (0.1% Vit.C+0.1% Vit. E) pH 5.5 pH 7 0 10 20 30 40 50 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 3 pH 3 to 5.5 pH 3 to 7 pH 3 to 5.5 (0.1% Vit. E+0.1% Vit.C) pH 3 to 7 (0.1% Vit. C+ 0.1% Vit. E) pH 5.5 pH 7 Figure 3-7. Effect of pH readju stment to neutral pH (5.5 and 7.0) antioxidants (0.1% ascorbic acid +0 .1% tocopherol), on developmen t of lipid hydroperoxides (PV) and thiobarbituric reactive substances (TBARS) in pH treated homogenates. Panel A and B represent lipid hydroperoxi des and TBARS in homogena tes made from Batch 1 mince, after 1 day of storage. Panels C and D represent lip id hydroperoxides and TBARS in homogenates made from Batch 2 mince, after 1 day of storage (4C). A control homogenate at pH 3 is also shown. Homogenates were held at pH 3 for one hour prior to readjustment. Each sampling poi nt is representative of the mean with standard deviations from triplic ate analysis, repeated twice. (A) (B) (C) (D)

PAGE 74

74 0 5 10 15 20 024681012 Time (hours)PV(mmol LPH/ kg of homogenate) pH 11 pH 11 to 5.5 pH 11 to 7 pH 11 to 5.5 (0.1% Vit. C+0.1%Vit. E) pH 11 to 7 (0.1% Vit. C+0.1% Vit. E) 0 20 40 60 80 100 120 140 160 180 200 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 11 pH 11 to 5.5 pH 11 to 7 pH 11 to 5.5 (0.1% Vit. C+0.1% Vit. E) pH 11 to 7 (0.1%Vit. C+0.1% Vit.E) 0 5 10 15 20 024681012 Time (hours)PV (mmol LPH/ kg of homogenate) pH 11 pH 11 to 5.5 pH 11 to 7 pH 11 to 5.5 (0.1% Vit. E+0.1% Vit. E) pH 11 to 7 (0.1% Vit.C+0.1%Vit. E) pH 5.5 pH 7 0 10 20 30 40 50 024681012 Time (hours)TBARS (umol MDA/ kg of homogenate) pH 11 pH 11 to 5.5 pH 11 to 7 pH 11 to 5.5 (0.1% Vit. C+0.1% Vit. E) pH 11 to 7 (0.1%Vit.C+0.1% Vit.E) pH 5.5 pH 7 Figure 3-8. Effect of pH readju stment to neutral pH (5.5 and 7.0) antioxidants (0.1% ascorbic acid +0 .1% tocopherol), on developmen t of lipid hydroperoxides (PV) and thiobarbituric reactive substances (TBARS) in pH treated homogenates. Panel A and B represent lipid hydroperoxi des and TBARS in homogena tes made from Batch 1 mince, after 1 day of storage. Panels C and D represent lip id hydroperoxides and TBARS in homogenates made from Batch 2 mince, after 1 day of storage (4C). A control homogenate at pH 11 is also shown. Homogenates were held at pH 11 for one hour prior to readjustment. Each sampling poi nt is representative of the mean with standard deviations from triplic ate analysis, repeated twice. (C) (B) (A) (D)

PAGE 75

75 CHAPTER 4 THE EFFECTS OF THE ALKALI/ACIDIC AI DED ISOLATION PROC ESS ON LIPID AND PROTEIN OXIDATION Introduction Oxidative processes inherently occur during handling and/or storage of muscle foods. Muscle foods are particularly prone to oxidative degradation due to proportionally high levels of unsaturated fatty acids, metals (heme bound or free) and multitudes of other catalysts and/ or propagators ( 80 ). The actual mechanisms by which oxidative processes change muscle food proteins are not well understood. Increasing interests in the acid/alkali aided isolation process warrant study of quality changes in muscle tissue during th e process and the final isolated protein. According to various studies, protein oxidation is implicated in va rious protein functionali ty changes: gelation, emulsification, viscosity, solubility, and hydration ( 78, 114, 184-186 ). Various physical and chemical protein changes are attributed to the reaction of protein with reactive oxygen species (ROS) as well as lipid and prot ein oxidation products, resulting in polymerization, degradation, and formation of other protein complexes ( 102, 124, 125, 130, 133, 187-191 ). The main intent of the acid/alkali aided isola tion process is to obtain quality isolates from underutilized fish sources, namely pelagic sp ecies, by-catch, and fish byproducts. The majority of these underutilized sources have significant levels of dark musc le (red muscle). Dark muscle is particularity difficult to work with due to its metabolic nature. Dark muscle is heavily reliant on oxidative metabolism, and relative to white muscle, depends on higher concentrations of heme proteins for increased blood flow, mitoc hondrial function, and lipid s for sustained energy ( 192 ). Although, much of these problematic cons tituents are removed during the isolation process, residuals can pose difficultie s in obtaining a high quality isolate ( 5 ). To date, studies addressing the simultaneous effects of extreme pH and/or readjustment to neutral pH on protein

PAGE 76

76 and lipid oxidation have not been conducted. The objectives of this study, using Spanish mackerel as a model pelagic species, were to 1) investigate how the extent of lipid oxidation in low or high pHadjusted homogenates affects lip id oxidation development in the final Spanish mackerel protein isolate, and 2) investigate how lipid oxidation in acidic/alkali pH adjusted homogenates and in the final protein isolate relate to pr otein oxidation and functionality. Information regarding overall func tionality and quality of isolates obtained from the acid/alkali aided separation is important for improvement and successful implementation of this process industrially. Materials and Methods Raw Materials Spanish mackerel was freshly caught during the fall season (September) in Cedar Key, Florida. The freshly caught Spanish mackerel corresponds to Batch 2 (Materials and Methods, Chapter 3). Fish were manually skinned, degutte d, and filleted. One half of the fillets were minced using a large scale grinder (The Hoba rt MFG. Co., Model 4532, Troy, Ohio). Fillets and ground fish (100 gram quantities) were vacuum p acked and stored at -8 0C until needed. The initial average moisture a nd lipid content and pH were 77.76 %, 6.49.015, and and 2.65 0.07, respectively (Table 3-1). Chemicals Tricholoracetic acid, sodium chloride, sodium tripoly phosphate chloroform (stabilized with ethanol), methanol, ammonium thiocyanate, iron sulfate, barium chloride dihydtrate, 12 N hydrochloric acid were obtained from Fischer Scie ntific (Fair Lawn,New Jersey). Thiobarbituric acid, 1,1, 3,3 tetraethoxypropane, streptomycin sulfate SigmaMarker molecular weight standard, and EZ-Blue stain solu tion were obtained from Sigma Ch emical Co., (St. Louis, MO). Propyl gallate and (ethlenediaminetetraacetic acid 99%), EDTA, TRIS, Potassium Chloride

PAGE 77

77 (KCL) were obtained from Acros Organics (New Jersey, USA). Pre-cast linear 4-20% Tris-HCL gradient gel, Laemmli buffer, and B-mer captoethanol, Power Npac 300 power supply were obtained from Bio-Rad laborator ies (Hercules, CA). The Coomassie PlusBetter Bradford Assay kit was obtained form Pierce (Rockford, Il ). The Zentech PC Test -Protein Carbonyl Enzyme ImmunoAssay Kit was obtained from Biocell Corporation. LTD.(New Zealand). Storage studies and Preparation of Homogenate and Isolate Frozen, minced Spanish mackerel (-80C) wa s thawed in vacuum bags under cold running water and was immediately placed in zip-lock ba gs at 4C for 4 days. From preliminary studies, storing freshly caught fish for four days would mimic industrial fish quality for the isolate process. In addition, raw material storage w ould ideally allow for detectable measurement changes in both lipid and protein quality. Homogenates were prepared by diluting the mince with cold deionized water (1:9) and mixing (45 second at 4C, speed 3) usi ng a conventional kitchen blender equipped with a rheostat (Staco Inc., Dayton, Ohio, 3p 1500B). The pH of the homogenate was approximately 6.2; which is lowe r than homogenate prepared from fresh mince. Homogenates were slowly adjusted to pH 3, pH 6.5 (control), or pH 11 through dropwise addition of 2 N HCL or NaOH. After a specif ied holding time (1 to 8 hrs) at 4C, the homogenates were adjusted to pH 5.5. Th e homogenates were centr ifuged at 10,000 g for 20 minutes at 4C to collect the pr ecipitated proteins. Prot ein isolates kept on ice during analysis were assessed for moisture, pH, and protein content prior to analysis. Isolates were then held for 3 days in one gallon Ziplock polyethylene bags at 4C and we re analyzed again. Thiobarbituric Acid Reactive Substances (TBARS) TBARS were assessed according to Lemon ( 193 ) as modified by Richards and Hultin ( 43 ). Briefly, one gram of sample was homogeni zed with 6 mL of trichloracetic acid (TCA) solution composed of 7.5% TCA, 0.1% propyl gallate, and 0.1% EDTA dissolved in deionized

PAGE 78

78 water. The homogenate was filtered through What man No. 1 filter paper. The TCA extract (2 mL) was added to 2 mL of thiobarbituric acid (TBA) solution (0.23% dissolved in deionized water). The TCA and TBA solutions were freshly prepared and heated if needed. The TCA/TBA solution was heated for 40 minutes at 100C. Th e samples were centrifuge d for 10 minutes (4C) at 2,500 g and were read at 532 nm using a spectrophotometer (Agilent 8453 UV-Visible, Agilent Technologies, Inc., Palo Alto, CA). A standard curve was constructed using 1,1, 3,3 tetraethoxypropane ( 193 ). Protein Quantification and Characterization Protein isolate concentrati ons were assessed using a modified Biuret method ( 194 ). The Bradford method was used to quantify proteins leve ls, either in low concentration (<1 mg/mL) or limited quantity ( 195 ) using the Coomassie PlusBetter Bradford Assay kit. Protein composition in isolates was analyzed with el ectrophoresis. SDS-PAGE (sodium dodecyl sulfate polyacrylamide gel electrophoresis) ( 196 ). Protein concentrations we re adjusted to 3 mg/ mL. SDS-PAGE samples were prepared by comb ining 0.33 L of protein sample, 0.66 L of Laemelli buffer, and 50 L of Bmercaptoetha nol (1mg/ mL protein concentration). Samples were boiled for 5 minutes in 1.5 mL cryogenic tube s and were then cooled on ice. To elucidate changes in disulfide bonds, electrophor esis samples were prepared without -mercaptoethanol using 50 L of Laemmli buffer instead. Pre-cast lin ear gradient 4-20% acrylamide gels were run for 1 h at 200 V; mA 120 with a Power Npac 300 power supply. Samples were applied to the gels in 10 and 15 L aliquots. Molecular we ight standards (205,0006,500 Mw) were run in 5 and 15 uL quantities. Gels were set with 12% TC A for 1 h and were stained with EZ-Blue stain solution (minimum of 1 hr with agitation). Gels were then destai ned with deionized water for 18 hrs with agitation, changing out the water numerous times.

PAGE 79

79 Rheology To compare gel quality among isolates, osci llatory rheology was conducted using an AR 2000-Advanced Rheometer (TA Instruments, New Castle, DE). Viscoelastic behavior and thermal stability of the protein isolates were ev aluated and characterized according to Kristinsson and Demir ( 13 ). Prior to analysis, protei n isolates were adjusted to an average moisture content of 85% through addition of deionized water. Sodium tripolyphoshate (0.3%) and sodium chloride (600 mM) were added to the moisture adjusted isolates according to Liang and Hultin ( 197 ) Isolate additives were incorporated using a mortar and pestle. Samples were adjusted to pH 7.2 through dropwise addition of 1 N NaOH. Fr eshly prepared gels were analyzed in duplicate using an oscillating acry lic plate (40 mm). Approximate ly 20 grams of sample were place on the acrylic plate (5C). The gap between the head and acrylic plate was set to 500 m. After the gap was reached, access sample was carefu lly removed with a stainless steel spatula. Two temperature ramps were used to measure ge l forming ability. A heating ramp (5-80C) and a cooling ramp (80-5C) at rates of 2C/ mi n. Oscillation procedures were conducted using 0.01 maximum strain and an angular frequency 0.6284 rad/s. Changes in the storage modulus (G) as a function of temperature were analyzed. The st orage modulus is indicat ive of elasticity ( 198 ). Protein Oxidation Protein carbonyl content was measured for the pH derived isolates using the Zentech PC Test (Protein Carbonyl Enzyme Immuno-Assay K it). Briefly, protein ca rbonyls are derivatized with dinitrophylhyrazine (DNP). The DNP bound proteins bind to the Elisa plate. The plate bound DNP is then reacted with an anti-DNP-bio tin antiobody. A stretavidin-linked horseradish peroxidase is then bound to the complex. The ch romatin reagent (contain s peroxide) is then added. The addition of peroxide initiates 3,3,5,5-tetramethyl benzidine (TMB) oxidation and thus a color reaction. The appli cation of the stopping reagent (an acid) terminates the reaction.

PAGE 80

80 The yellow chromophore is measured at 450 nm in a microplate reader (Molecular Device Corporation, Sunnyvale, CA). A standard curve is constructed using stan dards provided in the kit, concentration (nmol/g) agains t absorbance (450 nm). The inter-a ssay variation was within kit protocol recommendations (<15%). Prior to analysis, all sample s (0.5 g) were homogenized 1:1 with 20 mM Tris-HCL, pH 7.5, 0.5 M KCL buffer and a glass homogenizer. This buffer rather than water was used to maintain pH and increase isolate solubility according to Kristinsson and Hultin ( 199 ). Connective tissue and/or insoluble components were centrifuged out at 2,000 g prior to analysis. A glass homogenizer versus a motorized tissue tearer was used to reduce foaming. All samples were normalized to 1 mg/mL and dilutions were performed when necessary using the Enzyme Imm uno Assay (EIA) kit buffer. All sa mples and standards were run in at least triplicate. Analysis of Moisture and pH Moisture content was assessed using a moistu re balance (CSI Scientific Company, Inc., Fairfax, VA). The pH was measured using an elec trode ( Thermo Sure-Flow electrode (Electron Corp., Waltham, MA) and pH meter (Denver Instrument, model 220, Denver, CO). The homogenate samples were directly measured for pH. Isolates samples were diluted with deionized water (1:10) to measure pH. Statistical Analysis The SAS system was used to evaluate data. A two-way ANOVA model was used to evaluate statistical significance in lipid oxidation (TBARS, PV), and protein oxidation, with time and among pH treatments as factors. All analyses were performed in at least duplicate and all experiments were replicated twice. Least s quare means were used for post hoc multiple comparisons with Tukey adjustment controlling for Type I error.

PAGE 81

81 Data were reported as mean standard devi ation. The analysis of variance and least significant difference tests were conducted to identify differences among means, while a Pearson correlation test was conducted to determin e the correlations among means. Statistical significance was declared at p < 0.05. Results Initially, to ensure that the quality of fish was approximately equi valent to studies in Chapter 3, Spanish mackerel batch 2 mince was st ored at 4C for four days. Homogenates were subjected to pH 3 as per findings in Chapter 3, in which the greates t overall levels of homogenate lipid oxidation with respect to pH occurred at pH 3 and 3.5. Because the overall levels of lipid oxidation with respect to homogena tes 3 and 3.5 were insignificant from each other, and the fact that pH 3 is an intermediate of pH values ( 2.5, 3, and 3.5) tested in Chapter 3, pH 3 was chosen as the primary acid treatment. Likewise, no si gnificant differences in lipid oxidation were observed among homogenates at alkaline pH of 10.5, 11, or 11.5, so the intermediate pH 11 was chosen as a treatment for this study. This pH is commonly used in the alka line process. Prior to experimentation, moisture content and pH of the starting homogenate were 97.6.707% and 6.215.077. Oxidative Stability of Acid/Alk ali Derived Protein Isolates Levels of lipid oxidation were assessed th roughout the isolation pr ocess using one and eight hour exposure times at either pH 3 or pH 11. The isolates were obtained through centrifuation at pH 5.5. The isoel ectric point of fish proteins is close to pH 5.5, enabling for lowest repulsion and highest yield of extracte d proteins (mainly of myofibrillar nature)( 17, 200, 201 ). Isolates were assessed for oxidation right after is olation and after 3 days of 4C storage. Average moisture and protein contents of acid, al kaline, and control (pH 6.5) isolates from one and 8 hour hold times were similar and are summarized in Table 4-1.

PAGE 82

82 The first centrifugation step (Figure 2-2) was omitt ed in this experiment to increase protein isolate yield and reta in more lipids ( 13 ). Kristinson and Demir ( 13 ) observed a gel-like sediment in the soluble phase following the first centrifuga tion of acidified mullet. Similarly, a gel-like sediment was observed after readjusting Span ish mackerel homogenates from extreme and neutral pH to pH 5.5 (Figure 4-1). This pha se was grayish white, soft, slimy, and easily disrupted. Unlike the sediment obs erved by Kristinsson and Demir ( 13 ), the gel like sediment was found above and adjacent to th e collected proteins. Despite differences in processing, this sediment was observed. This sediment, mainly neut ral lipids, was removed with a plastic spatula. When comparing the two processes, it is evident th at more lipids are retain ed as exhibited by the Spanish mackerel sediment layer when th e first centrifugation step is excluded. Holding times (1 or 8 h) did not significantly affect lipid oxidati on in homogenates (See Figures 4-2,3,4). The supernatant from the isola tion of homogenate at pH 3 for eight hours had the greatest level of TBARS (mol MDA/kg of ti ssue) compared to all other pH homogenates, isolates, and supernatants at pH 3, 6.5, and 11 (Figure 4-3).The levels of TBARS in the supernatant (Figure 4-3) held at pH 3 for 8 h ( 94 ) were significantly (p<0 .001) greater compared to the corresponding isolates, or 52.59 5.48 and 31.17.97 umol M DA/kg tissue, respectively. The level of lipid oxidation in the pH 3 isolat e after 1 hour storage was significantly (p <0.03) less than the corresponding supernatant. The amo unt of oxidation measured in the pH 11 derived isolate, after an 8 hour holding ti me, was significantly less than th e same isolate stored for three days (p<0.0291) (Figure 4-3). Levels of oxidation in pH 11 derived isolate (Figure 4-4), after 8 hours of holding, were significantly lower than the corresponding isolate at pH 6.5 (p<0.0289). TBARS levels measured in the pH 6.5 supernat ant, after an 8 hour hold, were significantly greater than TBARS levels in isolate made after an 8 hour hold at pH 11 (p<0.0053).

PAGE 83

83 Significantly (p<0.0068) higher TBARS formed in the homogenate exposed to pH 3 for one hour compared to the pH 11 isolate, derived from an 8 hour homogenate holding time (Figure 4-3). As shown in Figure 4-3 and 4-4, isolate from the pH 11 treatment obtained after 8 hours of holding had significantly greater levels of oxidation after 3 days of storage than the corresponding freshly prep ared isolate (p <0.03). Levels of protein oxidation were assessed among the isolates after 1 hour, 8 hours, and 8 hours followed by three days of storage (See Figu re 4-5). Overall levels of oxidation were significantly different (p<0.0001) with respect to pH treatment: pH 11> pH 6.5> pH 3 and isolate: 1 hr> 8hr> 8hrs+ 3days. Specifically, isolate from the pH 11 treatment (1 h holding) contained significantly highe r levels of protein oxidation compar ed to all other combinations of isolate and pH, with exception of these isolates from the 6.5 treatment (1 h holding) and the pH 11 treatment (8 hr holding) (Figure 4-5). The isolate produced from holding the 6.5 homogenate for one hour, was significantly higher in carbonyls than the corresponding isolate at pH 11 (p< 0.0031). The isolate formed from a holding ti me of 8 hours at pH 11 had significantly (p<0.0001) greater levels of pr otein oxidation than the corres ponding pH 11 and pH 6.5 isolates (p <0.0006) formed from a one hour holding time. The pH 11 isol ate derived from an 8 hour holding time was more oxidized than the corres ponding pH 6.5 isolate (Figure 4-5). The pH 6.5 isolate carbonyl levels from the 8 h holding trea tment were significantly less than carbonyls after 1 hour of holding. After the isolates from the 8 hr holding treatment were stored for three days, a significant decrease (p<0.03) in protein carbonyls compared to the fresh 8 hour isolates was observed among pH 11and 6.5 treatments (Figure 45). Mean values were comparable to an average carbonyl content of 1.5 nmol /mg of protein by Liu and Xiong ( 189 ) in red and white muscle of chicken.

PAGE 84

84 Gel Forming Ability of Isolates In terms of rheological differences, the storage modulus (G) of isolates was not significantly affected by holding time or storage; however, the elasticity (G) of isolates as significantly (p 0.0002) affected by pH treatment with pH 11 isolate having the highest G (Pa), followed by pH 6.5, and then pH 3 (Figures 4-6). Examples of rheograms generated during the os cillatory testing of is olate pastes derived from 1 hour, and isolates derived from 8 hr pH holding times in addition to three days of refrigerated storage are shown in Figures 4-7,8,9. Initial G values were similar in the pH 11 and 6.5 isolates. These initial G values were greater th an G for pH 3 isolates. Initial increase in the storage modulus (G) on heating were observed at 30C for all samples at pH 3.0, 6.5 (control), and pH 11 (Figures 4-7,8,9). During heating, on e noticeable transition temperature (Tm) was observed at 40C for the isolates at pH 6.5 (control) (Figure 4-9). The same Tm, although significantly less noticeable, was also observed for pH 3.0 and pH 11 isolates (Figure 4-7,8). Significant increases in G were observed between 50 and 60C for isolates derived from pH 3 and 6.5 homogenates and between 40 and 55C for isolates derived from pH 11 homogenates (Figures 4-7,8,9). G values were stabilized be tween 70 an 80C. Significant increases in G during the cooling step were observed in all isol ates. During cooling, the greatest increase in G occurred in the pH 11 isolates, fo llowed by pH 6.5 (control), and pH 3. Protein Composition of Homogenates and Recovered Isolate Fractions SDS Page, with and without, mercaptoethanol showed differences in protein composition among isolate fractions. The inclus ion versus the exclus ion of mercaptoethanol in samples was applied to the gels to assess whether differen ces among protein fractions if any, were due to covalent (S-S) bonds or non-cova lent cross-linking, po ssibly attributed to protein oxidation. Cross-linking was not observed in any of the sa mples; however, hydrolysis was observed for all

PAGE 85

85 pH (11,6.5, 3). Hydrolysis was evident in is olates derived from adjusting and holding homogenates at acidic (3), alka line (11) or control (6.5) pH fo r 1 hour, followed by readjustment to pH 5.5 in Figure 4-10: Lanes 2-5 (pH 11); La nes 10 and 11 (pH 6.5); Lanes 14 and 15 (pH 3). Hydrolysis was also observed in isolates derived from adjust ing and holding homogenates at acidic (3), alkaline (11) or c ontrol (6.5) pH for 8 hour, followe d by readjustment to pH 5.5 in Figure 4-11: Lanes 4 and 5 (pH 6.5); Lanes 8 and 9 (pH 3); and Lanes 13 and 14 (pH 11). Proteolysis was evident in isolates derived from adjusting and holding homoge nates at acidic (3), alkaline (11) or control (6.5) pH for 8 hours, followed by readjustment to pH 5.5, and finally followed by 3 days of refrigerated storage in Fi gure 4-12: Lanes 4 and 5 (pH 6.5), Lanes 6 and 7 (pH 11); Lanes 14 and 15 (pH 3). Hydrolysis wa s also evident in pH 3, 11, and (6.5) control homogenates at initial pH adju stment (Figure 4-13) and 1 and 8 h holding (Figures 4-14,15), at pH 3, 6.5 (control), and 11. The most prevalent proteins fractions were myosin heavy chain (MHC)(~205 Kda) and actin ( ~ 43kDa), and tropomyosin (35.5 kDa). Ot her identified proteins included desmin (~56 kDa), troponin T (41 kDa), and topomyosin-beta (~38 kDa). Discussion The isolation of fish proteins following pH r eadjustment from extreme pH is performed for greater protein recovery and to recover proteins with modifi ed functional properties. With greater protein recovery, co-ext raction of unwanted components (l ipids and heme proteins) and consequentially oxidative degr adation could occur, negative ly affecting the quality and performance of the isolated proteins. Changes in pH of Mince with Storage The pH (6.2) of mince stored for 4 days was le ss than initial pH va lues of unstored mince homogenate (6.45) as reported in Chapter 3, indicative of meta bolite accumulati on and cellular

PAGE 86

86 degradation post mortem ( 31 ). Various postmortem changes can influence oxidative occurrence within the fish muscle: decreases in ATP a nd increases in ATP byproducts, loss of reducing compounds (ascorbate, NAD(P)H, glut athione, etc.), increased leve ls of low molecular weight transition metals (copper, iron), oxidation of ir on in heme proteins, disruption of cellular structure, loss of antioxidants, and accumulation of calcium and other ions. With mincing, postmortem muscle tissue can e xperience conditions very sim ilar to ischemia-reperfusion. Consequentially, production of supe roxide and hydrogen peroxide can occur, possibly leading to protein and lipid oxidation ( 5 ). Lipid Oxidation During Extreme pH Adjustment and Readjustment to 5.5 It was observed that through se paration of Spanish mackerel proteins by pH readjustment from extreme pH (3, 6.5, 11) to neutral pH (6.5) that lipids were retained (Figure 4-1). Kristinsson and Demir ( 13 ) reported greater lipid levels in ac id mullet isolates when the first centrifugation step was skipped. This could be explained by the co-preci pitation of membrane and neutral lipids with proteins at isolelectric pH. Co-precip itation is most likely a result of increased protein hydrophobicity and surface ac tivity with pH adjustment as observed by Kristinsson and Hultin ( 18 ) with cod myosin. Increases in hydrophobicity result in enhanced emulsification ability and inter action with hydrophobic residues of lipids and heme proteins. Increases in hydrophobicity is most likely attr ibuted to denaturation of myofibrillar and sarcoplasmic proteins, as supported by low extrac tability of rockfish proteins during the acid/alkali aided process ( 202 ). Kristinsson and Hultin ( 199 ) have shown that significant structural changes occur in myosin upon inco mplete refolding of cod myosin with the readjustment to pH 7.5 from acid and alkaline tr eatment. Partial refolding of myosins tertiary structure could facilita te hydrophobic interactions. Furtherm ore, low levels of retained hemoglobin upon aggregation can pose lipid oxidation problems. Undeland et al. ( 67 ) found that

PAGE 87

87 extensive hemoglobin chelation by 0.2% STPP and 0.2% EDTA during acid isolation of herring still gave rise to low TBARS levels. These resu lts emphasize that despit e significant removal of hemoglobin in the pH cycle, hemoglobin could s till be problematic rega rdless of concentration found in the isolate. The greatest level of TBARS were found in th e supernatant from th e isolation of pH 3 Spanish mackerel homogenate held for eight hours re lative to all other pH homogenates, isolates, and supernatants at pH 3, 6.5, and 11 (Figure 4-2 and 4-4). Additionally, the pH 3 supernatant, after 1 h of holding, was greater than the corr esponding isolate (Figure 4-2). According to Undeland ( 16 ), lipids were presen t in the highest amounts in th e emulsion (supernatant) layer followed by the gel sediment in th e acidic pH (2.7) extr action of herring. Lowe r lipid levels were observed from alkaline extracti on (10.8) in which the emulsion layer had 37% lipid and the isolate had 18% lipid ( 16 ). Just as Undeland ( 67 ) reported, the Spanish mackerel supernatant fraction at acidified pH (3) contained the hi ghest TBARS levels (F igure 4-3). During the conventional production of horse mackerel surimi performed by Eymard et al. ( 138 ), significant TBARS were found in the wash water. Lipid ox idation products formed during acid processing could be water soluble or at least suspendable, which partially explaines why the supernatant fractions (Figures 4-2 and 4-3) had such high TBARS. Overall, lipid oxidation in the pH treated is olates were not signifi cantly different (Figure 4-4). Lack of significant differe ntiation in TBARS between alkali and acid processing compared to other published studies is attr ibuted to skipping the first ce ntrifugation step. Kristinsson and Demir ( 13 ) observed significant differences among co lor values when inco rporating the first centrifugation in the separation of mullet and croaker proteins; howev er, in absence of centrifugation both acid and al kaline isolates were negatively a ffected in color value, suggesting

PAGE 88

88 more retention of heme proteins. With both centrifugation steps, the alkaline treatment was whiter in color (high L value) and was less yello w (low b value) than acid isolate. This was attributed to increased removal of heme pr oteins, found in the supernatant of the first centrifugation step ( 13 ). Kristinsson and Hultin ( 51 ) concluded that alkaline treatment resulted in greater hemoglobin solubility a nd reduced co-precipitation at pH 5.5 as per discussion of homogenate pH adjustment in Chapter 3. From th e lipid oxidation studies, it appears that acidic and alkaline pH processing of Spanish mackerel with exclusion of the first centrifugation is equally problematic. Holding times did not significantly aff ect the amount of oxidation among Spanish mackerel isolates within treatments (Figur es 4-2,3,4). According to Undeland et al. ( 203 ), the holding time at acidic pH (4, 30, or 70 minutes) pr ior to acidic isolation of herring isolates did not affect overall levels of oxi dation. In another study, Kristinsson ( 51 ) found that the prooxidative activity of trout hemoglobin at pH 3.0 and 3.5 versus pH 2.5 was unaffected by holding time prior to pH readju stment. Kristinsson et al. ( 204 ) reported no significance difference in lipid oxidation am ong the alkaline or acidic pH is olation of catfish proteins; however, when the first centrifugation was skippe d, significantly more lipid oxidation resulted relative to the other treatments. The lack of significance among the acid and alkali treatment concurs with some previous findi ngs. The significant difference be tween alkaline and control pH isolates, after 8 hours of holding, is intere sting (Figure 4-4). Kr istinsson and Liang ( 205 ), reported reduced levels of lipid oxidation in an al kaline isolate relative to a surimi control and acid isolates from Atlantic croa ker. As suggested by Kristinsson ( 12 ), this could be a result of hemoglobin adopting a more stable tertiary stru cture upon alkaline treatment compared to acid treatment ( 51 ). Additionally, oxidative products were most likely removed through

PAGE 89

89 centrifugation at 5.5, impart explaining a reductio n in lipid oxidation pr oduct in the alkaline isolate made an after 8 h holding (Figure 4-3). Lipid oxidation was observed among all isolate treatments (Figure 4-4). Similarly, the greatest increase in lipid oxidation during 7 days of storage of b eef heart surimi, washed with buffer at 5.5 and 7.0, was observed in th e 5.5 treatment by Srinivasan et al. ( 136 ). It would be interesting to investigate the effects of readjustme nt to pH 7, as previous studies suggest it could pose less lipid oxidation problems ( 190 ); however, reduced protein recovery would most likely result. With the final protein is olate at pH 5.5, it is to be expe cted that lipid oxidation would proceed significantly faster than at pH 7 as hemoglobin is more pro-oxidative at pH below neutrality as observed in a washed cod model systems ( 43, 51, 206 ). Lipid oxidation in isolates, after storage at low pH (5.5) was evident. This can be explained by rapid autoxidation of heme proteins in the isolates at pH 5.5. According to Richards and Hultin ( 43 ), increases in TBARS were due to decreases in pH from 7.6 to 6.0 in a hemoglobin mediated washed cod system. Protein Oxidation Overall levels of protein oxi dation (Figure 4-5) were high est at pH 11, followed by pH 6.5, and pH 3. The effect of pH on lipid and protei n oxidation has been studied in various meat systems. Similar to observed pH af fects in this study, Xiong et al. ( 124 ) and Wan et al. ( 123 ) have shown that myofibrils from bovine cardiac muscle improved in functionality with decreases in lipid oxidation. In a study c onducted by Srinivasan et al. ( 190 ), beef heart was washed with various pH buffers, ranging from pH 7.0 to 5.5. Afte r seven days of storage, the levels of lipid and protein oxidation were signif icantly higher at pH 5.5 and 6.0 than at pH 7.0. Srinivasan ( 136 ) reported a simultaneous increase in markers of lipid (TBARS, conjugate d dienes) and protein oxidation (carbonyls) with lowering of pH in bovine cardiac surimi. Kenney et al. ( 207 ) found that washed beef heart surimi ha d superior textural pr operties compared to fish surimi. Similar to

PAGE 90

90 dark muscled fish, beef heart contains an abund ance of heme proteins, iron, and polyunsaturated fatty acids ( 208 ). The mince had significantly lower levels of protein oxidation compared to all isolation treatments at pH 3, 6.5, and 11 (See Figure 4-5). Hultin and Luo ( 209 ) reported that protein modifications will occur with cha nges in muscle temperature, sa lt concentration, or pH, leading to significant degradation and polymerization of proteins due to protein oxidation ( 209 ). This phenomenon could explain why the mince had lower levels of oxidation, as the isolates were obtained through pH processing. Inte restingly, carbonyl levels were higher in the isolates made after one hour holding at pH 6.5 and 11 compared to pH 3. According to Srinivasan et al. ( 136 ), bovine heart surimi washed with antioxidants ab ove pH 5.7 produced stronger gels than surimi washed below pH 5.7. The differences in G at various washing pH was a ttributed to oxidative modification of protein residue s and or protein-lipid oxidatio n interactions, affecting the antioxidant efficacy. The observa tions by Srinivasan et al. ( 136 ) could explain the differences in the protein oxidation in Spanish mackerel isolates with pH, as different protein interactions occur in the presence of oxidized lipids or proteins, more so at high pH. Kristinsson and Hultin ( 199 ) reported partial refolding of c od myosin with pH 7.0 readjustme nt from alkaline pH. Partial refolding at high pH could make alkaline treated isolates more su sceptible to protein oxidation. In the pH 11 Spanish mackerel isolates, pr otein carbonyl content si gnificantly increased after 8 hours of holding compared to one hour of holding (Figur e 4-5); while there was an apparent decrease in TBARS after 8 hour hold compared to the 1 hour hold. Similar trends between protein and lipid oxidation were observed by Srinivasan and Hultin ( 210 ). In this study, significant increases in protein and lipid oxidation were observed when washed cod muscle was subjected to free radical exposur e at pH 6.8 for 2 hours. Interes ting kinetics between protein and

PAGE 91

91 lipid oxidation were observed in this system. With storage, prot ein oxidation (carbonyl) levels increased quickly, while there was a lag phase in lipid oxidation. Li pid oxidation began to accelerate at a point where protein oxidation dec lined slightly. Protein oxidation levels then experienced another rapid increase and lipid oxidation levels also increased. Hultin ( 210 ) concluded that perhaps protein or lipid oxidation could propagate e ach other. Similarly, longer pH holding times may afford greater production of ROS thr ough lipid oxidation, providing adequate initiation of protein oxidation ( 211 ). Significant decreases in protei n oxidation after 3 days of st orage occurred among the pH 11 and 6.5 isolates compared to the freshly prepared isolate (Figures 4-5). This is in contrast to other storage studies, in which, protein and lipid oxida tion increased over a duration equivalent to 7 days or more ( 63, 70, 136 ). It is possible that oxidized, aggregated proteins or insoluble proteins could not be adequately assessed via the carbonyl assay, as solubilized proteins are applied to the protein carbonyl kit. With storag e, protein degradation wa s observed in the gels. Protein aggregation may have possibly occured in the isolates as protein bands were observed at the top of the SDS-PAGE gels. It is also possi ble that protein oxidati ve products served as intermediary reactive species during the oxidative process, becoming pro or antioxidive with changes in pH and conformation environment ( 212, 213 ). Protein oxidation products could have served as antioxidants for the al kaline pH and control pH homogenate treatments and as oxidants in the pH 11 derived isolate, st ored at pH 5.5. It has been propos ed that protein radical species are highly dynamic and even transient, depe nding upon interaction with other molecules and environmental conditions ( 213 ). In a BSA hydrogen peroxide activated pr otein oxidation model study conducted by Ostdal et al. ( 211 ), BSA successfully oxidized urate, linoleic acid and a

PAGE 92

92 complex amino acid residue. Thus, protein radicals could play various role s in the oxidation of muscle foods either as endproducts or intermediates. Carbonyl measures are readily performed a nd acceptable measures of complex-muscle system samples ( 112, 210 ). Carbonyls are a global form of oxidation and can be formed as result of various reactive species ( 76 ), unlike other protein oxida tion products. Strict reliance on carbonyl levels as an indicator of protein oxidation is tenuous, as various other compounds form as result of protein oxidation. A more comprehens ive attempt to measure protein oxidation could include the measure of dinitrotyrosine, 3dinitrotyrosine, and modified AA residues ( 214 ). Protein carbonyl content in isol ates would be best analyzed by a method requiring chemical digestion. Chemical digestion would eliminate so lubility problems and possible underestimation of protein oxidation levels, upon analysis and/or plating. Kj aersgard et al. ( 215 ) have successfully used two-dimensional immunoa ssaying followed by mass spectroscopy in the identification and quantification of oxidized proteins and residues in fish species. Gel Forming Ability Noticeable differences in rheograms among the isolates derived from pH 3, 6.5 (control), and pH 11 homogenates were observe d. All isolate pastes showed initial increases in G at 30C. Yougsawatdigul and Park ( 200 ) also observed initial increases in surimi as well as alkaline and acid isolates at 34C ( 200 ). Unlike the acid (pH 3) and alkalin e (pH 11) isolates (Figures 4-7,8), a significant decrease in G wa s observed during the heating step in the pH 6.5 (control) isolates. The lowest G was observed at approximately 47C for isol ates derived from pH 6.5 homogenates (Figure 4-9). Yongsawatdgul and Park ( 200 ), Kristinsson and Liang ( 205 ), and Kristinsson and Hultin ( 18 ) observed similar rheology differences among surimi and pH treated samples of rock fish, Atlantic croaker, and c od muscle proteins, respectively. The Tm (40C) observed in the control (6.5) isol ate was not observed in the pH tr eated isolates. Similarly, Hultin

PAGE 93

93 and Kristinsson ( 18 ) reported absence of this peak in pH treated cod proteins compared to untreated cod proteins, and this observation wa s attributed to diffe rent protein-protein interactions among the control and pH treated proteins ( 18 ). Ingadottir ( 216 ) and Theordore ( 217 ) observed significant rheologi cal differences among surimi and pH treated samples for tilapia and catfish, respectively. These rheologica l differences between c ontrols and acid/alkaliaided isolates may be due to protei n denaturation with pH treatment ( 218 ). In agreement with findings by Kristinsson and Hultin ( 18 ) and Kristinsson and Liang ( 205 ) for washed rockfish and croakers, a large drop in storage modulus occu rred in pH 6.5 (control) isolate samples. According to Kristinsson and Liang ( 205 ), a large decrease in G could be attributed to actyomyosin gel formation. Non-covalent protein in teractions between acti n and myosin are due to various conformational changes. The Tm obser ved at 40C in the pH 6.5 isolates marks the beginning of gel network formation through vari ous conformational changes including myosin light chain dissociation from myosin heavy chains (MHC) ( 219, 220 ), exposure of hydrophobic groups, namely tryptophan, and transformation of the heavy myosin rod (tail) from a helix to a random coil. G significantly increased on cooling among all isolate samples. The pH treatment had a significant affect on gel st rength, as alkaline pH gave stronger gels than the control (6.5) an d the acid (pH 3) prepared samples (F igure 4-6). This is in disagreement with previous findings by Kristin sson and Liang, in which croake r alkaline and acid isolates had greater G than surimi. Kristinsson and Hultin ( 18 ) reported similar G values among acid (pH 2.5), alkaline (ph 11), and untreated (pH 7.5) cod muscle proteins; however, different mechanisms of gelation among the treatme nts were observed. Shikha et al ( 221 ) studied the effect of acid pH shifting in wa lley pollack. Surmi that was acidified and then readjusted to neutral pH had lower gel strength (g/cm2) than the control (pH 7.2 mi nce). Various other studies

PAGE 94

94 reported difference in G values among alkaline, acid, and untreated sample s for catfish, tilapia, mullet, and Spanish mackerel ( 13, 216, 218 ). Theodore ( 217 ) reported G trends comparable to the Spanish mackerel isolate samples, in which the alkali treated catfish samples were followed by surimi and then the acid samples in G valu es. In agreement with this study, Kristinsson and Demir ( 13 ) found that alkaline processing of Spanis h Mackerel and mullet gave greater G values than surimi processing, followed by acid processing. Higher G in the alkaline samples indicates greater protei n-protein interactions on cooling, en abling better gel-network formation upon setting. According to Kristinsson and Hultin ( 199 ), proteins can unfold through extreme pH and upon readjustment only partially refold. Specific changes in protein structure with unfolding and refolding could expl ain the differences in G among the isolates and control (pH 6.5) treatment. At low and high pH, proteins unfold and denature. Kristinsson and Hultin ( 199 ) reported dissociation of the cod myosin head group with unfolding at extreme pH. With readjustment to pH 7.5, the alkaline and acid trea ted myosin partially refolded; however, alkaline treated cod myosin remained more dissociated. Due to more flexibility than acid isolates upon refolding, alkaline treated proteins could form a better gel network. Incr eased exposure of thiol groups could also explain the difference in G among the alkaline (pH 11), control (pH 6.5), and acid isolates (pH 2.5). Kristinsson and Hultin ( 199 ) found an increases in thiol exposure when cod myosin was adjusted to alkaline pH and re adjusted to pH 7.5. This increase in sulfhydryl group (-SH) reactivity could explai n the higher G in the Spanish mackerel alkaline isolates versus the control (pH 6.5) and acid isolates (Figure 4-6). The acid isolates did show an increase in sulfhydryl reactivity with refolding according to Kristinsson and Hultin ( 199 ), but this increase was less compared to alkaline refoldi ng. A lower G in the acid versus alkaline processing could be due to less thiol interacti on among proteins.

PAGE 95

95 Textural Changes and Prot ein and Lipid Oxidation Greater overall levels of protei n oxidation (Figure 4-5) in th e alkaline treatments could explain high G (Figure 4-6). Protein oxidation could lead to te xtural changes in muscle food products ( 86 ), and could explain the differences in ge l strength among isolates. As a consequence of oxidation, proteins unfold and exposes th eir hydrophobic residues. Srinivasan and Xiong ( 130, 136 ) reported extensive sulfhydryl loss with protein oxidation. This increase in hydrophobic affinity could promote protein intera ction and aggregation, improving gel strength. Protein oxidation proceeds in a manner compar able to lipid oxidation in the presence of oxidizing lipids and/or other reactive oxygen species (ROS) ( 101, 222 ). Free radicals will react with either amino acid ( 21 ) side chains and/or peptide backbones. Proteins will often fragment and complex upon free radical attack ( 86 ). Hydroxyl groups are predominantly responsible for protein oxidation in the absence of lipids ( 106 ). Certain AA groups are preferentially oxidized, the most susceptible is cysteine. Myosin, the most abundant and importa nt muscle food protein, has 40 free thiol groups. ( 223 ). The abundance of myosin compared to sarcoplasmic proteins ( 224 ) warrants investigation of how myosin is affected during th e acid/alkali aided process ( 13 ). Protein carbonyls are product of: AA residue oxidation, peptide backbone degradation, proteinreducing sugar interaction, and interaction with nonprotein carbonyl groups ( 86 ). Oxidized proteins will nucleophilically att ack proximal proteins, forming cross-linkages. According to Xiong and Decker ( 87 ), protein cross-linkages result in pr otein polymerization, aggregation, and insolubility. Cross-linking ma y cause protein aggregation and increased G values ( 70 ). The quality of muscle foods is greatly impact ed by these physiochemi cal protein changes. Intermolecular interactions may be weak with protein oxidatio n; however, oxidatively induced polymerization may counteract these weak bonds and strengthen non-covalent interactions, through increased rigidity and low mobility within the gel network.

PAGE 96

96 The duration of storage for the Spanish mackerel isolates was short and the development of carbonyls did not exceed 2 nmol/mg (Figure 4-5). According to Lee et al. ( 104 ) and Nishimura et al. ( 225 ) there is a very high, positi ve correlation (r=0.99) in b eef heart surimi, between protein carbonyls and G when there is sufficien t protein oxidation and st orage. It was observed that increases in G occurred with increas es in protein carbony ls to at least ( 136, 190, 191 ). According to other studies ( 136, 190, 191 ), protein oxidation improved gel forming ability in muscle tissue through greater gel network formati on. In contrast to thes e studies, the level of oxidation was significantly less 15 nmol/mg protein. W ith increased storage, a greater correlation between G and protein oxida tion as well as lipid oxidati on may have been observed. Lipid oxidation in the alkaline isolate after 8 h holding was significantly less than with 1 hour holding (Figure 4-4). Protein oxidation levels also signifi cantly increased after 8 hr of holding compared to 1 hr of holding (Figure 4-5). A possible correlation between protein oxidation and lipid oxidation and G in the isolates exists. As per discussion in prev ious Protein oxidation section, protein oxidati on could be propagated by lipid oxidation. According to Saeed and Howell ( 70 ), the elastic modulus in Atlantic Macker el fillets stored at -20C and -30C was significantly decreased with the incorporation of vitamin E (500 ppm). Because the transfer of free radicals or protein-lipid interacti ons could proceed from lipid oxidation ( 70 ), it was inferred that vitamin E quenched lipid radicals and reduced protein oxidation formation. This study elucidates that lipid oxi dation is an initiator of protein oxidation. Protein oxidation did not proporti onally increase with lipid oxidation (F igure 4-4, 4-5) in the isolates. This is in contrast to other st udies in which increases in protein oxidation are proportional to increases in lipid oxidation. According to Smith ( 122 ), decreases in myofibrillar gel strength were observed in deboned turkey w ith increases in TBARS, and decreases in

PAGE 97

97 solubility and myosin ATPase activity ( 122 ). The authors attributed the decrease in solubility and ATPase activity to protein oxidation. The oxidativ e effects of the acid/alkali aided process on lipid and proteins has received little attention and deserv es further investigation. SDS page Electrophoretic patterns (Figures 4:7-12) revealed hydrolysis in alkaline and acid samples. Hydrolysis is a concern as it could negatively affect gel stre ngth due to gel softening ( 226, 227 ). Several studies suggest that proteolysis duri ng pH processing can result in poorer gels. According to various studies, the alkaline/acid-a ided process can initia te protease acitivity ( 16, 201, 228 ). Enzymes can be activated at low pH and at the isoelectric pH 5.5 ( 201, 229 ). Similar to this study, hydrolysis was observed in aci d/alkali aided samples by Kristinsson and Demir ( 13 ), Choi and Park ( 16, 201 ), Undeland et al. ( 36 ) and Ingadottir ( 216 ) in catfish, spanish mackerel, mullet, croaker, pacific whiting, herr ing, and tilapia. A band of approximately 150 KDA was observed in the gels (4-10,11,12, 13, 14, 15), similar to other studies analyzing the affects of pH processi ng on fish proteins ( 205, 230-232 ). Myosin heavy chain (MHC) degradation is indicat ed by this band ( 230 ). The observed hydrolysis among the Spanish mackerel isolates is most likely attributed to protease activation versus alkaline or acid adjustment as the control (pH 6.5) also contai ned hydrolysis. Because proteolytic hydrolysis was observed among all treatments and it did not negativ ely affect the gel stre ngth of either alkaline or control treatments, hydrolysis could not explain the differences in gel st rength in the alkaline, acid, or neutral isolates (Figure 4-6). SDS page results do therefor e suggest that the differences in gel strength are impart due to changes in comformation with pH tr eatment, or these are chemically induced changes linked to oxidati on. With storage, protein aggregation or degradation was not observed in th e gels; however, it is possible th at aggregates were too large to enter the gel.

PAGE 98

98 Conclusion Protein oxidation and lipid oxi dation were significantly aff ected by holding time and pH. The pH 11 isolate contained the highest overall levels of protein oxidation, followed by pH 6.5, and pH 3. Overall, levels of lipid oxidation were similar among the pH treatments, and significant removal of lipid oxida tion occurred with centrifugation at pH 3, after an 8 h holding. A positive correlation between lipid oxidation an d protein oxidation was not apparent, as observed in other studies. G was consistently higher in alkaline isolates compared to pH 6.5 and pH 3. Increases in G may be attributed to conf ormational changes at higher pH (6.5 and 11); and or protein modifications due to protein oxidation. As research currently exists positive correlations between lipid and protein oxidation and gel forming ability can only be tentatively made The study of protein oxidation in muscle foods has only just begun and further find ings are necessary to substantia te the influence of lipid and protein oxidation on protein physio chemistry and functionality.

PAGE 99

99 Table 4-1. Moisture content (%) and protein conten t (%) of isolates obtain ed after readjustment from pH 3, 11, or 6.5 to pH 5.5, and held for 1 and 8 hours. Moisture and protein content means and standard deviations we re calculated from triplicate analysis. Isolate 1 (1 hour hold) Isolate 2 (8 hour hold) Control (pH 6.5 5.5) Acid (pH 3 5.5) Alkaline (pH 11 5.5) Control (pH6.5 5.5) Acid (pH 3 5.5) Alkaline (pH 11 5.5) Moisture Content (%) 81.3 .353 82.5 .41 81.5 .707 80.8 .354 80.05 .20 81.3 .707 Protein Content (%) 10.9 2.02 12.3.198 12.29.505 10.4.52 12.3.500 11.8.940 Figure 4-1. Schematic of the observed centrifugati on layers with direct readjustment from extreme pH (3 and 11) and neutral pH (6.5) to pH 5.5. Centrifugation prior to readjustment was not employe d. This sediment layer was skimmed off with a plastic spatula. Supernatant: neutral lipids, saturated fatty acids, and ( sarco p lasmic ) p roteins Sediment: Connective tissue, membrane and neutral lipids Protein Isolate: primarily myofibrillar proteins, heme proteins, and other precipitated/ aggr egated proteins

PAGE 100

100 a bc abc abc ab abc abc abc abc c abc abc0 10 20 30 40 50 60 70 intial homogenatehomogenate after one hour isolatesupernatant Isolation TreatmentsTBARS (umol MDA/kg tissue) pH 3 pH 6.5 pH 11 Figure 4-2. Lipid oxidation (TBARS ) during the production of acid (pH 3), alkaline (pH 11) and control (pH 6.5) derived is olates. Homogenates were exposed to the given pH values for one hour prior to obtaining an isolate and corresponding isolates. Each bar in the graph is representative of TBARS (umol MDA/kg of tissue) with standard deviations. Different small letter s indicate a significant difference. bc a bc b bc b b b bc abc b bc c bc bc0 10 20 30 40 50 60 70 intial homogenate homogenate after 8 hours isolate supernatant8 hr + 3 days Isolation TreatmentsTBARS (umol MDA/kg tissue) pH 3 pH 6.5 pH 11 Figure 4-3. Lipid oxidation (TBARS ) during the product of acid (p H 3), alkaline (pH 11) and control (pH 6.5) derived is olates. Homogenates were exposed to the given pH values for 8 hours prior to obtaining and an isolate and corresponding isolates. Each bar in the graph is representative of TB ARS (umol MDA/kg of tissue) with standard deviations. Different small letters indicate a significant difference.

PAGE 101

101 ab ab ab a a ab a b abab 0 10 20 30 40 50 60 70 0 hr1 hr8 hr8 hr + 3 day storage Isolation TreatmentsTBARS (umol MDA/kg tissue) pH 3 pH 6.5 pH 11 4 day storage Figure 4-4. Lipid oxidation (TBARS ) of acid (pH 3), alkaline (pH 11) and neutral (pH 6.5) protein isolates derived from 1 and 8 hour pH exposure, analyzed right after preparation and after three da ys of refrigerated storage. Each bar in the graph is representative of TBARS (u mol MDA/kg of tissue) with standard deviations. The level of lipid oxidation in mince stored for four days (0 hr) is also shown. Different small letters indicate a significant difference. de e de e d b de a c f0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0 hr1 hr8 hr8 hr + 3 day storage Isolation TreatmentsProtein Carbonyl (nmol/ mg protein) pH 3 pH 6.5 pH 11 4 day storage Figure 4-5. Levels of protein oxidation as indi cated by protein carbonyls (nmol/mg protein) in mince stored for four days (0 hr), isolat es derived from 1 hour, and 8 hr pH holding times in addition to 3 days of refrigerated storage. Different small letters indicate a significant difference.

PAGE 102

102 c c cb ab b a abab0 5000 10000 15000 20000 25000 30000 35000 40000 45000 1 hr8 hr8 hr + 3 day storageIsolation Treatment TimeG'(Pa) pH 3 pH 6.5 pH 11 Figure 4-6. Storage Modulus (G) expr essed in terms of Pascals (P a) for isolate pastes derived from 1 hour, and isolates derived from 8 hr pH holding times in addition to three days of refrigerated storage. Isolate past es (85% moisture) were made through incorporation of sodi um tripolyphoshate (0.3%), sodium chloride (600 mM), and cold deionized water. Samples were adjusted to a final moisture content and of 85% and pH 7.2 Oscilation test were conducted usi ng a heating ramp (5-80C) and cooling ramp (80-5C) at rates of 2C/ min at 1% strain and 0.6284 Hz. Different small letters indicate a significant difference. 0 100 200 300 400 500 600 700 800 900 1000 0102030405060708090 Temperature (C)G' (Pa) 1 hr 8 hr 8 hr + 3 day storage Figure 4-7. Examples of rheograms from acid (pH 3) protein isolates derived from 1 and 8 hour pH exposure, analyzed right after prepar ation and after three days refrigerated storage. The rheograms show storage modul us (G) using a heating ramp (5-80C) and cooling ramp (80-5C) at rates of 2 C/ min at 1% strain and 0.6284 Hz. Isolates pastes (85% moisture) were made through incorporation of sodium tripolyphoshate (0.3%), sodium chloride (600 mM), and cold deionized. Samples were adjusted to a final moisture content and of 85% and pH 7.2. Cooling Heating

PAGE 103

103 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 0102030405060708090Temperature (C)G' (Pa) 1 hr 8 hr 8 hr + 3 day storage Figure 4-8. Examples of rheograms from alkaline (p H 11) protein isolates derived from 1 and 8 hour pH exposure, analyzed right after prep aration and after thr ee days refrigerated storage. The rheograms show storage modul us (G) using a heating ramp (5-80C) and cooling ramp (80-5C) at rates of 2 C/ min at 1% strain and 0.6284 Hz. Isolates pastes (85% moisture) were made through incorporation of sodium tripolyphoshate (0.3%), sodium chloride (600 mM), and cold deionized. Samples were adjusted to a final moisture content and of 85% and pH 7.2. 0 5000 10000 15000 20000 25000 30000 35000 0102030405060708090 Temperature (C)G' (Pa) 1 h 8 h 8 h +3 day storage Figure 4-9. Examples of rheograms from control (p H 6.5) protein isolates derived from 1 and 8 hour pH exposure, analyzed right after prep aration and after thr ee days refrigerated storage. The rheograms show storage modul us (G) using a heating ramp (5-80C) and cooling ramp (80-5C) at rates of 2 C/ min at 1% strain and 0.6284 Hz. Isolates pastes (85% moisture) were made through incorporation of sodium tripolyphoshate (0.3%), sodium chloride (600 mM), and cold deionized. Samples were adjusted to a final moisture content and of 85% and pH 7.2. Cooling Heating Cooling Heating

PAGE 104

104 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Figure 4-10. Protein composition of isolates derived from adjus ting and holding homogenates at acidic (3), alkaline (11) or control (6.5) pH for 1 hour, followed by readjustment to pH 5.5. Unless stated the sample were prepar ed and applied to the gel with inclusion of mercaptoethanol. Lanes 1: molecular weig ht standards (5 uL). Lanes 2-5 represent pH 11 isolates (15 uL, 10 uL, 10 uL, and 5 uL). Lane 6 and 7 represent pH 11 protein isolate (15 and 10 uL) prepared with no me rcaptoethanol. Lanes 8 and 9 represent pH 6.5 isolate (15, 10 uL). Lanes 10 and 11 repres ent pH 6.5 isolate ( 15, 10 uL) with no mercaptoethanol. Lanes 12 and 13 represent pH 3 isolate (15, 10 uL). Lane 14 and 15 represent pH 3 isolate with no mercaptoethanol (15, 10 uL). 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Figure 4-11. Protein composition of isolatesx derived from adjusting and holding homogenates at acidic (3), alkaline (11) or control (6.5) pH for 8 hour, followed by readjustment to pH 5.5. Unless stated the sample were prepar ed and applied to the gel with inclusion of mercaptoethanol. Lanes 1,12, 15: molecular weight standards (5 uL). Lanes 2 and 3 represent pH 11 protein isolate (15, 10 uL). Lanes 4 and 5 represent 6.5 protein isolate with no mercaptoethanol (15, 10 uL). Lanes 6 and 7 represent pH 6.5 isolate (15 and 10 uL). Lanes 8 and 9 represent pH 3 isolate without me rcaptoethanol. Lanes 9 and 10 represent pH 3 isolate (15 and 10 uL). Lanes 13 and 14 represent pH 11 isolate without mercaptoethanol. 205 kDa 116 kDa 97 kDa 84 kDa 66 kDa 55 kDa 45 kDa 36 kDa 29 kDa 24 kDa 20 kDa 14 kDa

PAGE 105

105 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Figure 4-12. Protein composition of isolates derived from adjusting and holding homogenates at acidic (3), alkaline (11) or control (6.5) pH for 8 hours, followed by readjustment to pH 5.5, and finally followed by 3 days of re frigerated storage. Unless stated the sample were prepared and applied to the ge l with inclusion of mercaptoethanol. Lanes 1, 10, and 11: molecular standards (5 uL). La nes 2 and 3 represent pH 6.5 isolate (15, 10 uL). Lanes 4 and 5 represent pH 6.5 isol ate without mercaptoethanol. Lanes 6 and 7 represent pH 11 isolate without mer captoethanol (15, 10 uL). Lanes 8 and 9 represent pH 11 isolate. Lanes 12 and 13 repr esent pH 3 isolate (15, 10 uL). Lanes 14 and 15 represent pH 3 isolates without mercaptoethanol. 1 2 3 4 5 6 7 8 9 Figure 4-13. Protein composition of in itial homogenates at acidic (3), alkaline (11) or contol(6.5) pH. Unless stated the sample were prepared and applied to the ge l with inclusion of mercaptoethanol. Lane 9: molecular standard s (5 uL). Lanes 1 and 2 represent pH 3 homogenate (15, 10 uL). Lanes 3 and 4 re present pH 6.5 homogenate. Lanes 6 and 7 represent pH 11 homogenate (10, 15 uL). Lanes 7 and 8 represent intial mince homogenate before pH adjustment.

PAGE 106

106 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Figure 4-14. Protein composition of homogenates held for one hour at acidic (3), alkaline (11) or control (6.5) pH. Unless stated the sample we re prepared and app lied to the gel with inclusion of mercaptoethanol. Lanes 1,4,15: mo lecular standards (5 uL). Lanes 2 and 3 represent pH 11 homogenate (15, 10 uL ). Lanes 5 and 6 represent pH 3.0 homogenate without mercaptoethanol. Lanes 7 and 8 represent pH 3 homogenate (10, 15 uL). Lanes 9 and 10 represent pH 6.5 ho mogenate, without mercaptoethanol (15, 10 uL). Lanes 11 and 12 represent pH 6.5 homogenate (10, 15 uL). Lanes 13 and 14 represent pH 11 homogenate without mercaptoethanol. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Figure 4-15. Protein composition of homogenates held for 8 hours at acidic (3), alkaline (11) or neutral (6.5) pH. Unless stated the sample we re prepared and app lied to the gel with inclusion of mercaptoethanol. Lanes 1,10,15: mo lecular standards (5 uL). Lanes 2 and 3 represent pH 11 homogenate (15, 10 uL ). Lanes 4 and 5 represent pH 11 homogenates without mercaptoethanol (10, 15 uL). Lanes 6 and 7 represent pH 6.5 homogenate (10, 15 uL). Lanes 8 and 9 represent pH 6.5 homogenate, without mercaptoethanol (15, 10 uL). Lanes 11 a nd 12 represent pH 3 homogenate (10, 15 uL). Lanes 13 and 14 represent pH 11 homogenate without mercaptoethanol (15,10 uL).

PAGE 107

107 CHAPTER 5 EFFECT OF ACID PROCESSING ON LIPI D OXIDATION IN A WASHED SPANISH MACKEREL MUSCLE MODEL SYSTEM Introduction Lipid oxidation in muscle foods is predominan tly detrimental to overall quality and storage stability ( 29, 233 ). An array of tissue components contribut e to lipid oxidation in muscle foods. Integral components in lipid oxi dation postmortem as elucidated by studies thus far are heme proteins, phospholipids, and components in the aqueous phase of the muscle ( 155, 234 ). Of the various heme proteins found in dark muscle, the major promoter of lipid oxidation is hemoglobin ( 46, 206, 235 ). According to various sources ( 5, 13, 16, 51, 67 ), acid processing, in contrast to alkaline processing, results in rapid development of lipid oxida tion as well as poor protein functionality and color. It is known that pro-oxidative activity of hemoglobin is greatly increased at acidic pH ( 12, 43, 57, 206 ) and that cytosolic extracts have both anti and pro-oxidative effects in vitro and in vivo like systems ( 41, 155, 236-239 ). The primary substrates for lipid oxidation are membrane lipids and/or lipid hydroperoxides according to various sources ( 36 ). It is important to understand the inte raction between cellular membra ne lipids, hemoglobin, and inherent cytosolic compounds, as the quality of da rk muscle foods could be favorably controlled. The purpose of this study was to investigate how key components in Spanish mackerel muscle tissue contribute to oxida tion during adjustment to low pH and readjustment to neutral pH using a washed fish muscle basic model system. In addition, this study was conducted to investigate whether the effects of lipid oxidation in a Spanish mackerel basic model system are congruent with the extent of lipid oxidation obs erved in acid homogenates and isolates made from the same fish muscle source.

PAGE 108

108 Materials and Methods Raw Materials Spanish mackerel was freshly caught during th e Fall season in Cedar Key, Florida. Within approximately an hour, fish was transported on ice to the Aquatic Food and Bimolecular Research Laboratory, University of Florida. Upon arrival, fish were manually skinned, degutted, and filleted. One half of the fillets were minced using a large scale grinder (The Hobart MFG. Co., Model 4532, Troy, Ohio). Fillets and ground fish (100 gram quantities) were vacuum packed and stored at -80C unt il needed. The inital average moisture content and pH of the prepared mince from both batches was 80% and 6.6. Chemicals Tricholoracetic acid was obtai ned from Fischer Scientific (Fair Lawn,New Jersey). Thiobarbituric acid (TBA), 1,1, 3,3 tetr aethoxypropane, potassium phosphate (K2HPO4), sodium phosphate (NaH2PO4), fluorescein, SigmaM arker molecular weight standard, and EZBlue stain solution, methanol, and streptomycin su lfate were obtained from Sigma Chemical Co. (St. Louis, MO). Propyl gallate and (ethlenediaminetetraacetic acid 99%) EDTA were obtained from Acros Organics (New Jersey, USA). Pre-cast linear 4-20% Tris-HCL gradient gel, Laemmli buffer, and B-mercaptoethanol, Power Npac 300 power supply were obtained from Bio-Rad laboratories (Hercules, CA). The Coomassie Pl usBetter Bradford Assay kit was obtained form Pierce (Rockford, Il). 2,2 Azobis (2 -amidinopropane) dihydrochloride (AAPH) was obtained from Waco Chemicals, Richmond, VA.. Trolox (6-hydroxy-2,5,7,8tetramethylchroman-2-carboxylic acid) was obt ained from Aldrich Chem, Inc., MI, WI. Washed Fish Frozen (-80C ) Spanish mackerel was thawed under cold running water and was placed immediately on ice. Washed fish was pr epared according to Undeland et al ( 155 ). First, all dark

PAGE 109

109 muscle was removed. The white muscle was minced (Oster Heavy Duty Food Grinder, Sunbeam Corporation, Inc.) and then washed with 3 volumes of cold (4C) distilled water. The mixture was manually mixed with a plastic spat ula for 2 minutes and was held on ice for 15 minutes. The mince was then dewatered with 2 layers of cheese cloth. Two washes with 3 volumes of cold 50 mM NaCl solution (pH 5.5) followed. The mince was dewatered between each wash. For the first wash with NaCl, the mixt ure was stirred for 2 minutes and was held for 15 minutes. The second wash with NaCl wa s followed by 1 minute homogenization and centrifiguation (15,000 g, 20 minutes, 4 C). The final moisture cont ent of the washed mince was 75%. The mince was vacuum packed in 50 g quantities and was stored at -80C Preparation of Spanish Mackerel Hemolysate Freshly caught Spanish mackerel were anesth etized (0.5 g ethyl 3aminoenzoate/ L of water) and placed on ice. Blood was collected from the caudal vein using syringes (25 guage, 1 inch needles) according to Rowley ( 240 ). The syringes were pre-loaded with heparin solution (150 units/mL). Hemolysate was collected from Sp anish mackerel blood according to Fyhn et al. ( 241 ). Four volumes of extraction buffer (1.7% NaCl, 1 mM Tris, pH 8.0) was added to the collected blood. Samples were centrifuged at 700 g for 10 minutes, 4 C, using a tabletop centrifuge (Eppendorf Centrifuge 5702, Brinkman Instruments. Inc., Westbury, N.Y.). The supernatant (plasma) was discarded and the se diment (red blood cells) was washed in three volumes of the extraction buffer and again centrifuged at 700g. The supernatant was again decanted and the red blood cells (sediment) were lysed through addition of 3 volumes of 1mM Tris, pH 8.0 buffer for one hour. Then, 1 mM NaCl (1/10 volume) was added to precipitate and maximize stromal protein removal. The lysed red blood cells were centrifuged for 15 minutes, 28,000g, 4C. The hemolysate, primarily hemoglobi n, was stored in 1 mL cryogenic tubes at 80C.

PAGE 110

110 Hemoglobin Quantification Hemoglobin levels were quantif ied according to Bradford ( 195 ) using the Coomassie PlusBetter Bradford Assay kit. An approxi mate hemoglobin concentration was calculated using a BSA standard curve and the average molecular weight of hemoglobin (ca. 66,000). Spanish mackerel Pressed juice Spanish Mackerel pressed juice was prep ared according to Gunnarsson et al. ( 238 ). Prefrozen Spanish mackerel mince was thawed under cold running water and was added to centrifuge tubes (100 g) samples. The samples were spun at 22,000 g for 4 hours at 4 C. The supernatant was drawn off and was filtered through four layers of cheese cloth. The pressed juice was nitrogen flushed and stor ed at -80C until needed. Thiobarbituric Acid Reactive Substances (TBARS) TBARS were assessed according to Lemon ( 193 ). One gram of sample was homogenized with 6 mL of trichloracetic acid (TCA) soluti on composed of 7.5 %TCA, 0.1% propyl gallate, and 0.1% EDTA dissolved in deionized water. The homogenate was filtered through Whatman No. 1 filter paper. The TCA extract (2 mL) was added to 2 mL of thiobarbituric acid (TBA) solution (0.23% dissolved in deionized wate r). The TCA and TBA solutions were freshly prepared and heated if needed. The TCA/TBA solution was heated for 40 minutes at 100C. The samples were centrifuged at 2,500 g for 10 minutes at 4C. A standard curve was constructed using 1,1, 3,3 tetraethoxypropane ( 193 ). Spanish mackerel Pressed Juice protein Quantification and Characterization Protein isolate concentrati ons were assessed using a modified Biuret method ( 194 ). The Bradford method was used to quantify proteins leve ls, either in low concentration (<1 mg/mL) or limited quantity ( 195 ), using the Coomassie PlusBette r Bradford Assay kit. Protein composition in isolates were an alyzed using electrophoresis ( 196 ). SDS-PAGE (sodium dodecyl

PAGE 111

111 sulfate polyacrylamide gel electrophoresis). Briefl y, protein concentration was adjusted to 3 mg/ mL. A 1 mg/mL protein concentr ation was obtained by combining 0.33 uL of sample, 0.66 uL of Laemelli buffer, and 50 uL of B-mercaptoethanol. Samples were boiled for 5 minutes in 1.5 mL cryogenic tubes and were then cooled on ice. Precast linear 4-20% gels we re run for 1 hr at 200 V; mA 120 with the Power Npac 300 power supply. Sa mples were applied to the gels in 10 and 15 uL volumes. Molecular weight standards (6,500 205,000 Mw) were run in 5 and 15 uL quantities. Gels were then set with 12% TCA for 1 hr and were stained with EZ-Blue stain solution (minimum of 1 hr with agitation). Gels were then destai ned with deionized water for 18 hrs with agitation, changing out the water numerous times. Preparation of Low Moisture Washed Spanish Mackerel Mince Oxidation Model System The model system was prepared according to Undeland et al. ( 155 ). Thawed washed Spanish mackerel (75% moisture) was divided out into 15-20 gram portions with 200 ppm streptomycin sulfate added. The raw pressed juic e had a protein concentration of approximately 100 mg/mL and was diluted to 30 mg/mL. To test the antiand pro-oxidative effects of pressed juice and hemoglobin treated at lo w pH (3) and readjustment to physiological pH (6.8), pressed juice and hemoglobin were adjusted to appropriate pH and were held for 1 hour and readjusted to pH 6.8 if needed. Pressed juice a nd water were added to the washed muscle to bring the moisture content of the overall system to 82%. The presse d juice was added to the washed muscle such that ~ 4 fold dilution was achieved. The final pressed juice protein c oncentration was ~ 8-10 mg/mL. In control samples, pressed juice was replaced with equal amounts of water. Samples were adjusted to pH 6.8 through drop wise addi tion of NaOH and thoroug h mixing with a mortar and pestle. The system was peri odically checked for pH. Once the desired pH was reached, the samples were plated in petri dishes. Hemoglobin was added in 6 M concentrations and was well incorporated with at least 100 turns of a spatula. Thorough mixing of the hemoglobin was

PAGE 112

112 indicated by a homogenous color. Plated samp les were stored at 4C for several days. The combinations involved pH adjustment of one, tw o, or three oxidation model constituents: washed fish muscle (WFM), pressed juice ( 103 ), or hemoglobin (Hb). The various combinations were conducted together and included: One way Adjustment: Control (H b (pH 6.8)+ PJ (pH 6.8) + WFM ( pH 6.8)) Control without PJ (Hb (pH 6.8)+ WFM (pH 6.8)) Hb (pH 3-6.8) +PJ (pH 6.8) +WFM (pH 6.8) Hb (pH 6.8) +PJ (pH 3-6.8) + WFM (pH 6.8) Hb (pH 6.8)+ PJ (pH 6.8) + WFM (pH 3-6.8) Multiple Adjustment: Control (Hb (pH 6.8) + PJ (pH 6.8) + WFM (pH 6.8)) Hb (pH 3-6.8)+ PJ (pH 6.8) + WFM (pH 3-6.8) Hb (pH 3-6.8) + PJ (pH 3-6.8) + WFM (pH 6.8) Hb (pH 6.8) + PJ (pH 3-6.8) + WFM (pH 3-6.8) Hb (pH 3-6.8) + PJ (pH 36.8) + WFM (pH 3-6.8) Samples were sniffed several times day to ev aluate relative lipid oxi dation progress. Based upon sensory, the 0.5-1 g plugs were taken once a da y until odor intensity began to subside. After plugs were taken, the remaining samples were remixed and flattened and were returned to 4C. Samples were stored in aluminum foil at -80C until analysis was to be performed. Preparation of Hemoglobin and Pressed Juice Oxidation Model System To mimic the washed system pressed juice and hemoglobin concentration upon addition to the washed fish muscle, the same volumes were prepared. Typically, 4.3 to 4.5 mL of pressed juice (30 mg/mL of protein) and 0.391 L of hemoglobin (6 M) were added to 11 grams of

PAGE 113

113 washed fish (75% moisture). Approp riate aliquots of pressed juice ( 103 ) and hemoglobin (Hb) were prepared and were adjusted to pH 3 or pH 6.8 with addition of 1 N NaOH or HCL (5 l at a time). The pH treated samples were held for an h our in glass test tubes for one hour. After one hour, the pH 3 adjusted pressed juice and/or hem oglobin were readjusted to pH 6.8 with addition of NaOH or HCL, except for controls, which rema ined the same pH. The final treatments were as follows: Hemoglobin /Pressed juice treatments: Hb (pH 3-6.8) PJ (pH 6.8) Hb (pH 3-6.8) PJ (pH 3-6.8) Hb (pH 6.8) + PJ (pH 6.8) Hb (pH 6.8) + PJ (pH 3-6.8) Hb (pH 3-6.8) + PJ (pH 6.8) Hb (pH 3-6.8) + PJ (pH 3-6.8). Oxygen Radical Absorbance Capacity (ORAC) To investigate oxidative changes in the washed system as result of adding pressed juice and hemoglobin, antioxidant activities were evalua ted. The antioxidant activity of pressed juice and hemoglobin alone and in combination were assessed using the ORAC assay according to Cao et al. ( 242 ). Florescein is used a marker of oxi dation inhibition, induc ed by AAPH. Briefly, samples were centrifuged at a low speed (1,000 g, Eppendorf Centrifuge 5415 D, Brinkmann Instruments. Inc., Westbury, N.Y.) to remove in solubles. Samples were normalized to 1 mg/ mL with water using the Bradford method ( 195 ). The Bradford protocol were conducted according to the Coomassie PlusBetter Bradford Assay kit. Sample were then diluted 1:10 with ORAC buffer (0.75 M K2HPO4 and 0.75 M NaH2PO4, pH 7) buffer. The sample (50 uL) was applied to

PAGE 114

114 the 96 well Costar microplate (F ischer Scientific, Fair Lawn, New Jersey). A florescein stock solution (50 nM in methanol) was diluted with ORAC buffer (1:1250) and was added (100 l) to the microwells. Before each run, AAPH was fres hly made from an AAPH stock solution (90 mg /mL ORAC buffer) and was quickly added to th e microwells (50 l). The microplate was immediately placed in a pre-heated (42C) 96-well florescent Molecular Device fmax microplate reader (Molecular Devices Corporation, Sunnyvale, CA) set to a 15-20 s pre-mix period, 485 nm excitation and 538 emission). Rela tive florescent unit (RFU) readings were recorded every 2 minutes in kinetics mode using SoftMax Pr o software (Molecular Devices Corporation, Sunnyvale, CA). The total assay time is 1 hour and 10 minutes. Trolox standards were applied in 25-200 M concentrations. Buffer, trolox standards, and samples were applied in at least duplicate. The area under the florescein decay curve for samples and standards was calculated and final values were ca lculated using a quadratic equation. AUC (Area under the curve): 0.5 +f5/f4+f6/f4 +f7/f4 +f34/f4+ f35/f4) x CT Where: F4 = fluorescent reading at cycle 4 Fi = flouresenct reading at cycle i. CT = cycle time in minutes. All final ORAC value were expressed in micr omoles of Trolox equivalents (TE)/ mg of protein. Data was calculated a nd tabulated in Microsoft Ex cel (Microsoft, Redmond, WA). Total Antioxidant Potential Total antioxidant capacity was assayed us ing the Cayman Chemical Antioxidant Kit (Cayman Chemicals, Ann Arbor, MI). In th is kit, hydrophilic and lipophilic compounds are tested for their abil ity to inhibit metmyoglobin induced oxidation of 2,2Azino-di(3ethylbenzthiazoline sulphonate ABTS. The presence of the oxidized form of ABTS, ABTS +, is measured at 750 nm. With successful inhibiti on of oxidation, lower absorbancies are observed.

PAGE 115

115 Antioxidant capacities are calcula ted relative to standard conc entrations of Trolox, a water soluble analogue of tocopherol. Samples were no rmalized to 1 mg / mL of protein and were centrifuged to remove insoluble compounds prior to analysis. Total antioxidant capacity was expressed in mmol Trolox e quivalents/mg of protein. Moisture Content and pH Moisture content was assessed using a moistu re balance (CSI Scientific Company, Inc., Fairfax, VA). The pH was measured using an elec trode ( Thermo Sure-Flow electrode (Electron Corp., Waltham, MA) and pH meter (Denver Instrument, model 220, Denver, CO). The homogenate samples were directly measured for pH. Isolates samples were diluted with deionized water (1:10) to measure pH. Statistical Analysis The SAS system was used to evaluate data. A two-way ANOVA model was used to evaluate statistical significan ce for lipid oxidation (TBARS). The two factors for the TBARS data were treatment and time and they were an alyzed with cross design. Antioxidant capacities for pressed juice, hemoglobin, and combina tions were analyzed with treatment nested within time. All analyses were performed at leas t twice and all experiments were replicated twice. Least square means were us ed for post hoc multiple comparisons with Tukey adjustment controlling for Type I error. Data were reported as mean and standard deviation. The analysis of variance and least significant differe nce tests were conducted to identify differences among means, while a Pearson correlation test was conducted to determine the correlations among means. Statistical significan ce was declared at p < 0.05.

PAGE 116

116 Results To gain an understanding of how key components in Spanish ma ckerel tissue contribute to oxidation through adjustment to acidic pH and then readjustment to neutral pH, three experiments were conducted using a washed musc le model system (WFM) in which: (1) One of two components in a washed Spanish mackerel sy stem were adjusted to pH 3 for one hour and readjusted to pH 6.8 (2) Two and three components in the washed Spanish mackerel system were adjusted to pH 3 for one hour and then readjust ed to pH 6.8. Lastly, non substrate components, including hemoglobin and th e cytosolic extract, alone or in co mbination were adjusted to pH 3 and readjusted to pH 6.8. Aqueous extract (pressed juice (PJ), hemogl obin (Hb), and WFM (phospholipids source) were prepared from freshly caught Spanish m ackerel. PJ (30 mg/mL) Hb (6 uM) and WFM (75%) moisture were combined. The components were combined resulting in a 4 fold PJ dilution. Prior to acidic adjustment, the pH of the pressed juice and blood were ~6.08 and ~6.68, respectively. Moisture contents of the washed muscle prior to experimentation ranged between 74 and 75%. Protein contents of hemoglobin and pressed ju ice were 186.19.71 g/mL and 102.44 3.41 mg/mL, respectively. This Spanish m ackerel pressed juice protein content was greater than the pressed juice (51 mg/ mL) obtained by Undeland et al. ( 155 ) from cod muscle. Auto-oxidation was apparent with pH 3 adju stment within minutes, as indicated by a chromophore change from red to rusty brown a nd coagulation of hemoglobin. As reported by Lawrie ( 243 ), denaturing agents will oxidize hemoglobin to a globul ar met or deoxy hemoglobin state with a brown, and sometime s grayish color. In agreement with the change of color and aggregation, the globin is most likely detache d, but the hematin nucleus mostly likely remains intact ( 243 ). Hemoglobin retained the same color and aggregated appearance with readjustment to pH 6.8.

PAGE 117

117 One and Two Way adjustment Overall lipid hydroperoxides (LHP) in the one component adjustment experiment (Figure 5-1a) were significantly differe nt (p>0.02) with WFM (pH 3-6.8) and NPJ (pH 6.8) >PJ (pH 36.8)> Control (pH 6.8) and Hb (pH 3-6.8). Hydrope roxide levels began declining in the WFM (pH 3-6.8) treatment at day 3, or as TBARS incr eased (Figure 5-1b). Increasing LHP levels were in agreement with levels of formed TBARS, with LHP (mmol/kg tissue) significantly (p<0.001) increasing in WFM (pH 3-6.8) and NPJ (pH 3-6.8) at 1 and 2 days, resp ectively. At day 7, WFM (pH 3-6.8) and NPJ (pH 3-6.8) reached the same levels of LPH (Figures 5-1a). Similar to TBARS, Hb (pH 3-6.8) and control PV values were comparable. In one component adjusted treatments, an overall significant increase (p<0.0001) in TBARS over 7 days was observed (Figure 5-1b). Overall TBARS formed in the following order: control without pressed juice (NPJ (pH 6.8))> WFM (pH 3-6.8)>PJ (pH 3-6.8) > control (pH 6.8) = Hb (pH 3-6.8). Significant increases among tr eatments were observed over time (p<0.0001); however, the control (pH 6.8) and Hb (pH 3-6.8) treatments did not significantly increase over time nor did they reach rancidity levels (20 m ol MDA/kg tissue) (Figur e 5-1b). The greatest lag phase (7 days) with respect to TBARS levels for one component tissue adjustment was observed in the control and Hb (pH 3-6.8), followed by PJ (pH 3-6.8)>NPJ>WFM (pH 3-6.8). The WFM (pH 3-6.8) and NPJ (pH 3-6.8) treatments formed rancidity levels of TB ARS after one and two days refrigerated storage (p< 0.0001). After one day of storage, TBARS levels in the NPJ (pH 6.8) treatment were greater (p<0.044 2) than the PJ (pH 3-6.8) at day 0. Appreciable levels of lipid oxidation in the PJ (3-6.8) compared to Hb (pH 3-6.8) and the control (pH 6.8) occurred after three days of storage, and rancidity levels were not reache d until approximately, 7 days of storage (Figure 5-1b). After 7 days of storage, the NPJ (pH 6.8) and PJ (pH 3-6.8) treatments were still increasing in levels of oxidation (p<0.0388) and (p< 0.0054) after 7 days of storage.

PAGE 118

118 The PJ (3-6.8) treatment reached equal levels of oxidation as the WF (3-6.8) treatment after 7 days of storage (Figure 5-1b). Two and Three Component Adjustment Increases in LHP values among treatments we re observed over 7 days. Overall levels (p<0.01) of LPH were as follows: Hb+ WFM (p H 3-6.8), Hb +PJ (pH 3-6.8), and PJ +WFM (pH 3-6.8)> Control (pH 6.8) = Hb+PJ+WFM (pH 3-6.8) The control (pH 6.8) and Hb +WFM (pH 3-6.8) treatment primary and secondary levels of oxidation appeared to follow the same trends. TBARS levels (Figure 5-2b), were significantl y less in the control (pH 6.8) than in all other treatments. The treatment with all three compone nts adjusted (Hb+PJ+WFM) from pH 3 to 6.8 had significantly (p<0.0001) grea ter levels of oxidation compar ed to (Hb+ WFM) adjusted from pH 3 to 6.8 with the excl usion of adjusted PJ samples. At day 1, significant (p<0.001) increases in TBARS for all treatments occurre d, excluding the control (pH 6.8), as there was not a significant lag phase among tr eatments. At day 7, Hb+PJ (pH 3-6.8) incurred higher levels (p<0.0151) than the PJ +WFM (pH 3-6.8) and PJA +WFM (pH 3-6.8). Comparison of one way and multi-compon ent pH adjustment in Washed System After 7 days of storage, several differences among pH adjusted and readjusted components were observed (Figure 5-1b,2b). The NPJ (pH 6.8) WFM (pH 3-6.8), PJ (3-6.8), Hb (pH 3-6.8) +PJ (pH 3-6.8) had significantly greater TBARS levels than the control(s) overall (p<0.0002). Of the treatments significantly differe nt than the control, the NPJ (pH 6.8) had the greatest peak levels of TBARS after 7 days (p<0.0001). WF M (pH 3-6.8) TBARS were not significantly different from Hb (pH 3-6.8) +PJ (pH 3-6.8) or PJ (pH 3-6.8). The Hb+PJ (pH 3-6.8) treatment was not significantly different from the PJ (p H 3-6.8) treatment, and both had greater TBARS than Hb (pH 3-6.8). The TBARS trend among one way and multi-component pH treatments was

PAGE 119

119 as follows: NPJ (pH 6.8)>WFM (pH 3-6.8)= Hb+P J (pH 3-6.8) = PJ (pH 3-6.8) > all other treatments. The overall greatest rate and extent of oxidation (TBARS and PV) over seven days was found in the WFM (3-6.8), aside from the NPJ (p H 6.8) treatment (Figure 5-1,2). PJ (pH 3-6.8), Hb (pH 3-6.8) +PJ (pH 3-6.8), PJ (3-6.8) + WFM (3 -6.8) + Hb (3-6.8), PJ (3-6.8) + WF (3-6.8) were equally pro-oxidative. PV levels corres ponded with TBARS levels among the treatments. Lipid hydroperoxides began increasing (p<0.0001) at the fastest rate and occurred to greatest extent (aside from the pressed juice) in the WF M (pH 3-6.8) followed by equivalent trends in the Hb+PJ (3-6.8) and PJ (3-6.8) tr eatment (Figures 5-1a, 5-2a). Antioxidant mechanisms of non-substrate components WFM model system ORAC Test To better understand the effect s of acidic pH and readjustme nts to pH 6.8 on PJ and Hb, the same samples from the WFM model system were analyzed for antioxidant activity, using ORAC and TEAC assay (Figure 5-3 a,b). Relative concentrations of Spanish mackerel PJ (30 mg protein/ mL) and Hb (6 M) alone or in combin ation remained consistent with the WFM model system in the absence of Spanish mackerel WFM. The radical scavenging ability of PJ and Hb alone or in combination was measured by the oxygen radical absorbance capacity (ORAC) assay (Figure 5-3a). To confirm that Spanish mackerel Hb and PJ had scavenging ability at physiological pH, untreated Hb (pH 6.6) and PJ (pH 6.08) were tested. ORAC values were si gnificantly different (p<0.0001) among native Hb and PJ, with values of 0.719.123 and 0.221 0.107 umol Trolox equivalents (TE)/mg protein. The scavenging capacity of PJ and Hb at neutr al (pH 6.8) and acidic pH (pH 3) compared to physiological pH was also tested. The radical sc avenging ability of pressed juice and hemoglobin was reduced (p<0.001) with acidi fication and neutral pH co mpared to physiological pH

PAGE 120

120 (Figure 5-3a). Acidified hemoglobin and pre ssed juice had undetectable ORAC values. Readjustment of hemoglobin and pressed juice fro m pH 3 to neutral pH gave (p<0.001) ORAC values that were significantly less than the untr eated hemoglobin, and readjustment of acidified PJ to pH 6.8 gave ORAC values that were insign ificant from untreated pr essed juice. Untreated PJ (pH 6.6) ORAC values were not significantly differe nt from Hb (pH 6.8) nor were Hb (pH 6.8) and PJ (pH 6.8) significantly diffe rent from each other (Figure 5-2a). Similar to the WFM model system, the Hb and PJ were held at acidic pH for one hour and were then readjusted to neutra l pH for ORAC testing, either al one or in combination. As shown in Figure 5-2 a, one hour of holding for Hb (p H 6.8) gave increased ORAC values, or 0.993 0.122 mol TE/ mg protein; while, no change wa s observed for PJ (pH 6.8). Hemoglobin and PJ exposure to pH 3 for one hour followed by read justment to pH 6.8, gave significant (p<0.0001) increases in scavenging capacity relative to init ial values at pH 3; however, they were not significantly different from each other (Figure 53a). Interestingly, an accumulative effect of ORAC abilities was observed when PJ (pH 3-6.8) + Hb (pH 6.8) and PJ (pH 3-6.8)+ Hb (3-6.8) were combined. Overall increases in scave nging capacity increased when PJ (3-6.8) was combined with Hb at pH 6.8 or at pH 3-6.8 (Fi gure 5-3a). Combinations involving PJ at pH 6.8 with either Hb at pH 6.8 or pH 3-6.8 were not significantly different from each other, nor were they significantly different from PJ at pH 6.8. TEAC Assay Trolox equivalent antioxidant capa city (TEAC) test was used to test the same samples. In contrast to ORAC, no significant difference was observed among untreated or acidified Hb and PJ. An adjustment to pH 6.8 gave rise to incr eases in TEAC for PJ (Figure 5-3b). This increase in PJ antioxidant capacity following pH readju stment was not observed, using ORAC (Figure 53a). After one hour of holding, Hb and PJ at pH 6.8 both increased (p<0 .001) in antioxidant

PAGE 121

121 capacity with corresponding values of 0.318 0.052 and 0.378 0.025 mM TE/mg protein. Combined PJ and Hb treatments were either e qual to or less than corr esponding PJ and Hb pH 6.8 treatments. Similar to ORAC, PJ (pH 3-6.8)+ Hb (6.8) was significantly greater Hb (pH 36.8) and PJ (36.8), and pJ (pH 6.8)+ Hb (pH 3-6.8) with p<0.0001. SDS Page The electrophoretic pattern of Spanish mack erel pressed juice was comparable to cod pressed juice as found by Undeland et al. ( 155 ), with exception of the 8 kDa fraction observed in the cod pressed juice. Identif iable proteins in the Spanish mackerel pressed juice include subunits of myosin (97, 75 kDa, 67 kDa), Desmin and Vimentin (55 and 52 kDa), actin (49 kDa), actin subunit (36 kDa), and troponin T (29 kD a), and myoglobin and/or subunits of hemoglobin (14 kDa). Discussion The pH sequence to which fish muscle com ponents are exposed in the acid/alkali aided process can activate oxidative mechanisms. In ge neral, the effect of pH can vary depending upon the complexity of the system and the catalys ts involved. The specific tissue components proved to be most important in the lipid oxidation of muscle foods are sti ll under much debate. To investigate the roles of tissue comp onents in the lipid oxidation of dark post-mortem fish muscle, Spanish mackerel hemoglobin, pressed juice, and washed muscle were combined and exposed to acidic (pH 3) and then pH 6.8 readjustment. The pH 3 adjustment treatment was utilized as per discussion in Chapter 3, in which the pH 3 homogenates along with pH 3.5 homogenates incurred the greatest levels of lipid oxidation (PV and TBARS). Furthermore, other studies have shown significant oxidation in washed cod systems adjusted to pH 3. Kristinsson ( 51, 152 ) observed a rapid increase in TBARS in a hemogl obin mediated washed cod system in which washed cod and hemoglobin were adjusted to acid pH (2.5-3.5). Similarly, Pazos et al. ( 161 )

PAGE 122

122 observed a rapid onset and rate of lipid oxidatio n in a pH 3.5 adjusted wash cod system with hemoglobin ( 161 ). Because holding times (1 and 8 hour s) had no effect on lipid oxidation development in the acid isola tion of Spanish mackerel, a one hour holding time of tissue components was implemented. Kristinsson and Hultin ( 51 ) found that increased holding times from 90 sec to 20 minutes increased levels of oxidation in a pH 3.0 adjusted washed cod hemoglobin mediated system; however, at pH 2.5 no noticeable changes were observed with different holding times. Additionally, pH 6.8 was c hosen as the readjustment pH due to findings made by Pazos et al. ( 179 ). In a hemoglobin-catalyzed lipid oxidation system, the extent and rate at which cod microsomes oxidized fastest occurred at pH 6.8 ( 179 ). The optimal pH range for sarcoplasmic reticulum (SR) oxidation for in vitro fish models has been found to be between pH 6.5 and 6.9 ( 5 ). The washed Spanish mackerel muscle system serves as the phospholipid matrix. No pressed juice treatment (control without pressed juice) Lipid hydroperoxides and TBARS we re the highest in the cont rol treatment with no added pressed juice (NPJ) in which Spanish mackerel washed fish muscle (WFM) and hemoglobin (Hb) were combined at pH 6.8 (Figur e 5-1b). According to Pazos et al. ( 161 ), there were appreciable levels of lipid oxi dation at pH 6.8 in a microsom al cod system to which cod hemoglobin was added ( 161 ). Above pH 6.8, lipid oxidati on decreased with increased microsomal pH adjustments to 8.4. Adjustment s below 6.8 reduced rates of lipid oxidation further at pH 6 and at pH 3.5 comp ared to pH 8.4. Pazos et al. ( 161 ) also investigated the effects of pH adjustment on the development of lipid oxidation in a washed fish cod matrix with hemoglobin added, and an inverse correlation be tween lipid oxidation and pH was observed compared to the tested microsomal system Longer lag phases and reduced TBARS were observed with decreases in pH from 7.4 to 6.8. Decr easing the pH to 3.5, resu lted in even greater extent and rate of oxidation ( 161 ). The latter findings in the microsomal system agree with the

PAGE 123

123 rapid rate of oxidation seen at pH 6.8 in the NPJ treatment in the presence of 5 M Spanish mackerel hemoglobin (Figure 5-1a,b). Kristinsson ( 152 ) found that in the absence of hemoglobin, TBARS did not form in a washed co d system. Thus, the pro-oxidative activity of hemoglobin exhibited in the study by Kristinsson ( 152 ) concurred with the NPJ treatment in which only hemoglobin and washed fish muscle were combined at pH 6.8. Undeland et al. ( 155 ) found that there was significant lipid oxidation in a system very similar to the current study, including washed cod and 5 M trout hemoglobin ( 155 ). In a study implementing native sperm whale myoglobin and washed cod, significan t levels of oxidation were observed ( 244 ). Washed cod in the absence of sperm hemoglobin develo ped very little levels of oxidation (TBARS, LHP). Lipid oxidation in the control samples The fact that in the control treatment, untre ated pH hemoglobin did not catalyze lipid oxidation in the presence of pressed juice co mpared to the no presse d juice (NPJ) control treatment (Figure 5-1a,b), confirms the antioxida tive capacity of presse d juice. According to Undeland et al. ( 155 ), TBARS and PV formation were co mpletely inhibited in a hemoglobin mediated washed cod muscle system, in the presence of cod pressed juice ( 155 ). In the Spanish mackerel WFM system, the peak levels of oxidati on were highest in the control without pressed juice (Figure 5-1a,b). The inhibito ry activity of pressed juice and the highly pro-oxidative nature of hemoglobin on lipid oxidation are equally evident from the NPJ (pH 6.8) and control (pH 6.8) treatments (Figures 5-1a,b). It is evident that the incorporation of aqueous extracts native to fish cytosol would enable for protection from reac tive oxygen species in the phospholipid membrane post mortem, similar to native tissue. According to Undeland et al. ( 155 ), the formation of hydroperoxides and painty odor wa s significantly delayed or i nhibited in the hemoglobin mediated cod oxidation system when cod aqueous extracts were added. Similar antioxidative

PAGE 124

124 capacities have been observed in other fish species including haddock, dab, and winter flounder ( 155 ). In a study conducted by Decker and Hultin ( 245 ), pressed juice from light mackerel muscle effectively inhibited metal catalyzed oxidation ( 245 ). It was evident from the low levels of lipid oxidation in the contro l (pH 6.8) compared to the NPJ (pH 6.8) treatment, that a four fold d ilution of the Spanish mack erel pressed juice was effective in inhibiting lipid oxi dation (Figure 5-1a,b). It is well known that the antioxidative nature of compounds is affected by environm ental changes in pH and concentration ( 64-66 ). Undeland et al. ( 155 ) found that dilutions up to 6 fold signi ficantly prolonged the lag phase by at least 4 days; whereas, whole pressed juice delaye d the onset of oxidation by >11 days. A < 1 kDa dialysis retentate and whole pressed juice we re equally inhibitive of lipid oxidation in the washed cod system. Thus, the antioxidative nature of the pressed juice is attributed to low molecular weight compounds rather than antioxida nt proteins (superoxide dismutase, catalase, and glutathione peroxidase). Chicken pressed juice was most effective with low and high molecular weight compounds present as found by Rong et al. ( 246 ). When ascorbate oxidase (200 M) was added to the chicken pressed juic e in the presence of he moglobin, significantly more oxidation occurred. These findings suppor t the importance of low molecular weight compound including and not limited to as corbate, glutathione, and urate ( 246 ). An antioxidative balance in the presence of hem oglobin, in the Spanish mackerel pressed juice, affirms the abundant antioxidative capacity of low molecular weight compounds as observed by Undeland et al. ( 66 ) in cod pressed ju ice and Takama ( 146 ) in trout pressed juice. The reasons for decreased prooxidative activity of hemoglobin in the presence of pressed juice have yet to be elucidated. Inhibitory mechanisms may be related to th e conversion of hemoglobi n to deoxyhemoglobin, or with autoxidation, i.e. conversion to metmyoglob in. This means as a cause for reduced lipid

PAGE 125

125 oxidation were tentatively c oncluded by Undeland et al. ( 155 ) in a hemoglobin/pressed juice aqueous system. These studies suggest that Span ish mackerel pressed ju ice may have inhibited lipid oxidation by preventing conversion of hem oglobin to prooxidative fo rms. Heme proteins are believed to promote oxidation through releas e of hemin from globin, enabling opportunities for hemin and lipids in muscle tissue to react ( 247-253 ). Grunwald and Richards ( 244 ) showed that hemoglobin dissocation into subunits incurr ed greater levels of lipid oxidation than undissoaciated hemoglobin. Hemoglobin was gene tically altered to achieve non dissociable subunits. With dissociation, myoglobin subunits re sult. According to Grunwald and Richards ( 244 ), spermwhale metmyglobin induced greater fo rmation rates and levels of oxidation in a washed cod muscle system versus ferrous myogl obin due to lower hemin affinity. Hemin is defined as the poryphrin ring with ferric (Fe3+) iron bound. Grunwald and Richards ( 244 ) found that hemin alone could activate lipid oxidation in the washed c od system, and hemin contributes to the initiation of lipid oxidat ion, but not the rate at which lipid oxidation occurs. Metmyoglobin is believed to be pro-oxidative due to a lower a ffinity for hemin and grea ter likelihood of hemin release. The Spanish mackerel pressed juice may have inhibited lipid oxidation by preventing hemoglobin oxidation, dissociation, and/or heme loss. The Effect of pH Adjustment on Tissue Components The pH effect on hemoglobin No significant difference was observed betw een the treatment in which hemoglobin was adjusted from pH 3 to 6.8 and th e control (Figure 5-1a,b). Previous studies would suggest that at low pH, the Spanish mackerel hemoglobin beca me more prooxidative. Lowering the pH of hemoglobin has been shown to activate fish hemoglobin ( 12, 35, 43, 51, 162 ). Lowering to pH 6 from neutrality (pH 7) results in increase d hemoglobin pro-oxidative activity due to the formation of deoxyhemoglobin as found by measur ement of lipid oxidation in a washed cod

PAGE 126

126 model system ( 206 ). The process of oxygen loss from hemo globin as the pH is lowered is known as the Bohr effect ( 50 ). A change in subunit dissociation could increase access and enable for hydroperoxide decomposition and autooxidation, produc ing potent ferryl-Hb radicals and even heme dissociation. Methemoglobin dissociation is 60 times more likely compared to oxy-and deoxyhemoglobin when exposed to pH, resulting in the release of hemin wh ich can interact with membrane layers due to its polar ity. In addition, the lowering of pH forces the heme-iron outside the plane of the protein In the case of heme disso ciation into dimers from tetramers, it has been found that autooxidation occurs 16 times faster ( 254 ). The increase of pr o-oxidative activity is due to the partial unfolding of the hemoglobi n, giving rise to a molten globular state ( 12 ). In another study, when hemoglobin was exposed to extreme pH (pH 2.5) versus neutral pH (7), lipid oxidation proceeded very quickly in a washed cod model system. This increase in lipid oxidation was attributed to vari ous property changes in the hem oglobin that allowed for greater association with the phospholip ids: unfolding, increased hydrophobi city, and detachment from the proximal histidine group ( 230 ). As dissociation of hemoglobin with pH adjustment is likely to occur, the specific affect of pH on myoglobin should also be considere d. At neutral pH (6.2), metmyoglobin is a potent oxidative catalyist. At acidic pH and in the presence of hydroperoxides, metmyoglobin is ve ry effective catalyst of lipid oxidation than at neutral pH. This was evident in increased levels of lipid oxidation in linoleate acid emulsion ( 255 ). Mechanistically, pH lowering can cause oxy-hemoglobin (Fe2+-O2) to auto-oxidize to Met (Fe3+O2), releasing superoxide ( 256 ). Further mutation can then resu lt in the formation of hydrogen peroxide. Hydrogen peroxide activates Met Hb/Mb to two hypervalent species, perferrylmyoglobin/hemoglobin and ferrylmyglobin/hemoglobin. Perferrylmyoglobin is a very transient species and its pro-oxida tive and initiative act ivity is negligible at pH values 5.5-6.5

PAGE 127

127 ( 211 ) and is quickly converted to the most stable hypervalent species, ferryl-Mb/Hb. In fresh meat (pH 5.5-5.8), the ferryl-Hb/Mb species are very short lived and are quickly autoreduced back to the deoxy-Mb/Hb; however, the ferryl-Hb/ Mb species is deemed very pro-oxidative and deemed the primary inducer of lipid oxidation. With aged fish, the antioxidative defenses dwindle, and the ability of this species to initiate lipid oxidation could occur with ease when subjected to reduced pH and lip id concentration; however, in a model linoleate system is was found that the met-Hb form was as oxidative as the ferryl-Hb form ( 47 ). In contrast to the ferryl species, it is observed that the oxidative potential of oxy/deoxy/ and met-Hb is dependant upon hydrogen peroxide concentration ( 43 ). This accumulation of the met-form occurs with drops in post-mortem pH and or acid-aid processing ( 43, 257 ). In the presence of hydroperoxides, Me t-Hb catalyzed lipid oxidation occurs and gives rise to products th at can propagate oxidation ( 46, 258, 259 ). Studies have found that met and deoxyhemoglobin/ myoglobin to be the more pro-oxidant forms versus oxymyoglobin/hemoglobin ( 43, 47, 206 ). Oxygen release from the heme protein may provide greater access to oxidative substrates a nd or reactive oxygen species ( 52 ). It is evident that prooxidative effects by lowering pH were not evident in Hb (p H 36.8). In contrast to various studies, an increase in prooxidative activity with pH lowering was not evid ent in the hemoglobin (pH 3-6.8) treatment (Figure 5-1a,b). Thus, the readjustment to pH 6.8 or other factors affected the catalytic activity of hemoglobin. Hemoglobin adjustment to low pH and readjustment to pH 6.8 In the presence of pressed juice and the s ubstrate (WFM), hemogl obin (pH 3-6.8) proved ineffective in promoting lipid oxidation (Fi gure 5-1a,b). The onset of rancidity occurred significantly earlier in a hemoglobi n mediated cod system adjusted to pH 2.5 compared to pH 3 and 3.5 in a study conducted by Kristinsson and Hultin ( 60 ). It was found that the rate of lipid

PAGE 128

128 oxidation was greater when trout hemoglobin, adjust ed to pH 2.5 and then readjusted to pH 7, was added to washed cod compared to when hemoglobin and washed cod were pH adjusted together. The main difference between the current study and that which was conducted by Kristinsson ( 152 ) at low pH followed by neutralization of blood, is that the current experiment was conducted in the presence of pressed juice. Th e results with Spanish mackerel thus clearly demonstrate how effectively the PJ functions as an antioxidant. Kristinsson and Hultin ( 51 ) investigated the changes in trout hemoglobin conformation with pH adjustment. UV-spectra revealed significant heme peak de gradation and blue shifting as well and reduced oxygen peaks at low pH. Hemogl obin readjusted to neutrality from low pH retained the conformation cha nges and remained somewhat unfolded. These conformational changes are in part indicative of autoxidation, but also unfolding of the protein molecule ( 60 ). The autoxidation of hemoglobin with pH lowe ring provide evidence for the increased prooxidative activity of hemoglobin compared to native hemoglobin (pH 7) in a washed model system. The increased pro-oxidative activity at low pH and with readjustment could possibly explain a quicker onset of oxidation ( 152 ). Hemoglobin aggregation was observed at low pH and after pH readjustment. According to Kristinsson ( 152 ), aggregation due to increases in hydr ophobicity at low pH, due to unfolding, could encourage interactions between hemogl obin and phospholipids within the washed fish matrix. Kristinsson and Hultin ( 51 ) found that despite greater rates of lipid oxidation, peak levels of oxidation with hemoglobin adjusted to low pH followed by neutral pH adjustment did not exceed oxidation at neutral pH in a washed cod sy stem. This could be due to aggregation of the hemoglobin upon partial refolding, which decreased access of the hemoglobin to substrates. The pro-oxidative activity of hemoglobi n is negligible at alkaline pH (10.5-11) in a washed muscle

PAGE 129

129 system and lipid oxidation was significantly suppre ssed compared to the neutral pH in the wash fish Hb-initiated oxidation system ( 51 ). Similar to studies conduced by Kristinsson and Hultin ( 218 ), the Spanish mackerel hemoglobin could have indeed been more pro-oxidative with acid treatment and readjustment to neutral pH; howev er, either deleterious aggregation due to one hour in this study versus 20 minutes in th e study conducted by Kristinsson and Hultin ( 51 ) prevented interaction with the washed fish matrix. Kristinsson and Hultin ( 199 ) have shown that longer holding times at low pH prior to readjust ment significantly reduced hemoglobins ability to refold and leads to reduced solubility. He moglobins propensity to aggregate is due to increased exposure of hydrophobic groups as well and increased release of heme/hemin. Heme can interfere with refolding of heme proteins ( 260 ). Heme has a high affinity towards an aggregative state. The pH treat ed hemoglobin solution was most likely a mix of species that easily aggregated ( 261 ). A possible species would incl ude the permanently denatured hemoglobin known as a hemichrome, which may have formed with a one hour holding time at pH 3. Severe aggregation woul d inhibit interaction with the washed system. Additionally, pressed juice may have contributed to the neglig ible effects elicited by pH treated hemoglobin by quenching any existent pro-oxidative activity. Pressed Juice adjustment To date, published studies on the pH adjust ment from extreme pH of a hemoglobin mediated WFM muscle system, with added PJ, do not exist. It was observed that adding the pH adjusted PJ (3-6.8) and PJ (3-6.8) + Hb (3-6.8) to the washed fish muscle resulted in equivalent levels and rates of oxidation (Figures 5-1b, 5-2 b). All treatments containing readjusted pressed juice had greater overall values of oxidation rela tive to the control; how ever, all values were comparably less than the WFM (3-6.8) and the NPJ (pH 6.8) treatments (Figures 5-1b, 5-2 b). The observation that treated pressed juice compar ed to untreated pressed juice resulted in

PAGE 130

130 significantly higher levels of oxi dation shows that the antioxidativ e effects are inhibited and less able to control hemoglobin mediated lipid oxid ation. Additionally, pH adjusted hemoglobin did not significantly effect the overa ll TBARS levels as the PJ (3-6.8) and PJ (3-6.8) + Hb (3-6.8) were equivalent. Similarly, all other treatment s in which pH adjusted hemoglobin was added were very low and were comparable to levels found in the control. In the pressed juice treatments, further evidence of re duced and/or inhibited pro-oxida tive status of hemoglobin is presented. Undeland et al. ( 155 ) found that heating of presse d juice suppressed or destroyed antioxidative capacities in chicken and herring pr essed juice compared to cod pressed juice. Reduced antioxidative capacities in chicken and herring pressed juice upon heating were attributed to higher levels of hemoglobin (3.2-11.6 uM) compared to cod pressed juice (1.8 uM). It is possible that through the he ating process, iron catalyst such as heme and iron are released ( 262 ), and thus the antioxidative ba lance is shifted toward a more pro-oxidative one. Rong et al. ( 246 ) found that heated chicken pressed juice wa s significantly less effective in inhibiting oxidation in a cod hemoglobin mediated system. By chelating low molecular weight iron in the heated pressed juice before addition to the wa shed muscle, a reduction in TBAR formation by 23% was observed ( 246 ). In this study, the pH treatment of pressed juice could activate low molecular weight iron and hemoglobin inheren tly present and provide less lipid oxidation protection as observed in the pres sed juice treatments: PJ (3-6.8) and Hb (3-6.8) +PJ (3-6.8) juice treatment. Undeland et al. ( 155 ) investigated the effects of pH in a washed fish system. Decreasing the pH of the system with the inclus ion of pressed juice from pH 6.6 to 6.0 reduced the effectiveness of the pressed juice. At pH 6.6 versus 6.0, the lipid oxidation lag phase of hydroperoxides was reduced from 5 to 2 days. Similarly, pH adjustment from pH 3 and

PAGE 131

131 readjustment to pH 6.8, significantly reduced the antioxidative properties of Spanish mackerel pressed juice. It is possible that at low pH the protei ns contained within th e pressed juice could undergo protein oxidation and resulting protein ra dicals could propagate si gnificant increases in lipid oxidation, relative to th e untreated pressed juice cons tituents. Ostdal et al. ( 211 ) found that oxidized BSA radicals effectively initiated ox idation in the presence of other molecules. Albumin was effectively oxidized by ferrylmyoglobin at low pH in a model system as indicated by increases in dityrosine ( 211 ). Pressed juice does cont ain blood components, including albumin and other proteins. The increases in pro-oxi dative activity in the PJ (3-6.8) + Hb (3-6.8) treatment, as hemoglobin did not effectively pr omote oxidation, and decreases in antioxidative capacity in the PJ (3-6.8) treatment could be in part due to the forma tion of reactive protein radicals in the pressed juice upon pH adjustment and readjustment. Lipid oxidation in the adjusted washed fish muscle It is important to address changes in the ma in constituents of washed fish with pH. The washed cod muscle tissue should be virtually fr ee of proand antioxidants and is mainly composed of phospholipids and myofibrillar proteins ( 244 ). In various studies, the susceptibility of the muscle to lipid oxidation is determined by the pro-oxidants and available membrane lipids. In general, fatty fish store their lipid (triacylg lycerol deposits) intraand intercellularly; while lean fish store their main sources of lipid in the liver. Unlike triacyglyerol content, phospholipids maintain a consistent presence ( 0.5-1% w/w) among fish species ( 25 ). According to Pazos et al. ( 161 ), the washed cod system utilized for hemoglobin mediated lipid oxidation was less than one percent lipid. Phospholipids comprised 86% of the total lipids present ( 161 ). In a washed cod system (0.1% fat), aromas indicative of lipid oxi dation were observed. When increasing the phospholipids content by six times, the rate no r the extent of rancidity was significantly affected. However, no rancidity developed when hemolysate was added to a myosin

PAGE 132

132 preparation, emphasing that there needs to be some lipid for the oxidation to occur (but very low levels) ( 58 ). Furthermore, an enzyme iron-catalyzed system was used to initiate oxidation in microsomes. The phospholipid oxidation increased the rate of co-suspended triacyglycerol oxidation ( 236 ). In another study ( 57 ), triacyglycerols in the form of menhaden oil and hemoglobin were added to a washed cod model syst em. No significant differences in the rates of oxidation were observed with or without menhade n oil. In addition, the levels of hemoglobin correlated with extents and rates of lipid oxidation. When hemoglobin was not added to the washed system, lipid oxidation proceeded very slowly; while doubling the amount of hemoglobin doubled the extent of oxidation, and d ecreased the lag phase. In fact, the samples containing only phospholipids oxidized faster th an the treatments with the menhaden oil added ( 57 ). Hemoglobin may preferentially oxidize phos pholipids due to: 1) the surface area of phospholipids and 2) the polar nature of phospholip ids allowing for water soluble hemoglobin to acquire better access and interaction (Borst et al, 2000). From these studies, the washed fish muscle matrix serves as an appropriate subs trate for hemoglobin mediated lipid oxidation studies. The greatest levels of oxidation, aside from th e NPJ treatment formed in the washed fish muscle (WFM) system that was first adjusted to pH 3 for one hour and then readjusted to pH 6.8. TBARS levels were negligible and comparable to the control, when hemoglobin was treated with and without WFM. These results provide evidence for significant changes in washed fish muscle that make it less resistant to oxidation. Confor mational changes occurring in the washed fish muscle most likely facilitate increased native hemoglobin association wi th the less resistant WFM (3-6.8).

PAGE 133

133 Pazos et al. ( 161 ) also tested the affects of adjusti ng a microsomal cod system to pH 3.5 (30 minutes) followed by readjustment to pH 6.0 and 8.0 in the presence of 5 M cod hemoglobin. Results indicated that there was greater ra te and onset of oxidation in the control samples (pH 6.0 and 8.0) than in the microsomal samples first adjusted to pH 3.5. Microsomal fractions at pH 6.0 and 8.0 incurred greater levels of oxidation than at pH 3.5 with adjustment to neutral pH. Less lipid oxidation occurred with pH lowering before readjustment to pH 8.0 compared to pH 6.0. This lowering affect is somewhat analogous to the significantly reduced levels of oxidation in the WF (3-6.8) treatment, compared to the NPJ treatment (Hb and washed fish adjusted to pH 6.8). It was suggested by Pazos et al. ( 161 ) that pH effects on the st ructure of both the blood and the microsomes play a role in a reduced levels of lipid oxidation, with previous adjustment to pH 3. Additionally, in the same study ( 161 ), a washed cod model system was studied with the addition of 3 mol of hemoglobin/kg of tissue. Lipid oxidation results from the washed cod muscle system varied greatly from the microsomal model system. In the instance of washed cod, oxidation at pH 3.5 proceeded very quickly and to a greater extent compared to pH 6.8. Washed cod at pH 3.5 and pH 6.8 proceeded faster than sy stems adjusted to pH 7.8. Interestingly, peak levels of oxidation were equi valent among pH treatments. Th ese findings by Pazos et al. ( 161 ) concur with results found by Kristinsson and Hultin ( 51 ), in which a washed cod muscle system with trout hemoglobin were adju sted to low pH 2.5 and neutral pH. The washed cod system developed significantly highe r levels of oxidation at pH 2.5 than at pH 7 ( 51 ). Interestingly, Pazos et al. ( 161 ) found that microsomes adjusted to pH 3.5 for 6 hours in the absence of hemoglobin lost significantly more sulfydryl groups than other pH treatments. In the presence of hemoglobin, TBARS trends and loss of sulfhydryls within the microsome sy stem correlated well

PAGE 134

134 with all pH treatments, excluding pH 3.5. The lack of correlation in indi cators of lipid and protein oxidation could be i ndicative of phospholipid changes that occur at low pH ( 161 ). Changes within the phospholipids structure coul d explain the high levels of lipid oxidation observed in the WFM (3-6.8) treatment. The overall greatest rate and extent of lipid oxidation among all pH adjustments, occurred in the WFM (3-6.8) treatment (Figure 5-1a,b). Kristinsson and Hultin ( 18 ) addressed the affects of adjusting the washed fish to pH 2.5 and th en to pH 7, followed by the addition of untreated trout hemoglobin. Washed cod that was acidified for 20 minutes followed by pH 7 readjustments had a shorter lag phase than at pH 7. In the in stance of alkaline adjustment and readjustment to neutrality, reduced susceptibility to oxidation occurred in the washed cod compared to the control. When hemoglobin was alkaline treated and readjusted to pH 7 and added to washed muscle, equivalent rates and levels of oxi dation occurred, compar ed to the control ( 18 ). Changes in susceptibility to lipid oxidation when washed fish was treated, substantiates evidence that the washed fish (phospholipids and myo frillar proteins) play an esse ntial role in lipid oxidation development with pH adjustment. Studies on the effects of myofibrillar and myosin proteins may give insight as to why the pH treated washed fish matrix would result in greater lipid oxidation. Kristinsson and Hultin ( 18 ) investigated the affects of low (2.5) and high pH (11) adjustment and readjustment on myosin and myofibrillar proteins. The myofibrillar fraction is obt ained through three washes and centrifugation to remove soluble compounds ( 230 ), including sarcoplasmic proteins. The myofibrillar fraction is equivalent to a washed cod system. Myosin and myofibrillar proteins experienced improved functionality (gelation, solubility, and emulsification) with pH readjustment from low and high pH to pH 7.5. In creased functionality was due to conformational

PAGE 135

135 changes within the protein structure including increased hydrophobicity and surface activity. In most cases, the myosin proteins revealed slight ly greater increases in functionality with pH readjustment ( 18 ). These results may explain the greater levels of lipid oxidation in the WFM (36.8). Changes in protein conforma tion may affect the interaction between proteins and lipids. Conformational changes within th e WFM may enable hemoglobin to better interact with the WFM and effectively promote oxidation. Grunwald and Richards ( 244 ) summarized the heme protein mediated lipid oxidation mechanism in washed cod muscle. Metmyoglobin has a low affinity for heme relative to ferrous heme. Hemin, released from the globin, will pref erentially interact with phospholipids through hydrophobic interaction ( 263 ). Dissociation of hemoglobin will cause more oxidation, due to low hemin affinity of subunits compared to intact hemoglobin or myoglobin ( 252, 264 ). These findings substantiate hemoglobins proxidative ability in a WFM model system. Hydrophilic portions of the hemin can also further the inter action with phospholipids by binding to the polar head groups ( 263 ). Miyazawa ( 265 ) has reported the presence of preformed lipid hydroperoxides in phospholipid membranes, in vivo. Lipid hydroperoxides and hemin will react to form various reactive oxygen species (ROS). These ROS can further oxidize phospholipids by extracting hydrogen. Additionally, hemin is known to degrade upon reactions with ROS ( 244 ). It is further concluded that hemin is a reactant versus a ca talyst during lipid oxida tion. This was concluded because hemin can be degraded by lipid oxid ation products and that lipid oxidation is not proportional to increases in hemoglobin concentr ations, in a washed cod muscle system ( 154 ). Antioxidant Capacity as related to oxidation in washed muscle Antioxidant potential of the hemoglobin and pressed juice added to the washed Spanish mackerel were assessed to possibly elicit reas ons as to why there ar e differences in lipid oxidation among washed Spanish mackerel pH treatments. The oxygen radical absorbance

PAGE 136

136 capacity (ORAC) and Trolox equivalent antioxidant capacity (TEAC) test were used to generate an antioxidant profile. Antioxidant mechanisms generally fit into one of two categories: hydrogen atom transfer (HAT) or single electron transport (SET). HAT mech anisms measure the ability of antioxidant(s) to donate a hydrogen atom to free radicals; wher eas, SET reactions measure the ability of an antioxidant to donate an elec tron and reduce various compounds including, metals, carbonyls, radicals, etc ( 266 ). HAT and SET mechanisms occur simu ltaneously in oxidative reactions. The extent to which each occurs is dependant upon reaction environment as well as antioxidant properties. The oxygen radical absorbance capacity (ORAC) me thod is a HAT based pr otocol. It is an in vitro assay, modeling antioxida nt-lipid interactions in food and physiological environments. Hydrophilic protocols are used quite extensively, and hydrophobic ORAC methods are currently being adapted. ORAC protocols involve analysis over long dura tions (> 1 hour); whereas SET reactions, such as the Trolox equivalent anti oxidant (TEAC) capacity assay, require short incubation times (4-6 minutes). TEAC, or AB TS (2,2-azinobis (3-e thylbenzothiazoline-6solfonic acid) assays measure antioxidant scav enging capacity for a long lived radical anion, ABTS +. The TEAC method is food and physiologica lly appropriate and can be used with varying conditions and media. Specifically, the assay can detect antio xidative capacity with changes in pH and can be used for both hydrophilic and hydrophobic an tioxidant assessment ( 267 ). The accumulated effects of various aq ueous and lipid antioxidants are measured including: vitamins, proteins, lipi ds, glutathione, ur ic acid, etc. ( 268 ). TEAC reactions may not concur with slower reactions, such as ORAC. Short incubation times may result in low TEAC values because the endpoint maybe prematurely measured for certain an tioxidants and/or media

PAGE 137

137 being assessed. Quantitatively, this assay has been given little merit acco rding to Van den Berg and Haenen( 269 ), due to lack of standardized endpoi nts; however, this assay could provide information for comparative analysis of samples and/or antioxidants ( 269 ). Due to the limitations of each assay, a discussion of ge neral trends observed with each test for Spanish mackerel blood and pressed juice follows. Initial hemoglobin and pressed juice samples in creased in antioxidant activity with pH adjustment from physiological pH (6.08 and 6.6) correspondingly, to pH 6.8 (Figures 5-3 a,b). These results concur with previous findings. As pe r the previous discussion of the pH affect on hemoglobin, reduced pro-oxidative activity occu rs with increases in pH above pH 6 ( 12, 42, 51, 161, 206 ). According to Antonenkov and Sies ( 270 ), rat liver cytosol optimally prevented lipid oxidation in a microsome system between pH 6.2 and 6.8. After a one hour holding period, pH 6.8 hemoglobin and pressed juice increased or re tained their antioxidativ e potential as observed in the TEAC and ORAC Assays (Fig ures 5-3a,b). Undeland et al. ( 155 ) observed a greater inhibition of lipid oxidation in a washed cod model system when cytosol was added at pH 6.6 compared to pH 6. When pH was lowered, the OR AC assay (Figure 5-3 a) revealed a significant reduction in ant-oxidative capacity for the Spanis h mackerel hemoglobin a nd pressed juice. This reduction is most likely attributed to increase in pro-oxidative ac tivity with decreases in pH and aggregation of proteins at pH 3. Denaturation and a ggregation most likely inhibit mobility and antioxidative activity of pr oteins in the pressed jui ce. Antonenkov and Seis ( 270 ) found the antioxidative assessment of rat cytosol unrelia ble at pH values below 5.8, due to protein aggregation. Antioxidative capac ity was regained in the hem oglobin and pressed juice with readjustment to pH 6.8 (Figure 5-3 a, b). It is possible that the denature d proteins could refold and thus serve as greater radical scavengers w ith readjustment to pH 6.8 compared to pH 3.

PAGE 138

138 The additive effects of pressed juice and hem oglobin combinations were equal to or higher than treated components alone. The combined mi xture of hemoglobin (pH 6.8) and pressed juice (pH 6.8) possessed high antioxidant potential. This observation confirms the low levels of lipid oxidation in the control treatment for the washed system. The relatively lower levels of lipid oxidation observed in the pressed juice (3-6.8) + Hemoglobin (6.8), compared to the WFM (36.8) treatment, parallel a relatively high level of antioxidant potential (Figure 5-2b). There seemed to be little correlati on between the different combin ations of pressed juice and hemoglobin and the level of lipid oxidation in the washed system. Extracts/samples from the entire washed system is necessary to determine th e levels of lipid oxidati on. As realized, the state of the washed system significantly affects inte ractions with the pre ssed juice and hemoglobin. Analysis of the WFM model system was attemp ted; however, results were undetectable or inconsistent. The problems were most likely due to dilutions or protein insolubility. Thus, methods and/or extraction techniques to better assess the antioxidati ve status in the whole system are necessary Antioxidative potential for each treatment did vary among the TEAC and ORAC assay; however, similarities among the assays was observ ed. Variability among assays could be due to various factors regarding assay function: si mplicity, required instrumentation, biological suitability, mechanisms, incubation and endpoint s, and lipophilic and/ or hydrophilic affinity ( 268 ). Differences among the aqueous system vs. the washed model system are that protein aggregates either in the hemoglobin through pH ad justment to pH 6.8 or adjustment from 3 to 6.8, must be centrifuged out prior to assessment. Th ese aggregates could still possibly play a pro or antioxidative role in WFM hemoglobin mediated lipid oxidat ion. The aggregation may reduce

PAGE 139

139 the overall antioxidative potential of treatments, but at the same time remove prooxidative compounds from the system. Lastly, addition of the pressed juice and hemoglobin combinations into the washed system resulted in 4 fold diluti on of pressed juice. Components in the aqueous system were in the same concentration used in the washed system. Pressed juice dilution still proved to be effective in dela ying or preventing oxidation in the washed system; however, the antioxidative profiles of the hemoglobin and pr essed juice are most likely reduced in the washed system compared to the wet system, due to dilution. Conclusion In conclusion, the role of hemoglobin as a reactant and pro-oxidant in a washed Spanish mackerel model system was demonstrated in th e NPJ (Hb (6.8) +WFM (6.8)) treatment, having the highest levels of lipid oxidation (PV, TBARS) compared to the control and pH treatments. The antioxidant activity of Spanish mackerel pre ssed juice was evident as the control (Hb (6.8) +PJ (6.8) +WFM (6.8)) had significantly lower le vels of lipid oxidation (PV, TBARS) compared to all other treatments. The levels of oxidation in the Hb (3-6.8) treatment were not significantly different from control; thus, it is likely that pro-oxidant activity of hemoglobin observed in the NPJ treatment was lost with adjustment from pH 3 to 6.8. The WFM (3-6.8) treatment had the highest levels of oxidations among thepH treatments. This finding pointed to significant changes within the phospholipids, enabling for greate r interaction with he moglobin. A significant increase in lipid oxidation with pH adjustment of Spanish mackerel PJ, compared to the control, provides evidence that the antioxidative activity of pressed juice wa s lost. Additionally, treatments containing adjusted pressed juice had equivalent levels of lip id oxidation. This result shows that the pH adjustment of pressed ju ice has a significant effect on lipid oxidation development in a WFM system. Low levels of li pid oxidation in the control agreed with the high levels of antioxidant activity in the aqueous system, as measured by the TEAC and ORAC

PAGE 140

140 methods. The results show that lipid oxidation incurred during acidic processing could be primarily due to phospholipids changes, secondary to changes in hemoglobin and/or aqueous extracts. Means to retain the antioxidative capacity of fish cytosolic extract with alkali and acidic processing needs further investigat ion. The effects of low pH and readjustment to neutral pH on pressed juice, hemoglobin, and m yofibrillar extract (washed fish) warrant further investigation.

PAGE 141

141 0 20 40 60 80 100 120 140 160 01234567 Time (days)PV ( mmol LPH/kg tissue) Control (pH 6.8) Control (pH 6.8) w/o pressed juice Hemoglobin (pH 3-6.8) Pressed juice (pH 3-6.8) Washed fish muscle (pH 3-6.8) 0 10 20 30 40 50 60 70 80 90 100 01234567 Time (days)TBARS ( umol MDA/kg tissue) Control (pH 6.8) Control (pH 6.8) w/o pressed juice Hemoglobin (pH 3-6.8) Pressed juice (pH 3-6.8) Washed fish muscle (pH 3-6.8) Figure 5-1. The role of acified (pH 3) and pH 6.8 tissue compone nts on lipid oxidation (A) Lipid hydroperoxides and (B) TBARS in a washed model system. Each treatment was a mixture of hemoglobin, and cytosolic extr act (Pressed Juice), and washed muscle from Spanish mackerel, except for the cont rol without pressed juice. One of three components for each treatment was first adju sted to pH 3 for one hour, followed by readjustment to pH 6.8. The control sample components were subject to a pH of 6.8. Samples were stored for 7 days at 4C. E ach sampling point is representative of the mean and standard deviation from triplicate analysis. (a) (b)

PAGE 142

142 0 20 40 60 80 100 120 140 160 01234567 Time (days)PV ( mmol LPH/kg tissue) Control (pH 6.8) (Hemoglobin + Washed fi sh muscle) pH 3-6.8 (Hemoglobin + Pressed juice) pH 3-6.8 (Pressed juice + Washed fish muscle) pH 3-6.8 (Hemoglobin + Pressed juice+ Washed fish muscle) pH 3-6.8 0 10 20 30 40 50 60 70 80 90 100 01234567 Time (days)TBARS ( umol MDA/kg tissue) Control (pH 6.8) (Hemoglobin + Washed fish muscle) pH 3-6.8 (Hemoglobin + Pressed Juice) pH 3-6.8 (Pressed juice + Washed fish muscle) pH 3-6.8 (Hemoglobin + Pressed Juice + Wa shed fish Muscle) pH 3-6.8 Figure 5-2. The role of two and three acified (p H 3) and neutralized (6 .8) tissue components on lipid oxidation (a) Lipid hydroperoxides a nd (b) TBARS in a stored washed model system. Each treatment was a mixture of hemoglobin, and cytosolic extract (Pressed Juice), Two to three components for each tr eatment were adjusted to pH 3 for one hour, followed by readjustment to pH 6.8. The control sample components were subjected to a pH of 6.8. Samples were stored for 7 days at 4C. Each sampling point is representative of the mean and standa rd deviation from triplicate analysis. (b) (a)

PAGE 143

143 f g efg c defg ef c a d de f b0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8Hb (pH 6 .08) PJ (pH 6 .6) H b ( pH 3 ) Hb (pH 6 .8) PJ (pH 3) P J ( pH 6.8) H b ( pH 3 -6.8) PJ (pH 3 -6.8) P J ( pH 3-6. 8) +H b ( pH 6.8) P J (pH 3-6. 8) + Hb (pH 3-6. 8) PJ ( pH 6 .8) +Hb (pH 6.8 ) PJ (pH 6.8)+ Hb (pH 3-6.8)TreatmentsORAC (umol Trolox equivalents/ mg protein) 0 hour hold 1 hour hold bc e de e e cde a a bc b a a a bcd0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45H b (pH 6.08) P J (pH 6.6) H b (pH 3) H b (p H 6.8) PJ ( p H 3) PJ (pH 6.8) Hb (pH 3-6.8) P J (pH 3-6.8) P J (pH 3-6.8 ) +Hb (pH 6. 8) PJ ( p H 3 -6.8) + H b ( pH 3 -6.8) PJ (pH 6.8) +Hb (pH 6.8) P J (pH 6.8)+ H b (pH 3-6.8)TreatmentsTEAC (mM Trolox equivalents/ mg protein) 0 hour hold 1 hour hold Figure 5-3. Antioxidant capacity (A) ORAC and (B) TEAC of Span ish mackerel hemoglobin and pressed juice alone or combination with ad justment to pH 3 for one hour followed by readjustment. Several combinations were im plemented. Each bar is representative of the mean and standard deviation from tr iplicate analysis. Different small letters indicate a significant different (p< 0.05). (A) (B)

PAGE 144

144 1 2 3 4 5 6 Figure 5-4. Protein composition of untreated Spanish mackerel pressed juice. Lanes 1 and 6: molecular standards (5 uL). Lanes 2 and 5 represent Spanish Mackerel pressed juice (5 uL). Lanes 3 and 4 represent Spanis h mackerel homogenates at 10 and 15uL. 205 kDa 116 kDa 97 kDa 84 kDa 66 kDa 55 kDa 45 kDa 36 kDa 29 kDa 24 kDa 20 kDa 14 kDa

PAGE 145

145 CHAPTER 6 CONCLUSION The acid/alkali-aided process promotes uti lization of by-catch a nd dark muscled fish species for human food. Dark muscle fish species pose various processing challenges due to the abundance of unstable lipids and heme proteins. To investigate the effects of the acid/alkali aided processing on lipid oxidation in dark muscle fish species, Spanish mackerel was used as a model. The effects of acid and alkaline processing on lipid oxi dation in Spanish mackerel varied among homogenates, isolates, and the washed fish muscle system. Lipid oxidation was prevalent at low pH compared to high pH in the Spanish mackerel homogenates. The initial quality of the Spanis h mackerel, storage, and pH holding times all significantly affected lipid oxida tion development in the homogenate s. Addition of ascorbic acid and tocopherol to the acid homogenates, eff ectively inhibited lipi d oxidation, but was not affected by the addition of chelat ors, EDTA and citric acid. The pH readjustment from acidic to neutral pH did not significantly affect the li pid oxidation in the homogenates; however, the addition of tocopherol (0.1%) and ascorbic acid (0.1%) in combin ation with pH readjustment proved to be highly oxidative. How adding ascorbic acid and tocopherol to the acid and alkali homogenates affects the final proteins isolate needs further investigation. Isolates were prepared from acid and alkaline processing of Spanish mackerel. Lipid oxidation was comparable among the alkaline and acid isolates, with holding time and storage. High levels of lipid oxidation were observed in the supernatant fractions from the acid, alkaline, and control treatments. Alkalin e isolates contained overall great er protein oxidation than the acid and control isolates, as determined by carbo nyl levels. Protein degredation occurring during acid and alkali processing. Rheological testi ng provided information regarding isolate gel strength. Greater gel strength wa s observed in the alkaline isol ates, followed by the control and

PAGE 146

146 acid isolates. Possible correlations between lipid and protein oxi dation and gel strength warrant further investigation. Additionally, the mechan isms by which protein and/or lipid oxidation develop with pH adjustment during the isolation process should be investigated. The addition of antioxidants in the final isolate could prevent lip id and protein oxidation, a nd thus affect isolate quality. The adjustment from pH 3 to pH 6.8 greatly affected all components in a washed Spanish mackerel model system. The pH adjustment significantly inhibited hemoglobins catylitic activity and pressed juice antioxidative activity. The washed fish muscle was significantly affected by pH adjustment. This was evident in significantly higher levels of lipid oxidation in WFM treatment (pH 3-6.8) compared to all othe r pH treatments. The addition of hemoglobin, adjusted from pH 3 to 6.8 to the washed Span ish mackerel muscle, gave lipid oxidation levels comparable to the control. Greater levels of oxi dation were observed in the pH treatments in which pressed juice (3-6.8) was added, compared to the control. The affect of pH adjustment from extreme pH to neutral pH on a washed fi sh muscle system, containing pressed juice, has not been investigated. Thus, there needs to be further understanding of the pH affects on the key components in fish muscle and how this relate s to lipid oxidation development. Methods to better access the antioxidative cha nges in the washed fish muscle require investigation. An understanding of components contributing to the high antioxidative acitivty in pressed juice is also needed. Results from this study point to various ch anges in lipid and pr otein oxidation due to alkaline and acidic treatment. Acid treatment resulted in high le vels of lipid oxidation in the homogenates and in the washed fish model system Acid and alkaline processing proved to be equally lipid oxidative in the prep aration of Spanish mackerel isolat es; however, greater levels of

PAGE 147

147 protein oxidation developed with alkaline treatment. Thus, measures of lipid and protein oxidation are imperative to understanding how fish protein structure and quality is altered.

PAGE 148

148 LIST OF REFERENCES (1) Cushman Jr., J. H., Cuts Sought in Wasteful Fish Kills. New York Times January 13, 1998,p2. (2) Windsor, M.; Barlow, S., Introduction to Fi shery Byproducts. Fishing news (Books) Ltd: London, 1981. (3) FAO, The production of Fish Meal and O il. In FAO Fisheries Technical Paper, No. 142,Rome, 1975. (4) Okada, M., Utilization of sma ll pelagic species for food. In Pr oceedings of the Third National Technical Seminar on Mechanical Recovery an d Utilization of Fish Flesh, Martin, R. E., Ed. National Fisheries Institute: Washington, D.C., 1980; pp 265-282. (5) Hultin, H. O., Oxidation of Lipids in Seafoods. In Seafoods: Chemistry, Processing Technology and Quality, Shahidi, F.; Botta, J. R., Eds. Blacki Academic: Glasgow, 1994; pp 49-74. (6) Hultin, H. O.; Kelleher, S. D., Surimi Pro cessing from Dark Muscle Fish In Surimi and Surimi Seafood, Park, J. W., Ed. Ma rcel Dekker: New York, 2000; pp 59-77. (7) Hultin, H. O.; Kelleher, S. D.; Feng, Y.; Kris tinsson, H. G.; Richards, M. P.; Undeland, I. A.; Ke, S. High Efficiency Alkaline Protein Extraction. 60,230,397, 2000. (8) Hultin, H. O.; Kelleher, S. D. Process for Isolating a Protein Composition from a Muscle Source and Protein Composition. 6,005,073, 1999. (9) Shewfelt, R. L., Fish Muscle Lipoly sisA review. J Food Biochem 1981, 5, 79-100. (10) Gandemer, G., Lipids a nd Meat Quality: Lipolysis, Oxidation, Maillard Reaction and Flavour. Sci. Aliment. 1999, 19, 439-458. (11) Hultin, H. O.; Kristinsson, H. G.; Lanier, T. C., Process for recovery of functional proteins by pH shifts. In Surimi and Surimi S eafood, 2nd ed.; Marcel Dekker: New York, 2004. (12) Kristinsson, H. G., Acid unfolding of flound er hemoglobin: Evidence for a molten globular state with enhanced pro-oxidative activ ity. J. Agric Food Chem 2002, 50, (26), 7669-7676. (13) Kristinsson, H. G.; Demir, N., Functional Fi sh Protein Ingredients from Fish Species of Warm and Temperate Waters: Comparison of Acid and Alkali-Aided Processing vs. Conventional Surimi Processing. In Advances in Seafood Byproducts, Alaska Sea Grant College Program: 2003. (14) Kelleher, S. D.; Feng, Y. M.; Hultin, H. O.; Livingston, M. B., Role of initial muscle pH on the solubility of fish muscle proteins in water. Food Biochemistry 2004, 28, (4), 279-292. (15) Hultin, H. O., Controlling lipid oxidation during processing at high and low pH. In Annual IFT Meeting, Las Vegas, Nevada, 2004. (16) Undeland, I. A.; Kelleher, S. D.; Hultin, H. O ., Recovery of functional proteins from herring (Clupea haerangus) light muscle by an acid or alkaline solubilization process. J. Agric. Food Chem. 2002, 50, 7371-7379. (17) Kristinsson, H. G.; Theodore, A. E.; Demir, N.; Ingatottir, B., Acid-aided and alkali-aided process. Food Chemistry and Toxicology 2005, 70, (4), C298-C306. (18) Kristinsson, H. G.; Hultin, H. O., Effect of low and high pH treatment on the functional properties of cod muscle proteins J. Agric Food Chem 2003, 51, 5103-110. (19) Theodore, A. E.; Kristinsson, H. G., A comp arative study between acidand alkali-aided processing and surimi processing for the recovery of proteins from channel catfish muscle. Journal of Food Science 2003, 70, (4), C298-C306.

PAGE 149

149 (20) Reinitz, G. L.; Orme, L. E.; Hitzel, F. N., Variations of body composition and growth among strains of rainbow trout (Salmo gairdneri). Trans. Am. Fish Soc. 1979, 108, 204207. (21) Kiessling, A. T.; Aasgaard, T. ; Storebakken, T.; Johansson, L.; Kiessling, K. H., Changes in the structure and function of the epaxial mu scle of rainbow trout (Oncorhynchus mykiss) in relation to ration and age. Aquaculture 1991, 93, 373-387. (22) Huss, H. H., Quality and Quality Changes in Fresh Fish. In Fisheries Technical Papers T-34, ed.; FAO: Rome, 1994; p 195. (23) Sevanian, A.; Hochstein, P., Mechanisms and consequences of lipid peroxidation in biological systems. Annu. Rev. Nutr. 1985, 5, 365-390. (24) Sies, H., Biochemistry of oxidative stress. Angew Chem. Int. Ed. Engl. 1986, 25, 1058-1071. (25) Ackman, R. G., Fish Lipids. Part 1. In A dvances in Fish Science and Technology, Fishing News Books Ltd: England, 1980; pp 86-103. (26) Wasasundara, U. N.; Shahidi, F., Canola extract s as an alternat ive antioxidant for canola oil. J. Am. Oil Chem. Soc. 1994, 71, 817-822. (27) Bateman, L.; Hughs, H.; Morris, A. L., Hydroperoxide decomposition in relations to the initiation of radical chain reacti ons. Disc Faraday Soc. 1953, 4, 190-194. (28) Gutteridge, J. M. C., Lipid peroxidation: So me Problems and Concepts. In Oxygen Radicals and Tissue Injury, Federation of American So cieties of Experimental Biology: Bethesda, MD, 1988; pp 9-19. (29) Pearson, A. M.; Gray, J. I.; Wolzak, A. M.; Horenstein, N. A., Safety implications of oxidized lipids in muscle foods. Food Technology 1983, 37, (7), 121-129. (30) Gray, J. I.; Pearson, A. M., Rancidity and warmed-over flavor. Meat. Res. 1987, 3, 221-269. (31) Erickson, M. C., Lipid Oxidation of Muscle Food. In Quality attributes of Muscle Foods, Kluwer Academic: Boston, 1999; pp 365-368. (32) Medina, I.; Saeed, S.; N., H., Enzymatic oxi dative activity in sardine (Sardina pilchardus) and herring (Clupea harengus) dur ing chilling and corre lation with quality. European Food Researh and Technology 1999, 210, (1), 34-38. (33) Mead, J. F., Free Radical in Biology and Medicine. Academic Press: New York, 1976; Vol. 1, p 51. (34) Floyd, R. A.; Lewis, A., H ydroxyl free radical formation fr om hydrogen peroxide by ferrous iron-nucleotide complexes. Biochemistry 1983, 22, 2645. (35) Richards, M. P.; Ostdal, H.; Andersen, H. J., Deoxyhemoglobin-mediat ed lipid oxidation in washed fish muscle. J Agri Food Chem 2002, 50, (5), 1278-1283. (36) Younathan, M. T.; Watts, B. M., Relations hip of meat pigments to lipid oxidation. Food Res. 1959, 24, 728-733. (37) Kanner, J.; Harel, S.; Granit, R., Initia tion of membranal lipid peroxidation by activated methemoglobin and metmyoglobin. Arch. Biochem. Biophys. 1985, 289, 130. (38) Hultin, H. O.; Kelleher, S. D.; Feng, Y.; Kristinsson, H. G.; Richards, M. P.; Undeland, I. A.; Ke, S. High Efficiency Al kaline Protein Extraction. 2000. (39) Rhee, K. S., Natural antioxidants for meat pr oducts. In Warmed-over fla vor of meat, A.J., S. A.; Bailey, M. E., Eds. Academic Press: Orlando, 1987. (40) Asgar, A.; Gray, J.; Buckley, D.; Pearson, A. M.; Booren, A. M., Perspectives on warmed over flavor. Food Technology 1988, 42, (6), 102-108.

PAGE 150

150 (41) Kanner, J.; Salon, M. A.; Harel, S.; Shegal ovich, I., Lipid peroxidatio n of muscle food: the roles of cytosolic fraction. J Agri Food Chem 1991, 39. (42) Richards, M. P.; Hultin, H. O., Effect of pH on lipid oxida tion using trout hemolysate as a catalyst: a possible role for deoxyhemoglobin. J. Agric Food Chem 2002, 48, 3141-3147. (43) Richards, M. P.; Hultin, H. O., Effect of pH on lipid oxida tion using trout hemolysate as a catalyst: a possible role for deoxyhemoglobin. J Agri Food Chem 2000, 48, 3141-3147. (44) Connell, J. J., Control of Fish Qual ity. Whitefriars Press, Ltd.: London, 1975; p 333. (45) Wheaton, F. W.; Lawson, T. B., Processing Aquatic Food products. John Wiley and Sons: New York, 1985. (46) Everse, J.; Hsia, N., The toxicities of na tive and modified hemoglobins. Free Rad. Bio. Med 1999, 22, 1075-1099. (47) Baron, C. P.; Andersen, H. J., Myoblobininduced lipid oxidation: A review. J Agri Food Chem 2002, 50, 3887-3897. (48) Nawar, W., Lipids. In Food Chem., 3 ed.; Fenemma, O. R., Ed. Marcel Dekker Inc.: New York, 1996; pp 225-320. (49) Gutteridge, J. M. C.; Halliwell, B., The measurement and mechanism of lipid peroxidation in biological systems. Trends Biochem. Sci. 1990, 15, 129-135. (50) Weber, R. E., Adaptions of oxygen tran sport: Lessons from fish hemoglobins. In Hemoglobin function in vertebra tes, Prisco, G. d.; Giardina B.; Weber, R. E., Eds. Springer: Verlag, Italia, 2000; pp 23-27. (51) Kristinsson, H. G.; Hultin, H. O., The effect of acid and alkali unfolding and subsequent refolding on the pro-oxidative activity of trout hemoglobin. J Agri Food Chem. 2004, 52, (17), 5482-5490. (52) Levy, A.; Rifkind, J. M., Low-temperature formation of distal histidine complex in hemoglobins: probe for heme pocket flexibility. Biochemistry 1985, 24, 6050-6054. (53) Astatke, M.; McGee, W. A.; Parkhurst, L. J., A flow procedure to determine oxygen binding isotherms for low affinity a nd easily oxidizable hemoglobins Comp. Biochem. Physiol. 1992, 101 B, 683-688. (54) Richards, M. P.; Dettmenn, M. A., Comp arative analysis of different hemoglobins: autoxidation, reaction with peroxide, and lipid oxidation. J Agri Food Chem 2003, 51, (13), 3886-3891. (55) Richards, M. P.; Dettmenn, M. A., Comp arative analysis of different hemoglobins: Autoxidation, reaction with peroxide, a nd lipid oxidation. J Agri Food Chem 2003, 51, (13), 3886-3891. (56) Reeder, B. J.; Wilson, M. T., The effects of pH on the mechanism of hydrogen peroxide and lipid hydroperoxide consumption by myoglobin: A role for the protonated ferryl species. Free Rad. Bio. Med 2001, 30, (1311-1318), 30. (57) Undeland, I. A.; Richards, M. P., Adde d triacylglycerols do not hasten hemoglobinmediated lipid oxidation in washed min ced cod muscle. J Agri Food Chem 2002, 50, 6847-6853. (58) Richards, M. P.; Hultin, H. O., Rancidity development in a fish muscle model system as affected by phospholipids. J. Food Lipids 2001, 8, 215-230. (59) Porter, P. J.; Kennish, J. M.; Kramer, D. E., The effects of ex sanguination of sockeye salmon on the changes of lipid composition duri ng frozen storage. In Seafood Science and Technology, Bligh, E. D., Ed. Fishing News Books: Oxford, 1992; pp 76-83.

PAGE 151

151 (60) Shikama, K., The molecular mechanism of autoxidation for myoglobi n and hemoglobin: an enerable puzzle. Experientia 1998, 98, 1357-1373. (61) Rogers, M. S.; Patel, R. P.; Reeder, B. J.; Sarti, P.; Wilson, M. T.; Alayash, A. I., Prooxidant effects of cross-linked haemoglobins explored using lyposome and cytochrome c oxidase vesicle model membra nes. Biochemistry 1995, 310, 827-833. (62) Undeland, I. A.; Hall, G.; Karin, W.; Ga ngby, I.; Rutgersson, A., Preventing lipid oxidation during Recovery of functional proteins from herring (Clup ea harengus) fillets by an aid solubilization process. J Agri Food Chem 2005, 53, 5625-5634. (63) Howell, N.; Saeed, S., Effect of lipid oxidation and frozen storage on muscle proteins of Atlantic Mackerel. j Science of Food and Agriculture 2002, 82, 579-586. (64) Witting, L. A., Vitamin E and lipid antioxidants in free-radical initiated reactions. In Free Radicals in Biology, Priyor, A., Ed. Acad emic Press: New York, 1995; Vol. IV. (65) Chow, C., Vitamin E and oxidati ve stress. Free Rad. Bio. Med 1991, 11, 215-21. (66) Cadenas, E.; Packer, L., Handbook of antioxi dants. Marcel Dekker, Inc.: New York, 1996; p 545-91. (67) Undeland, I. A.; Hall, G.; Wendin, K.; Gangby, I.; Rutgersson, A., Preventing Lipid Oxidation during Recovery of Functional Prot eins from Herring (Clupea harengus) fillets by an Acid Solubilization Process. J Agri Food Chem 2005, (53), 5625-5634. (68) Kelleher, S. D.; Silva, L. A.; Hultin, H. O., Inhibition of oxidation during processing of washed, minced Atlantic mackerel. J. Food Sci 1992, 57, 1103-1119. (69) Richards, M. P.; Kelleher, S. D.; Hultin, H. O., Effect of washing with or without antioxidants of quality retention of mackerel fillets during refrigerated and frozen storage. J Food Sci. 1998, 46, (4363-4371). (70) Howell, N.; Saeed, S., Rheological and differential scanning calorimetry studies on structural and textural changes in frozen Atla ntic mackerel (Scomber scombrus). J Science of Food and Agriculture 2004, 82, 579-586. (71) Dean, R. T.; Fu, S.; Stocker, R.; Davies M. J., Biochemistry and pathology of radicalmediated protein oxidation. Biochem. J. 1997, 324, 1-18. (72) Stadtman, E. R., Oxidation of free amino acids and amino acid residues in proteins by radiolysis and by metal-ca talyzed reactions. Annu. Rev. Biochem. 1993, 62, 797-821. (73) Davies, M. J.; Dean, R. T., Radical-mediate d protein oxidation: from chemistry to medicine. Oxford Univ. Press: Oxford, 1997. (74) Hawkins, C. L.; Davies, M. J., Generation and propagation of radical reactions on proteins. Biochem. Biophys. Acta 2001, 1504, 196-219. (75) Halliwell, B., Free radi cals, antioxidants and human di sease: curiosity, cause or consequence? Lancet 1994, 344, 721-724. (76) Shacter, E., Quantification and significant of protein oxidation in bi ological samples. In Drug Metab. Rev., 2000; Vol. 32, pp 307-326. (77) Hwang, E.; Kim, G.-H., Review : Biomarkers for oxidative stress status of DNA, lipids, and proteins in vitro and in vivo can cer research. Toxicology 2007, 229, 1-10. (78) Schaich, K. M., Free ra dical initiation in proteins and amino acids by ionizing and ultraviolet radiations and lipid oxidation. III: Free radical tr ansfer from oxidizing lipids. CRC Crit. Rev. Food Sci. Nutr. 1980, 13, (189-244). (79) Kanner, J., Oxidative processes in meat and meat products: quality implication. Meat Sci. 1994, 36, 169-180.

PAGE 152

152 (80) Johns, A. M.; Birkinshaw, L. H.; Ledward, D. A., Catalysits of lipid oxidation in meat products. Meat. Sci. 1989, 25, (209-220). (81) Haberland, M. E.; Fong, D.; Cheng, L., Malondi aldehyde-altered protein occurs in atheroma of watanabe hertable hyperlipi demic rabbits. Science 1988, 241, 215-218. (82) Kim, J. G.; Sabbagh, F.; Santanam, N.; W ilcox, J. N.; Medford, R. M.; Parthasarathy, S., Generation of polyclonal antibody against lipid peroxide-modif ied proteins. Free Rad. Bio. Med 1997, 23, 251-259. (83) Requena, J. R.; Fu, M. X.; Ahmed, M. U.; Jenkins, A. J.; Lyons, T. J.; Baynes, J. W.; Thorpe, S. R., Quantification of malondial dehyde and 4 hydroxynonenal adducts to lysine residue in native and oxidized human low density lipoprotein. Biochem. 1997, 322, 317-325. (84) Dean, R. T.; Fu, S.; Stocker, R.; Davies M. J., Biochemistry and pathology of radicalmediated protein oxidation. Biochem. 1997, 324, 1-18. (85) Berlett, B. S.; Stadtman, E. R., Protein oxidation in aging, disease, and oxidative stress. J. Biol. Chem. 1997, 272, 201313-6. (86) Xiong, L., Protein Oxidation and Implication for muscle food Quality. In Antioxidants in Muscle Foods, Decker, E. A.; Faustman, C.; Lopez-Bote, C. J., Eds. John Wiley and Sons, Inc.: New York, 2000. (87) Xiong, Y. L.; Decker, E. A., Alternations in muscle protein func tionality by oxidative and antioxidative processes. J. Muscle Foods 1995, 6, 139-160. (88) Martinaud, A.; Mercier, Y. ; Marinova, P.; Tassy, C.; Gatelli er, P.; Renerre, M., Comparison of oxidative processed on myofibrillar pr oteins from beef during maturation and by different model oxidation systems. J. Agric. Food Chem. 1997, 45, 2481-2487. (89) Garrison, W., Reaction mechanisms in radi olysis of peptides, polypeptides, and proteins. Chem. Rev 1987, 87, 381-398. (90) Garrison, W.; Jayko, M. E.; W., B., Radia tion-induced oxidation of protein in aqueous solution. Rad Research 1962, 16, 483-502. (91) Swallow, A., Effect of i onizing radiation on proteins, RCO groups, peptide bond cleavage, inactivation,SH oxidati on. John Wiley and Sonds: New York, 1960; p 211-224. (92) Schuessler, H.; Schilling, K., Oxygen effect in radiolysis of proteins. Part 2. Bovine Serum albumin. Int. J Radiat. Biol 1984, 45, 267-281. (93) Radi, R.; Bush, K. M.; Cosgrove, T. P.; Freman, B. A., Reaction of xanthine oxidasederived oxidants with lipid and protein of human pl asma. Arch. Biochem. Biophys. 1991, 286, 117-125. (94) Lii, C. K.; Chai, Y. C.; Zhao, W.; Thomas, J. A.; Hendrich, S., S-thiolation and irreversible oxidation of sulfhydrls on carbonic anhydrase III during oxidative st ress: A method for studying protein modification in intact ce lls and tissues. Arch. Biochem. Biophys. 1994, 308, 231-239. (95) Vogt, W., Oxidation of meth ionyl residues in proteins: tools, targets, and reversal. Free Rad. Bio. Med 1995, 18, 93-105. (96) Levine, R. L.; Mosoni, L.; Berlett, B. S.; Stadtman, E. R., Methionine residues as endogenous antioxidants in proteins Proc Natl Acad Sci 1996, 93, 15036-15,040. (97) Reddy, V. Y.; Desorchers, P. E.; Pizzo, S. V.; Gonias, S. L.; Sahakian, J. A.; Levine, R. L.; Weiss, S. J., Oxidative dissociations of human alpha 2-maroglobulin tetramers in dysfunctional dimers. J. Biol. Chem. 1994, 269, 4683-4691.

PAGE 153

153 (98) Gitlin, G.; Tsarbopoulos, A.; Patel, S. T.; Sydor, W.; Pramanik, B. N.; Jocobs, S., Isolation and characterization of a mono methioninesulfoxide variant of interferon alpha-2b. Pharm Res. 1996, 13, 762-769. (99) Stadtman, E. R.; Levine, R. L., Free radi cal-mediated oxidation of free amino acids and amino acid residues in proteins. Amino Acids 2003, 25, 207-218. (100) Uchida, K.; Kato, Y.; Kawakishi, S., A novel mechanism for oxidative damage of prolyl peptides induced by hydroxyl radicals Biochem. Biophys. Res. Commun 1990, 169, 265-271. (101) Gardner, H. W., Lipid hydrope roxide reactivity with protei ns and amino acids: A review. J Agri Food Chem 1979, 30, 1204-1208. (102) Srinivasan, S.; Xiong, Y. L., Sulfhydryls in antioxidant-washed beef heart surimi. J. Muscle foods 1997, 8, 252-263. (103) Verziji, N.; De Groot, J.; Oldehinkel, E. ; Bank, R. A.; Thorpe, S. R.; Baynes, J. W.; Bayliss, M. T.; Bijlsma, W. J.; Lafeber, F.; Tekoppele, J. M., Age accumulation of Mailard products in human articular cartila ge collagen. Biochem. 2000, 350, 381-387. (104) Lee, H. G.; Lee, C. M.; Chung, H.; Laver y, S. A., Sodium ascorbate affects surimi-gel forming properties. J. Food Sci. 1992, 57, (1343-1347.). (105) Uschida, K.; Stadtman, E. R., Covalent attachment of 4-hydroxynon eal to glyceraldehyde3-phosphate dehydrogenase. J. Biol. Chem. 1993, 268, 6388-6393. (106) Stadtman, E. R.; Oliver, C. N., Metal-catalyzed oxidation of proteins. J. Biol. Chem. 1991, (266), 2005-2008. (107) Stadtman, E. R., Metal i on-catalyzed oxidation of protei ns: Biochemical mechanism and biological consequences. Free Radi cal Biol. and Med. 1990, 9, (315-325). (108) Korashi, R.; Kohen, R.; Katzhendler, J.; Ch evion, M., Iron mediates paraquat toxicity in Escherichia coli. J. Biol. Chem. 1986, 261, 12,472-12,476. (109) Czapski, G., On the use of OH radical scave ngers in biologicals systems. Issrael J. Chem. 1984, 24, 29-32. (110) Samuni, A.; Aronovitch, J.; Godiner, D.; Chev ion, M.; Czapski, G., Cytoxicity of vitamin C and metal ions. A site-specific Fenton mechanism. Eur. J. Biochm. 1983, 137, 119-124. (111) Amici, A.; Levine, R. L.; Tsai, L.; Stadtm an, E. R., Conversion of amino acid residues in proteins and amino acid homopolymers to carbonyl derivatives by meta-catalyzed oxidation reactions. J. Biol. Chem. 1989, 264, 3341-3346. (112) Levine, R. L.; Garland, D.; Oliver, c. N.; A., A.; Climent, I.; Lenz, A. g.; Ahn, B. W.; Shaltiel, S.; Stadtman, E. R., Determination of carbonyl content in oxidatively modified proteins. Methods of En zym. 1990, 186, (464-477). (113) Schaich, K.; Karel, M., Fr ee radical reactions of peroxidizing lipid with amino acids and proteins: and ESR study. Lipids 1975, 11, 392-400. (114) Roubal, W. T.; A.L., T., Damage to pr oteins, enzyme, and amino acids by peroxidizing lipids. Arch. Biochem. Biophys. 1966, 113, 5-8. (115) Stadtman, E. R., Protein ox idation in aging and age-related diseases. Ann. N.Y. Acad. Sci. 2001, 928, 22-38. (116) Coux, O.; Tanaka, K.; Goldberg, A., Structur e and functions of 20S and 26S proteosomes. Annu Rev Biochem. 1996, 65, 801-847. (117) Moskovitz, J.; Berlett, B. S.; Poston, J. M.; Stadtman, E. R., Methionine sulfoxide reductase in antioxidant defens e. Methods Enzymol. 1999, 300, 239-244.

PAGE 154

154 (118) Starke-Reed, P. E.; Oliver, C. N., Prot ein oxidation and proteo lysis during aging and oxidative stress. Arch Bi ochem. Biophys. 1989, 275, 559-567. (119) Carney, J. M.; Starke-Reed, P. E.; Oliv er, C. N.; Landum, R. W.; Chang, M.; Wu, J.; Floyd, R. A., Reversal of agerelated increase in brain pr otein oxidation, decrease in enzyme activity, and loss in temporal and sp atial memory by chronic administration of the spin-trapping compound N-tert-butyl-alpha -phenylnitrone. Proc Natl Acad Sci 1991, 88, (3633-3636). (120) Levine, R. L., Carbonyl m odified proteins in cellular re gulation, aging, and disease. Free Radical Biol. and Med. 2002, 32, (9), 790-796. (121) Srinivasan, S.; Hultin, H. O., Chemical, phys ical, and functional prope rties of cod proteins modified by nonenzymatic free-radical genera ting system. J. Agric. Food Chem. 1997, 45, 310-320. (122) Smith, D. M., Functional and biochemical ch anges in deboned turkey due to frozen storage and lipid oxidation. J. Food Sci 1987, 52, 22-27. (123) Wan, L.; Xiong, Y. L.; Decker, E. A., Inhi bition of oxidation during washing improves the functionality of bovine cardiac myofibrill ar protein. J. Agric Food Chem 1993, 41, 22672271. (124) Xiong, Y. L.; Decker, E. A.; Robe, G. H. ; Moody, W. G., Gelation of crude myofibrillar protein isolated from beef heart under antioxidative conditions. J. Food Sci 1993, 58, (1241-1244). (125) Wang, B. W.; Xiong, Y. L., Functional Stability of an tioxidant-washed, cryoprotectanttreated beef heart surimi during frozen storage. J. Agric Food Chem 1998, 41, 2267-2271. (126) Liu, G.; Xiong, Y. L., Thermal denaturation a nd in vitro digestibility of oxidized myosin. 1997; Vol. 56-4. (127) Stone, A. P.; Stanley, D. W., Mechanisms of fish muscle gelation. Food Research International 1992, 25, 381-388. (128) Srinivasan, S.; Xiong, L., Thermal stability of oxidized mu scle proteins as analyzed by differential scanning calorimentry. 1996 IFT A nnual Meeting: Book of Abstracts Abst. No. 38-5. (129) Li., S. J.; King, A. J., Li pid oxidation and myosin enaturat ion in dark chicken meat. J. Agric. Food Chem. 1996, 44, 3080-3084. (130) Wang, B. W.; Xiong, Y. L.; Srinivasan, S., Chemical Stability of antioxidant washed beef heart surimi during frozen st orage. J. Food Sci 1997, 62, 939-945, 991. (131) Buttkus, H., The reaction of myosin with malonaldehyde. J. Food Sci. 1967, 32, 432-434. (132) Dillard, C. J.; Tappel, A. L., Fluorescent products from the reaction of peroxidizing PUFA with phophtdyl ethanolamine and ph enylalanine. Lipids 1973, 8, 183-189. (133) Decker, E. A.; Xiong, Y. L.; Clavert, J. T.; Crum, A. D.; Blanchard, S. P., Chemical physical and functional properties of oxidized turkey white musc le myofibrillar proteins. J Agri Food Chem 1993, 41, 186-189. (134) Jarenback, L.; Liljemark, A ., Ultrastructural changes duri ng frozen storage of cod. III: Effects of linoleic acid and li noleic acid hydroperoxides on myof ibrillar proteins. J. Food Technol. 1975, 10, 437-452. (135) Xiong, Y. L., Oxidative Processes in m eat. 49th Annual Reciproc al Meat Conference 1996, pp 79-86.

PAGE 155

155 (136) Srinivasan, S.; Xiong, Y. L.; Decker, E. A. Inhibition of Protein and Lipid Oxidation in Beef heart Surimi-like material by Antioxi dants and Combinations of pH, NaCL, and buffer type in the washing medi a. J Agri Food Chem 1996, 44, 119-125. (137) Liu, G.; Xiong, Y. L., Rheological properti es of oxidatively modified muscle, whey and soy protein composite systems. 1998 IFT Annua l Meeting: Book of Abstracts, Abst. No. 46C-1. (138) Eymard, S.; Carcouet, E.; Rochet, M.; Duma y, J.; Chopin, C.; Claude, G., Development of lipid oxidation during manufacturing of horse mackerel surimi. J Science of Food and Agriculture 2005, 85, 1750-1756. (139) Undeland, I. A.; Richards, M. P.; Hultin, H. O., Added triacylglycerols do not hasten hemoglobin-mediated lipid oxidation in washed minced cod muscle. J. Agric Food Chem 2001, 50, 6847-6853. (140) Lee, C. M.; Trevino, B.; Chaiyawat, M. A., A simple and rapid solvent extr action method for determining total lipids in fish tissue. J. AOAC Int. 1996, 79, 487-492. (141) Shantha, N. C.; Decker, E. A., Rapid, sensitive, iron-based. spectrophotometric methods for determination of peroxide values of food. lipids. J. AOAC Int. 1994, 77, (2), 421-424. (142) Buege, J. A.; Aust, S. D., Microsomal lip id peroxidation. In Methods Enzymol., Fleischer, S.; Packer, L., Eds. Academic Press: New York, 1978; p 302. (143) Petillo, D. Antioxidant loss es in the light ad dark muscle of Atlantic mackerel (Scomber scrombrus). Massachusetts, Amherst, 1996. (144) Hamilton, R. J., The chemistry of the Rancidity of foods. In Rancidity in foods, Allen, J. C.; Hamilton, R. J., Eds. Chapman and Hall: New York, 1994. (145) Monahan, F. J., Chapter 1. Oxidation of Lipids in Muscle Foods: Fundamental and Applied Concerns. In Antioxidants in Muscle Foods: Nutritional Strategies to Improve Quality, Decker, E. A.; Faustman, C.; Lopez-Bote, C. J., Eds. John Wiley and Songs, Inc.: New York, 2000; p 499. (146) Takama, K., Changes in flesh lipids of fish during frozen storage. Bull. Fac. Fish. Hokaido University. 1974, V, (25), 256-263. (147) Sato, K.; Hegarty, G. R., Warmed-over flavor in cooked meats. J Food Sci. 1971, 36, 1098-1102. (148) Decker, E. A.; Livisay, S. A.; Zhou, S ., Mechanisms of Endogenous Skeletal Muscle Antioxidants: Chemical and Physical Aspects. In Antioxidants in Muscle Foods, Decker, E. A.; Faustman, C.; Lopez-Bote, C. J., Ed s. John Wiley and Sons, Inc.: New York, 2000. (149) Uchida, K.; Stadtman, E. R ., Modifications of histidine resi dues in proteins by reaction with 4hydroxynonenal. Biochemistry 1992, 89, 4544-4548. (150) Eisenbrand, G.; Schuhmacher, J.; Golzer, P., The influence of glutat hione and detoxifying enzymes on DNA damage induced by 2-alkena l in primary rat hempatocytes and hum lymphhoblastoid cells. Chem. Res Toxicol. 1995, 8, (1), 40-46. (151) Alsavar, C.; Quantick, P. c.; Grigor, J. M ., Aroma compounds of fresh and stored Mackerel (Scomber scombrus) In Chapter 5 in Flavor a nd lipid chemistry of Seafoods., Shahidi, F.; Cadwallader, K. R., Eds. American Ch emical Society: Washington, 1997; pp 39-54. (152) Kristinsson, H. G. Conformation and f unctional changes of hemoglobin and myosin induced by pH: functional role in fish quali ty. University of Massachusetts, Amherst, 2002.

PAGE 156

156 (153) Richards, M. P.; Kelleher, S. D.; Hultin, H. O., Effect of washing with or without antioxidants on quality retention of mackerel f illets during refrigerated and frozen storage. J Food Sci. 1998, 46, (4363-4371). (154) Richards, M. P.; Hultin, H. O., Contributions of blood and blood components to lipid oxidation in fish muscle. J Agri Food Chem 2002, 50, 555-564. (155) Undeland, I. A.; Hultin, H. O.; Richards, M. P., Aqueous Extracts from Some Muscles Inhibit Hemoglobin-Mediated Oxidation of C od Muscle Membrane Lipids. J. Agric Food Chem 2003, 51, (3111-3119). (156) Riggs, A., Properties of Hemoglobin. A cademic Press: New York, 1970; Vol. IV, p 209-252. (157) Richards, M. P.; Hultin, H. O., Hemolysa tes from mackerel, herring and trout promote lipid oxidation at different rates. Fish Sci. 2003, 69, 1298-1300. (158) Hargrove, M. S.; Krzywda, S.; Wilkinson, A. J.; Dou, Y.; Ikedasaito, M.; Olson, J. S., Stability of myoglobin A model for the fold ing of heme-proteins. Biochemistry 1994, 33, (39), 11767-11775. (159) Kristinsson, H. G., Changes in trout he moglobin conformations and solubility after exposure to acid and alkali pH. J Agri Food Chem 2004, 52, 3633-3643. (160) Alayash, A. I., Hemoglobin-based blood subs titutes: oxygen carriers, pressor agents, or oxidants? Nat. Biotchnol. 1999, 17, 545-549. (161) Pazos, M.; Medina, I.; Hultin, H. O., Effect of pH on Hemoglobin-Catalyzed lipid Oxidation in Cod Muscle Membranes in V itro and in Situ. J Agri Food Chem 2005, 53, 3605-3612. (162) Richards, M. P.; Modra, A. M.; Li., R., Role of deoxyhemoglobin in lipid oxidation of washed cod muscle mediated by trout, poultr y and beef hemoglobins. Meat Sci. 2002, 62, 157-163. (163) Lee, S.; Joo, S.; Phillips, A. L.; Faustm an, C. In Oxymyoglobin and lipid oxidation in yellowfin tune (Thunnus albaca res). Annual Meeting and Food Expo., Anaheim, A, 2002; Anaheim, A, 2002. (164) Goto, Y.; Fink, A. L., Acid-induced fold ing of heme proteins. Methods of Enzym. 1994, 232, 3-15. (165) Cilliard, J.; Cilliard, P.; Cormier, M., Effect of experime ntal factors on the prooxidant behaviour of alpha-tocopherol. J. A. Oil Chem. Soc. 1980, 57, 255-261. (166) Benedict, R. C.; Strange, E. D.; Swift, C. E. Effect of lipid antioxidants on the stability of meat during storage. J. Agric Food Chem 1975, 23, 167-173. (167) WATTS, B. M., Oxidative rancidity and disc oloration in meat. J. Advances Food Research 1954, 5, (1). (168) Halliwell, B.; Gutteridge, J. M. C., Oxygen free radicals and iron in relation to biology and medicine: Some problems and concep ts. Arch. Biochem. Biophys. 1986, 246, 501-514. (169) Kanner, J., Mechanism of nonenzymatic li pid peroxidation in muscle foods. In Lipid Oxidation in Foods, St. Angelo, A. J., Ed. American Chemical Society Books: Washington, D.C., 1992; Vol. 500, pp 55-73. (170) Richards, M. P.; Li., R., Effects of releas ed iron, lipid peroxides, and ascorbate in trout hemoglobin-mediated lipid oxidation of wash ed cod muscle. J. Agric. Food Chem. 2004, 52, 4323-4329. (171) Buettner, G., In the absen ce of catalytic metals, ascorbat e does not autoxidize at pH 7: Ascorbates as a test of catalytic meta ls. J. Biochem. Biophys. Methods 1988, 16, 27-40.

PAGE 157

157 (172) Harris, R. S., General Discussion on the stab ility of nutrients. In Nu tritional evaluation of Food Processing, Harris, R. S.; von Loesecke, H., Eds. AVI: Westport, CT., 1971; pp 1-4. (173) Gregory III, J. F., Vitamins. In Food Chem ., 3rd ed.; Fenemma, O. R., Ed. Marcel Dekker, Inc.: New York, 1996; pp 532-609. (174) Lee, S. H.; Oe, T.; Blair, I. A., Vitamin C induced-induced decomposition of lipid hydroperoxides to endogenous genotoxins. Science 2001, 292, 2083-2086. (175) Buettner, G., The pecking order of free ra dicals and antioxidants: lipid peroxidation, alphatocopherol, and ascorbate. Arch. Biochem. Biophys. 1993, 300, 535-543. (176) Traber, M. G., Determinants in plasma vitamin E concentrations. Free Radical Biol. Med. 1994, 16, (229-239). (177) Petillo, D.; HO, H.; J, K. ; WR, A., Kinetics of antioxidant loss in mackerel light and dark muscle J Agri Food Chem 1998, 46 (10), 4128-4137 (178) Jia, T.; Kelleher, S.; Hultin, H.; Petill o, D.; Maney, R.; Krzynowek, J., Comparison of quality loss and changes in the glutathione an tioxidant system in stored mackerel and bluefish muscle J. Agric. Food Chem. 1996, 44, (5), 1195-1201. (179) Pazos, M.; Gonzalez, M., Preservation of the endogenous antioxidant system of fish muscle by grape polyphenols during frozen storage. European Food Researh and Technology 2005, 220, 514-519. (180) Erickson, M. C., Compositional Parameters a nd their relationship to oxidative Stability of Channel Catfish J Agri Food Chem 41, 1213-1218. (181) Decker, E. A.; Livisay, S. A.; Zhou, S ., Mechanisms of Endogenous Skeletal Muscle Antioxidants: Chemical and Physical Aspects. In Antioxidants in Muscle Foods, Decker, E. A.; Faustman, C.; Lopez-Bote, C. J ., Eds. John Wiley and Sons: New York, 2000. (182) McCay, P. B., Vitamin E: Interactions wi th free radicals and ascorbate. Annu. Rev. Nutr. 1985, 5, 323-340. (183) Kroger-Ohlsen, M.; Andersen, L., Acidcatalyzed autoreduction of ferrymyoglobin in aqueous solution studied by freeze quenching and ESR spectroscopy. Free Radic. Res. 1999, 30, (4). (184) Pokorny, J.; Davidek, J.; Chocholata, V.; Pa nek, J.; Bulantova, H.; Janitz, W.; Valentova, H.; Vierecklova, M., Interactions of oxidized lipids with protein. XVI: Interactions of oxidized ethyl linoleate with collagen. Die Nahrung 1990, 34, 159-169. (185) Kanner, J.; Karel, M., Changes in lysozyme due to reactions with the peroxidizing methyl linoleate in a dehydrated model system J. Agric. Food Chem. 1976, 24, 468-472. (186) Funns, J. A.; Weiss, U.; Karel, M., Effects of reaction c onditions and reactant concentrations on polymerization of lysozyme reacted with peroxidizing lipids. J Agri Food Chem 1982, 30, 1204-1208. (187) Kelleher, S. D.; Hultin, H. O.; Wilhelm, K. A., Stability of mackerel surimi prepared under lipid-stabilizing processing conditi ons. J. Food Sci 1994, 59, (269-271). (188) Smith, D. M., Functional and Biochemical ch anges is deboned turkey due to frozen storage and lipid oxidation. J. Food Sci 1987, 52, (22-27). (189) Lui, G.; Xiong, Y. L., Storage stability of antioxidant-washed myofibrils from chicken white and red muscle. J. Food Sci 1996, 61, (890-894). (190) Srinivasan, S.; Xiong, Y. L., Gelation of beef heart surimi as affected by antioxidants. J. Food Sci 1996, 61, 707-711.

PAGE 158

158 (191) Parkington, J. K.; Xiong, Y. L.; Xiong, S.; Blanchard, S. P. ; Srinivasan, S.; Froning, G. W., Chemical and functional properties of oxidati vely modified beef heart surimi stored at 2 degrees C. J Food Sci. 2000, 65, (3), 428-433. (192) Kisia, S. M., Structure of fish locomoto ray muscle. In Fish Morphology, Datta-Munshi, J. S.; Gutta, H. M., Eds. AA Balkema: Rotterdam, 1996; pp 167-178. (193) Lemon, D. W. An improved TBA test for Rancidity; 51 Halifax, NS, Canada, 1975. (194) Torten, J.; Whitaker, J. R., Evaluation of Biuret and Dye-bindi ng Methods for Protein Determination in Meats. J. Food Sci 1964, 29, (2), 168-174. (195) Bradford, M. M., A rapid and sensitiv e method for the quantif ication of microgram quantities of protein utilizing the principle of protein-dye bindi ng. Anal. Biochem 1976, 72, 248-254 (196) Laemmli, U. K., Cleavage of structural proteins during the assembly of the head of bacteriophase T4. Nature 1970, 227, 680-685. (197) Liang, Y.; Hultin, H., Functional protein isolates from mechanically deboned turkey by alkaline solubilization with isoelectric precipitation. J. Muscle foods 2003, 14, 195-206. (198) Sherman, P., Industrial Rheology: with part icular reference to foods, pharmaceuticals, and cosmetics. Academic Press: New York, 1970. (199) Kristinsson, H. G.; Hultin, H. O., Change s in Conformation and Subunit Assembly of cod Myosin at Low and High pH and after Subse quent Refolding. J. Agric Food Chem 2003, 51, 7187-7196. (200) Yongsawatdigul, J.; Park, J., Effects of alkali and acid solubilation on gelation characteristics of rockfish muscle proteins. j Food Sci. 2004, 69, C499-505. (201) Choi, Y. J.; Park, J. W., Acid-aided prot ein recovery from enzyme-rich Pacific whiting. J Food Sci. 2002, 67, 2962-2967. (202) Yongsawatdigul, J.; Park, J. W.; Esturk, O ., Gelation characteristics of alkaline and acid solubilization of fish muscle proteins. In Institute of Food Technologist (IFT), New Orleans, 2001. (203) 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 J. Agric Food Chem 2003, 51, 3992-3998. (204) Kristinsson, H. G.; Theodore, A. E.; Necla, D.; Ingadottir, B., A Comparative Study between Acid and Alkali-aided Processing a nd Surimi Processing for the Recovery of Proteins from Channel Catfish Muscle. Food Chemistry and Toxicology 2005, 70, (4), C298-C306. (205) Kristinsson, H. G.; Liang, Y., Effect of pH-shift proces sing and surimi processing on Atlantic Croaker (Micropogoni as undulates) Muscle Proteins. Food Chemistry and Toxicology 2006, 71, (5), C304-C312. (206) Undeland, I. A.; Kristinsson, H. G.; Hu ltin, H. O., Hemoglobin catalyzed oxidation of washed cod muscle phospholipids-effect from pH and hemoglobin source. J Agri Food Chem 2004, 52, 4444-4451. (207) Kenney, P. B.; Kastner, C. L.; Kropf, D. H., Muscle washing a nd raw material source affect quality and physiochemical properties of low-salt, low-fat, restructured beef. J. Food Sci 1992 57, 545-550. (208) USDA, Beef products: Raw, Processed and Prepared. In U. S. Department of Agriculture Handbook 8-13, U.S. Government Printi ng Office: Washington, D.C., 1990.

PAGE 159

159 (209) Luo, S.; Hultin, H. O., In vitro lipid oxida tion modifies proteins and functional properties of sarcoplasmic reticulum. Bioenergetics and biomembranes 1986, 18, (4), 315-323. (210) Srinivasan, S.; Hultin, H. O., Hydroxyl ra dical modification of fish muscle proteins. Journal of Food Biochemistry 1995, 18, 402-425. (211) Ostdal, H.; Davies, M. J.; Andersen, H. J., Reaction between protein radicals and other biomolecules. Free Rad. Bio. Med 2002, 33, (2), 201-209. (212) Olinescu, R. M.; Kummerow, F. A., Fibrin ogen is an efficient antioxidant. J. Nutr. Biochem. 2001, 12, 162-169. (213) Luxford, C.; Morin, B.; Dean, R. T.; Davi es, M. J., Histone H1and other protein-and amino acid hydroperoxides can give rise to fr ee radicals which oxidize DNA. Biochem. J. 1999, 344, 125-134. (214) Stadtman, E. R.; Berlett, B. S., Reactiv e oxygen-mediate protein oxidation in aging and disease. Chem. Res Toxicol. 1997, 10, 484-494. (215) Kjaersgard, I. V. H.; Norrelykke, M. R.; Baron, C. P.; Jessen, F., Identification of carbonylatd protein in frozen rainbow tr out (Oncorhynchus mykiss) fillets and development of Protein Oxidation during fr ozen storage. J Agri Food Chem 2006, 54, 9437-9446. (216) Ingadottir, B. The use of acid and alka li-aided protein solubila tion the precipidation methods to produce functional protein ingredie nts from tilapia. University of Florida, Gainesville, 2005. (217) Theodore, A. E. Bioactive and functional properties of catfish pr otein hydrolysates and catfish protein isolates. University of Florida, Gainesville, 2005. (218) Davenport, M.; Kristinsson, H. G., Effect of different acid and alkali-treatments on the molecular and functional propert ies of catfish muscle protei ns. Abstract: 49G-14. In IFT Annual Meeting, Las Vegas, Nevada,2004. (219) Samejima, K.; Takahashi, K.; Yasui, T., Heat induced denaturatio n of myosin total rod. Agric. Biol. Chem. 1976, 40, (12), 2455. (220) Samejima, K.; Ishioroshi, M.; Yasui, T., Rela tive roles of the head and tail portions of the molecule in the heat-induced gela tion of myosin. J. Food Sci. 1981, 46, 1412. (221) Shikha, F. H.; Hossain, M. I.; Morioka, K.; Kubota, S.; Yoshiaki, I., Effect of pH-shifting on the gel forming characteristi cs of salt-ground meat from walleye pollock. Fisheries Science 2006, 72, 870-876. (222) Schaich, K. Alterations in protein functionality due to oxid ative reactions. University of Kentucky, Lexington, 1993. (223) Strehler, E. E.; Strehler-Page, M. A.; Pe rriad, J. C.; M., P.; Nadal-Ginard, B., Complete nucleotide and encodes amino acid sequences of a malian myosin heavy chain gene: Evidence again intron-dependent evoluti on of the rod. J. Mol. Biol 1986, 190, 291-317. (224) Belitz, H. D.; Grosch, W.; Schieberle, P., Meat. 3rd ed.; Sp ringer-Verlag: Berlin, Germany, 2004. (225) Nishimura, K.; Ohishi, N.; Tanaka, Y.; Sa skura, C., Participation of radicals in polymerization by ascorbic acid of crude actomyos in from frozen surimi of Alaska pollock during a 40C incubation. Biosci Biotech Biochem 1992, 56, 24-28. (226) Lanier, T. C.; Lin, T. S.; Hamann, D. D.; Th omas, F. B., Effects of alkaline protease in minced fish on texture of heat-p rocessed gels. J. Food Sci. 1981, 46, 1643. (227) Lee, C. M., Surimi Process Technology. Food Tech. 1984, 38, 69.

PAGE 160

160 (228) An, H.; Weerasinghe, V.; Seymore, T. A. ; Morrissey, M. T., Cath epsin degradation of Pacific whiting surimi proteins J Food Sci. 1994, 59, (1013-1017, 1033). (229) Kristinsson, H. G.; Rasco, B. A., Hydrolys is of salmon muscle proteins by an enzyme mixture extracted from Atlant ic salmon (Salmo salar) pyloric caeca. J. Food Biochem. 2000, 24, 177-187. (230) Kristinsson, H. G., Evaluati on of different methods to isolate cod (Gadus morhua) muscle myosin. J. Food Biochem. 2001, 48, 3141-3147. (231) Takeda, H.; Seki, N., Enzy me catalyzed cross-linking and degradation of myosin heavy chain in Walleyed pollock surimi paste during setting. Fisheries Sci. 1996, 62, 462-7. (232) Kelleher, S. D.; Hultin, H. O., Surimi pro cessing from dark muscled fish. In Surimi and Surimi Seafood, Park, J. W., Ed. Marcel Dekker, Inc.: New York, 2000; pp 59-77. (233) Gray, J. I.; Pearson, A. M., Rancidity and warmed over flavour. Adv. Meat Res. 1987, 3, 221. (234) Erickson, M. C.; Hultin, H. O., Nonenzyma tic catalysis of lipid oxidation by the soluble fractions of mackerel muscle. J. Food Sci 1990, 55, 947-950, 953. (235) Richards, M. P.; Hultin, H. O., Hemolysate from mackerel herring and trout promote lipid oxidation at different rates. Fish Sci. (236) Slabyj, B. M.; Hultin, H. O., Lipid-pero xidation by microsomal fractions-isolated from light and dark muscles of herring (Cl upea-harengus). J. Food Sci 1982, 47, 1395-1398. (237) Han, H. T.; Liston, J., Li pid-peroxidation protection fact ors in rainbow-trout (SalmoGairdnerii) muscle cytosol. J. Food Sci 1989, 54, (4), 809-813. (238) Gunnarsson, G.; Undeland, I. A.; Thippe swamy, S.; Sanberg, A.; Lingard, A.; MattssonHulten, L.; Soussi, B., Inhibito ry Effect of Known Antioxidants and of Press Juice from Herring (Clupea harengus) Light Muscle on the Generation of Free Radicals in Human Monocytes. J Agri Food Chem 2006, 54, 8212-8221. (239) Rong, L., Characterization of aqueous compone nts in chicken breast mu scle as inhibitors of hemoglobin-mediated lipid oxid ation. J. Agric Food Chem 2005, 53, 767-775. (240) Rowley, A. F., Collection, separation and id entification of fish leukocytes. In Techniques in Fish Immunology, Stolen, J. S.; Fletcher, T. C.; Andersen, D. P.; Roberson, B. S.; Van Muuiswinkel, W. B., Eds. SOS Pub lications: New Jersey, 1990; pp 113-135. (241) Fyhn, U. E.; Fyhnm, H. J.; Davis, B. J. ; Powers, D. A.; Fink, W. L.; Garlick, R. L., Hemoglobin heterogenity in Amazonian fi shes. Comp. Biochem. Physiol. 1979, 62A, 3966. (242) Prior, R. L.; Cao, G., In vivo total antioxi dant capacity: comparison of different analytical methods. Free Radical Biol. Med. 1999, 27, (1173-1181). (243) Lawrie, R. A., Meat Science. 4th ed.; Pergamon Pr ess: New York, 1985. (244) Grunwald, E., Mechanisms of Heme Prot ein-Mediated lipid oxidation using hemoglobin and myoglobin variants in raw and heat ed washed muscle. Agriculture and Food Chemistry 2006, 54, 8271-8280. (245) Decker, E. A.; Hultin, H. O., Factors influe ncing catalysis of lipid oxidation by the soluble faction of mackerel muscle. J. Food Sci 1990, 55, 947-950,953. (246) Rong, L.; Richards, M. P.; Undeland, I. A ., Characterization of aqueous components in chicken breast muscle as inhibitors of he moglobin-mediated lipid oxidation. Agriculture and Food Chemistry 2005, 53, 767-775.

PAGE 161

161 (247) Chiu, D. T.; Van den Berg, J.; Kuypers, F. A.; Hung, I. J.; Wei, J. S.; Lui, T. Z., Correlation of membrane lipid peroxidati on with oxidation of hemoglobin variants: possibly related to hemin releas e. Free Rad. Bio. Med 1996, 21, 89-95. (248) Bamm, V. V.; Tsemakhovich, V. A.; Shak lai, M.; Shaklai, N., Haptoglobin phenotypes differ in their ability to inhi bit heme transfer from hemoglobin to LDL. Biochem. 2004, 43, 3899-3906. (249) Balla, G.; Jacob, H. S.; Eaton, J. W.; Belchler, J. D.; G.M., V., Hemin: a possible physiological mediator of low density li poprotein oxidation in endothelial injury. Arterioscler. Throm. 1991, 11, 1700-1711. (250) Grinshtein, N.; Bamm, B. B.; Tsemakhovich, V. A.; Shaklai, N., Mechanism of low density lipoprotein oxidation by haemogl obin derived iron. Biochemistry 2003, 42, 6977-6985. (251) Jorolim, P.; Lahav, M.; Liu, S. C.; Palek, J., Effect of hemoglobin oxidation products on the stability of red cell membrane skeletons and the association of skeletal proteins: correlations with releas e of heme. Blood 1990, 76, 2125-2131. (252) Richards, M. P.; Detmann, M. A.; Grunwal d, E., Pro-oxidative characteristics of trout hemoglobin and myoglobin: A role for released heme in oxidation of lipids. J Agri Food Chem 2005, 53, 10231-10238. (253) Vincent, S. H.; Grady, R. W.; Shaklai, N.; Snider, J. M.; Muller-Eberhard, U., The influence of heme-binding proteins in he me catalyzed oxidations. Arch. Biochem. Biophys. 1988, 265, 539-550. (254) Griffonm, N.; Baudin, V.; Dieryck, W.; Dum oulin, A.; Pagnier, J.; Poyart, C.; Marden, M. C., Tetramer-dimer equilibrium of oxyhemogl obin mutant determined from auto-oxidation rates. Protein Sci. 1988, 7, 673-680. (255) Baron, C. P.; Skibsted, L. H.; Andersen, H. J., Prooxidative activ ity of myoglobin species in linoleic acid emulsions. J. Agric. Food Chem. 1997, 45, (5), 1704-1710. (256) Storage, M.; Matsuoka, A.; Highmore, A.; Seaward, Y.; Shikama, K., The molecular mechanism of autoauxidation for huma n oxyhemoglobin. J. Biol. Chem. 1998, 273, 86078615. (257) Misra, H. P.; Fridovich, I ., Generation of superoxide duri ng autoxidation of hemoglobin. J. Biol. Chem. 1972, 247, (6960-6962). (258) Yoshida, Y.; Kashiba, K.; Niki, E., Freeradical mediated oxidation of lipid induced by Hemoglobin in aqueous dispersions Biochem. Biophs. Act 1994, 1201, 165-172. (259) Harel, S.; Kanner, J., Hydrogen per oxide-activated methae-myoglobin and other haemoproteins as initiators of membrana l lipid peroxidation. In Superoxide and Superoxide Dismutase in Chemistry, Rotilo, G., Ed. Elsevier Science B.V. (Biomedical Division): New York, 1986; pp 25-28. (260) Shen, L.; Hermans, J., Kinetics of conformation change of sperm-whale myoglobin. Biochem. 1972, III. Folding and unfolding of apomyoglobins and the suggested overall mechanism, 1845-1849. (261) Kibald, M.; Luck, T.; Holley, W.; Kloste rmeyer, H., Decolorization of slaughterhouse thick blood for the recovery of a colourless protein preparation and an iron-containing heme fraction. In Food proteins: structure and functionality, Schwenke, K. D.; Mothes, R., Eds. VCH: Weinheim, Germany, 1991; pp 103-105. (262) Ericksson, C. E.; Olsson, P. A.; S.G., S., Denatured hemoproteins as catalysts in lipid oxidation. JAOCS 1971, 48, 442-447.

PAGE 162

162 (263) Cannon, J. B.; Kuo, F. S.; Paternack, R. F. ; Wong, N. M.; Muller-Eer hard, U., Kinetics of the interaction of hemin liposomes with heme binding proteins. Biochem. 1984, 23, 37153721. (264) Hargrove, M. S.; Whitaker, T.; Olson, J. S.; Vali, R. J.; Mathews, A. J., Quaternary structure regulates hemin dissociation fr om human hemoglobin. J. Biol. Chem. 1997, 272, 17385-17389. (265) Miyawawa, T.; Suzuki, T.; Fujimoto, K. ; Kinoshita, M., Age related change of phosphatiylchonline hydroperoxide and phosphotidylethanolamine hydroperoxide levels in normal human red blood cells. Mech. Ageing Dev. 1996, 86, 145-150. (266) Wright, J. S.; Johnson, E. R.; Dilabio, G. A., Predicting the activity of phenolic antioxidants: theoretical method, analysis of substituent effects, and application to major families of antioxidants. J. Am. Chem. Soc. 2001, 123, 1173-1183. (267) Awiki, J. M.; Rooney, L. W.; Wu, X.; Prio r, R. L.; Cisneros-Zevallos, L., Screening methods to measure antioxidant activity of sorghum (Sorghum bicolor) and sorghum products. J. Agric. Food Chem. 2003, 51, 6657-6662. (268) Prior, R. L.; Wu, X.; Sc haich, K., Standardized Methods for the Determination of Antioxidant Capacity and Phenolics in Foods and Dietary Supplements. J. Agric. Food Chem. 2005, 53, 4290-4302. (269) Van den Berg, R.; Haenen, G. R. M. M.; Va n den Berg, H.; Bast, A., Applicability of an improved Trolox equivalent antioxidant cap acity (TEAC) assay for evaluation of antioxidant capacity measurements of mixtures. Food Chemistry 1999, 66, 511-517. (270) Antonenkov, V. D.; Sies, H., Some basic pr operties of ascorbate-dependent antioxidativedefense factors from rat liver cytoso l. Chem.-Biol. Interact. 1995, 94, 169-181.

PAGE 163

163 BIOGRAPHICAL SKETCH Ms. Holly Tasha Petty was born in Anchorage Al aska, and was raised in Nikiski, Alaska for the first 15 years of her life. She then live d in Oviedo, Florida, and received her high school diploma from Oviedo High school. She received he r Bachelor of Science degree in food science and human nutrition, with honors. In Fall 2002, Holly began her doctoral graduate studies in food science and human nutrition, and minor in applied physiology and kinesiology. As a graduate student, she served as the Florida Section Institute of F ood Technologists student representative from, 2005-2006; an d she served as the Food Scie nce Graduate Representative, 2005-2006. During her graduate schooling, Holly was awarded the Agricultural Womans Scholarship, 2006; Florida Section Institute of Food Technologists Student Fellowship, 2006; Aylesworth Foundation for the Advancement of Marine Sciences fellowship, 2005-2007; and the Institute of Food Technologists National grad uate scholarship, 2004-2005. She is the main inventor of the, Novel Low-Allergenic Food Bar, U.S. Provisional Patent No. 60/844,431, as well and co-inventor of Liquid Nutrient Composition for Improving Performance, UF. Disclosure No. 12237. Holly has contributed to va rious metabolic and hea lth-related studies as a participant and trainer through the Applie d Physiology and Kinesiology Department, University of Florida. Holly has been a me mber of the National Strength and Conditioning Association, Internat ional Association for Food Protection, American Oil Chemist Society, Institute of Food Technologists, Phi Tau Sigma Honor Society, and the University of Florida Womens Cycling Team. Upon graduation, Holly hope s to contribute to industry and academia, promoting health initiatives through her knowledg e of food science, nu trition, and exercise physiology; and her passion for the vita lity of people and the environment.