PLACE II RETURN BOX to roman this chockout mm your mood.
To AVOID FINES mum on or baton» dd. duo.

DATE DUE [{DATE DUE DATE DUE

L__[__-
L__JL___JL__|
——T:::J

MSU Is An Nﬁm‘aﬂvo Mon/Ema! Opportunity Institulon

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m1

 

THE EFFECT OF SALT ON NON-HEME IRON RELEASE AND LIPID
OXIDATION IN PORK

by

HSING-FENG LIU

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

DEPARTMENT OF FOOD SCIENCE AND HUMAN NUTRITION

1997

ABSTRACT

THE EFFECT OF SALT ON THE NON-HEME IRON RELEASE AND LIPID
OXIDATION IN PORK

By

HSING-FENG LIU

Sodium chloride is a well known prooxidant in the meat systems. However, the
prooxidant mechanisms of NaCl are unclear. This study speciﬁcally focused on the effect
of NaCl on the release of non-heme iron in ground pork and water-washed muscle ﬁbers
and its relationship to lipid oxidation.

Three non-heme iron quantitation methods were evaluated to better understand
why different methods realize different non-heme iron values when quantitating the same
sample. Ferrous chloride (0, 3, 6 and 9 ug Fe/g meat) was added to ground pork. The
pork was formed into patties. The modiﬁed Schricker, Igene, and ferrozine procedures
were used to quantitate the non-heme, heme and total iron concentrations. The modiﬁed
Schricker method gave greater overall recoveries of added iron than the ferrozine and
Igene methods. The non-heme and heme iron distributions for each pork sample was
different when evaluated using the different analytical methods. Greater non-heme iron
and smaller heme iron concentrations were found when using the modiﬁed Schricker
method.

The effect of NaCl concentration (0, 0.15, 0.30 and 0.45 M) and different salts
(NaCl, KC], NaBr and KBr) on non-heme iron concentrations and lipid oxidation was

studied in ground pork. As NaCl concentration increased from O to 0.45 M,

Thiobarbituric acid-reactive substances (TBARS) and peroxide values increased (p<0.05)
in both raw and cooked samples. Non-heme iron concentrations increased in raw samples
aﬁer 6 days at 4 °C and for cooked samples immediately after cooking (day 0).
Treatments containing different salts had higher (p<0.05) TBARS and peroxide values
than the control, and signiﬁcantly (p<0.05) increased non-heme iron concentration in raw
samples after 6 days. At the same molarity, lipid oxidation and non-heme iron
concentrations produced by different salts were not signiﬁcantly different.

The effects of NaCl, KCl, NaBr and KBr on non-heme iron release from
hemoglobin, myoglobin and recovery of added iron were studied in raw and cooked
water-washed muscle ﬁbers. The addition of various salts to the model system increased
TBARS values. However, it had no effect on the release of non-heme iron from
hemoglobin and myoglobin. The addition of various salts to cooked washed muscle ﬁbers
containing 3 ug Fe/g meat added iron produced greater TBARS values compared to the
model system without salt. Salt addition did not signiﬁcantly increase the non-heme iron
concentration in the model system. Sodium chloride promoted lipid oxidation without
increasing the non-heme iron fraction in this model system. It is proposed that a major
ﬁmction of NaCl in lipid oxidation is to reduce the integrity of the muscle structure.
Therefore, lipids are more accessible for the attack of prooxidants such as non-heme iron.

KEY WORDS: SALT, LIPID OXIDATION, NON-HEME IRON.

iii

DEDICATION

to my parents who always support me

to my young brother who takes the responsibility to take care of
my parents so I can pursue what I want

to my wife who always with me

to my little boy and girl

ACKNOWLEDGNIENTS

I would like thank my major professor Dr. AM. Booren for his guidance and
support during my education at MSU. I am also grateﬁrl to my committee members, Dr.
J. 1. Gray, Dr. M.J. Zabik, Dr. D.M. Smith and Dr. GM. Strasburg for serving on my
guidance committee and their critical review of this manuscript. Special thanks go to Dr.
J. 1. Gray for his understanding and support. All of the work was conducted in his
laboratory. Finally, I would like to thank all of my ﬁiends for their support during my stay

at MSU.

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

LIST OF TABLES -- -- ........... .... --..------.-VIII
LIST OF FIGURES - ..................... - -- - XI
INTRODUCTION ...................................................................................................... 1
REVIEW OF LITERATURE ...... - - ..... - - -5
Introduction ........ ......... - -- - -- -- - -- - -------- 5
Lipid Oxidation-- - -- - - - - ------- 6
l. Mechanism of lipid oxidation .... - ----- - -6
2. Role of Iron in lipid oxidation - - - -7
The Prooxidant Effect of Salt on Lipid Oxidation In Meat ------ 12
Possible Mechanisms for the Prooxidant Effect of NaCl ..... -- - - -- 14
1. Sodium chloride and free iron ion ................................................................. 15
2. Anion-promoted autooxidation ..................................................................... l7
3. Water activity-- ---------- -- - -- - - .............. 18
4. Physical state of meat ....... - - - -- ----------- l9
5. Enzyme activity .............................................................................................. 20
Non-heme Iron Measurement ............................................................................... 21
1. Consensus in terminology .............................................................................. 22

2. Differences in iron values from the same sample between non-heme iron
measurement methods .................................................................................... 23
References .............................................................................................................. 26

CHAPTER I

A COMPARISON OF THREE ANALYTICAL PROCEDURES FOR
DETERMINING NON-HEME IRON CONCENTRATION IN GROUND PORK33

 

 

 

 

Abstract-- -- - -- - - ---------- ----- ..... - -33
Introduction .......................................................................................................... 34
Materials and Methods - ----- - -- -- -- - - 37
Materials - - - -- -- - - - - - -...--.37
Sample preparation -_ ..... - - - 37
Data analysis ...................................................................................................... 41

vi

Results and Discussion -- - - - - 43

 

 

 

 

 

 

 

 

 

 

 

 

 

Recovery of added iron using three methods -- - - - -- 43
Evaluation of three sample quantitation procedures -------- -- - - 44
Evaluation of three sample preparation procedures -- ------- - - - - - 49
Summary and Conclusions - -- -- - - - -- -- -- - -- --59
References .............................................................................................................. 60
CHAPTER II
THE EFFECT OF VARIOUS SALTS ON NON-HEME IRON
CONCENTRATION AND LIPID OXIDATION IN GROUND PORK ................ 61
Abstract - - - -- ...... - 61
Introduction ................................... - - - ...... - - --62
Materials and Methods - - - -------- - - - - --65
Sample preparation - - -- -- - ----- - - ..... .65
Data analysis.-. -- - - - - - ..... - - - -.67
Results and Discussion ------------------------ -- - 68
Effect of NaCl concentrations on TBARS and Peroxide values ....................... 68
Effect of NaCl on non-heme Iron concentration ............................................... 72
The relationship between lipid oxidation, non-heme iron concentration and
NaCl-- -- .- - ..... -- 76
The effect of various salts on lipid oxidation and non-heme Iron concentration
in pork---. ---------- - -- -- ----- - ------- 77
Summary and Conclusions ................................................................................... 84
References .............................................................................................................. 85
CHAPTER [D

THE EFFECTS OF CHLORIDE AND BROMIDE SALTS ON NON-HEME
IRON CONCENTRATION AND LIPID OXIDATION IN A MODEL SYSTEM

 

 

CONTAINING VARIOUS FORMS OF IRON ...................................................... 88
Abstract-- - - -- - --------------------- - - --88
Introduction -- ------------------- --------- - --...-.89
Material and Methods .......................................................................................... 92

Sample preparation ........................................................................................... 92
Data analysis ...................................................................................................... 94
Results and Discussion .......................................................................................... 96
The effect of salt on non-heme iron concentrations in the WP system
containing heme proteins ................................................................................... 96
The effect of salt on lipid oxidation in the presence of heme iron .................. 104
Effect of salt on non-heme iron recovery and lipid oxidation in a model system105
Summary and Conclusions ................................................................................. 115
References ............................................................................................................ 116
SUMMARY AND CONCLUSIONS ..................................................................... 119
FUTURE RESEARCH .......................................................................................... 123

vii

LIST OF TABLES

CHAPTERI

Table 1-1 Comparison of iron quantitation procedures of raw pork extracted using
the Igene method - _ ......................... _- -- - - .......... 47

 

Table 1-2 Comparison of iron quantitation procedures of cooked pork extracted
using the Igene method ............................................................................... 48

Table 1-3 Comparison of iron quantitation procedures of raw pork extracted using
the ferrozine method - _ _- - _- - -- - _ - - - - ________ - __ 50

 

Table 1-4 Comparison of iron quantitation procedures of cooked pork extracted
using the ferrozine method - - ......... -_ - 51

 

Table 1-5 Comparison of iron quantitation procedures of raw pork extracted using
the modiﬁed Schricker method . ................................................................. 52

Table 1-6 Comparison of iron quantitation procedures of cooked pork extracted
using the modiﬁed Schricker method ......................................................... 53

Table 1-7 Heme, non-heme and total iron in the raw samples measured by Igene,
ferrozine and modified Schricker methods. ................................................ 54

Table 1-8 Heme, non-heme and total iron in the cooked samples measured by
Igene, ferrozine and modified Schricker methods ...................................... 56

viii

CHAPTER II

Table 2-1 The effect of NaCl concentrations on TBARS and peroxide values in raw

 

 

 

 

 

 

ground pork during refrigerated storage (4 0C) - -_ - -- _ - 2--“69
Table 2-2 The effect of NaCl concentrations TBARS and peroxide values in cooked
ground pork during refrigerated storage (4 °C) 71
Table 2-3 Total and non-heme iron concentration of raw ground pork during
refrigerated storage (4 °C) - - -- - -- - -- - - 74
Table 2-4 Total iron and non-heme iron concentration of cooked ground pork
during refrigerated storage (4 °C) - - -- - - 75
Table 2-5 The effect of different salts on TBARS and peroxide values in raw
ground pork during refrigerated storage (4 °C) ......................................... 78
Table 2-6 Total and non-heme iron concentration in raw ground pork during
refrigerated storage (4 °C) ........................................................................... 79
Table 2-7 The effect of different salts on TBARS and peroxide values in cooked
ground pork during refrigerated storage (4 °C) - - - - ............ - - 82
Table 2-8 Total iron and non-heme iron concentration in cooked ground pork
during refrigerated storage (4 °C) - - - - - 83
CHAPTER II]

Table 3-1 Total and non-heme iron in a raw water-washed muscle fiber (WF) model
system in the presence of selected heme proteins or salts ........................... 97

Table 3-2 Total and non-heme iron in a cooked water-washed muscle ﬁber (WF)
model system in the presence of selected heme proteins or salts ................ 99

Table 3-3 Effect of cooking and different salts on the iron release from heme
proteins in a model system stored at 4 °C for 2 days ................................ 101

Table 3-4 The effect of heme proteins and different salts on TBARS values in a raw
water-washed muscle fiber (W F) model system ....................................... 106

Table 3-5 The effect of heme proteins and different salts on TBARS values in a
cooked water-washed muscle ﬁber (WF) model system ............................ 107

Table 3-6 The effect of chloride and bromide salts on the recovery of added iron
in a raw water-washed muscle fiber (W F) model system ......................... 108

Table 3-7 The effect of chloride and bromide salts on the recovery of added iron in
a cooked water-washed muscle fiber (WF) model system - - - - 110

 

Table 3-8 The effect of iron and chloride and bromide salts on TBARS values in a
raw water-washed muscle fiber (WF) model system .............. - - 112

 

Table 3-9 The effect of iron and chloride and bromide salts on TBARS values in
cooked water-washed muscle fiber (WF) model system _ - 113

 

LIST OF FIGURES

CHAPTER I

Figure 1-1 Sample preparation scheme for comparing modiﬁed Schricker, ferrozine
and Igene methodology .............................................................. 38

Figure 1-2 Flow diagram for different non-heme iron measurement procedures ..... 39

Figure 1-3 Flow diagram for the evaluation of different iron quantitation
procedure-3m: ........ - -- - - - - - - ....... - - - -- - 40

 

Figure 1-4 The recovery of the added iron from raw ground pork using three
procedures for determining non-heme iron ..................................... 45

Figure 1-5 The recovery of the added iron from cooked ground pork using three
procedures for determining non-heme iron ..................................... 46

CHAPTER 111

Figure 3-1 Water-washed muscle ﬁbers preparation procedure ................................ 95

xi

APPENDIX

A. Modiﬁed Schricker Non-heme Iron Procedure..............................................126
B. Ferrozine Non-heme Iron Procedure...........................................................127
C. Igene Non-heme Iron Procedure128
D. Total iron analysis procedure .............................................................................. 129

E. The recovery of added iron from raw ground pork using three procedures for
determining non-heme iron ..... ........... _ ................. 130

 

F. The recovery of added iron from cooked ground pork using three procedures
for determining non-heme iron - - - -- - - - - - --.-..l31

 

xii

INTRODUCTION

Sodium chloride (N aCl) is an important additive and is widely used by the food
industry. It is added to food for the purpose of ﬂavor, preservation, or to attain proper
functional properties such as texture and water holding capacity. It is also an important
ingredient in surimi-based products such as kamaboko, chikuwa, and satsumage (Japanese
ﬁsh gel products). One of the adverse effects of salt in food products is its ability to
promote lipid oxidation .

Salt has been reported to act as a prooxidant of lipid oxidation in meat (Ellis et al.,
1968; King and Earl, 1988; Kanner et al., 1991b; Arnold et al., 1991; Osinchak et al.,
1992; Ahn et al. 1993a, b). Salt increased lipid oxidation in dark meat turkey patties
(King and Earl, 198 8). Minced turkey tissue with salt was more oxidized than minced
turkey tissue alone (Kanner et al., 1991b). Many possible mechanisms have been
proposed for the prooxidant effect of salt. One mechanism has linked the prooxidant
effect of NaCl to iron ions in meat (Kanner et al.,l99lb; Osinchak et al.,l992). Kanner et
al. (1991b) proposed that NaCl increased the availability of iron ions during lipid
oxidation. They suggested that when free iron ions were added to minced turkey dark

muscle, a large part of the added iron interacted with protein macromolecules. This

2

interaction prevented iron ions from contacting with membranous lipids and acting as a
catalyst of lipid oxidation. Addition of NaCl interrupted the interaction between the iron
and the protein macromolecules, therefore, more free iron ions were available to interact
with the lipid fraction and enhance lipid oxidation.

Osinchak et a1. (1992) used a different model system (phosphatidylcholine
liposomes) but a similar approach to study the prooxidant effect of NaCl. They conﬁrmed
that the prooxidant effect of NaCl involved iron ions. They also observed that when using
NaCl and a high molecular weight fraction (>10 kilodalton) of mackerel press juice as a
prooxidant in the model system, the amount of lipid oxidation increased 7-8 times
compared to using the high molecular weight ﬁ'action (>10 kilodalton) of mackerel press
juice alone as a prooxidant. They speculated that part of the iron was released from the
high molecular iron-containing protein in mackerel muscle press juice when NaCl was
present. Therefore, free iron dramatically increased lipid oxidation in this model system.

Iron in high molecular weight iron-containing proteins such as hemoglobin,
myoglobin and ferritin accounted for 51.6 to 74.5% of total iron depending on the type of
muscle (Hazell, 1982). It was demonstrated that iron-containing proteins could release
iron during cooking (Love and Pearson 1974; Oellingrath, 1988) or during storage
(Decker and Hultin, 1990). Cooking denatures the heme protein, degrades the heme and
subsequently releases iron (Oellingrath, 1988). Salt may have a similar effect as cooking
on the denaturation of iron-containing proteins. Salt can alter the stability of meat protein,
causing a reduction in denaturation temperature (Kijowski and Mast, 198 8). Ahn and
Maurer (1989) reported that NaCl will decrease the heat stability of hemoglobin and

myoglobin. Because of their high iron concentration and possible destabilization by NaCl,

3

it is possible that part of the free iron released is from iron-containing proteins in the

presence of NaCl.

Although Kanner et al. (1991b) and Osinchak et al.(1992) suggested that the
prooxidant effect of NaCl is to increase the availability of iron ions by either breaking the
iron-protein macromolecule or releasing the iron from the high molecular weight fraction
(>10 kilodalton) of mackerel muscle press juice. There are few data which illustrate the
relationship between NaCl and the release of iron ions in meat. There are also few data
available to identify the free iron source in the presence of NaCl. The purposes of these
studies are to determine the relationship between NaCl and iron ion release in meat and to
deﬁne the possible source of free iron in meat using a model system. These studies are
based on the hypothesis that the prooxidant effect of NaCl in meat is due to its
ability to enhance the availability of iron ions to promote lipid oxidation. This is
done by: either facilitating the release of iron-ions from iron-containing proteins
(hemoglobin and myoglobin); and/or by breaking the protein-iron complex in the
meat system.

Four objectives were established to verify this hypothesis:

1. To compare existing iron analysis methodologies and establish an iron analysis method
that would provide reproducible data in studies designed to evaluate the role of salt in
the oxidation process;

2. To establish the effect of NaCl on the release of non-heme iron and lipid oxidation in
ground pork;

3. To study the effect of NaCl on the release of iron from iron-containing proteins in a

model system; and

4

4. To determine the effect of NaCl on the binding of free iron to protein macromolecules

in a model system.

REVIEW OF LITERATURE

Introduction

Lipid oxidation is a major deteriorative reaction in meat and meat products during
storage. It has an adverse effect on ﬂavor, color, texture and nutritive value of meat. Iron
is believed to be the most important catalyst for lipid oxidation in biological systems
(Kanner et al., 1987). Different forms or complexes of iron can directly initiate lipid
oxidation or facilitate the generation of very reactive species such as the hydroxyl radical
which is capable of initiating lipid oxidation.

Sodium chloride has been reported to act as a prooxidant in lipid oxidation of
meat. Raw and cooked dark meat turkey patties with 1-2% NaCl oxidized more rapidly
than dark meat turkey patties that did not contain salt(King and Earl, 198 8). In earlier
studies, NaCl was reported to act either as an antioxidant (Chang and Watts, 1950) or as a
prooxidant (Ellis et al., 1968). The prooxidant effect of NaCl in the lipid oxidation is
further complicated by the fact that salt may contain metal contaminants, which could
serve as catalysts of lipid oxidation. The exact mechanism for the prooxidant effect of
NaCl is still unclear. This review provides an overview of the lipid oxidation mechanism,

how iron catalyzes lipid oxidation in biological systems, the prooxidant effect of various

6

salts in meat, possible mechanisms involved in the prooxidant of salt and the problems

related to non-heme iron measurement methods of analysis.

Lipid Oxidation
Lipid oxidation is a major deteriorative reaction in meat and meat products during
storage. It leads to the development of oﬂlﬂavors, nutrient and shelf life reduction (Ho
and Chen, 1994). Lipid oxidation products are implicated in the disruption of biological
membranes, the inactivation of enzymes, and damage to proteins. (Frankel, 1984; Kubow,

1993; Marshall and Elswyk, 1995; McCord, 1994).

l. Mechanism of lipid oxidation
Initiation: LH —' Lo
propagation: L- + 02 _’ L000
mm + LH “’ LOOH + L-
LOOH —’ L00 + 00H
Lo- + LH—’ LOH + L-
termination L'OOO + L"OO- —» L'CO + L"CHzOH + Oz
L'OO- + L" —v L'OOL" (stable products)
L'0 + L". —* L'L"
Lipid oxidation is a free radical chain reaction (Bolland and Gee, 1946).
Generally, it can be divided into three distinct phases: initiation, propagation and
termination. Lipid oxidation is initiated by abstracting a hydrogen atom from lipid (LH)

and producing a free radical (L0). The resulting carbon-centered radical (L0) can react

7
with oxygen to form a peroxyl radical (L000). The peroxyl radical can abstract hydrogen

from adjacent fatty acid side chains and form a hydroperoxide and another free radical
which will ﬁIrther propagate the chain reaction of lipid oxidation. Hydroperoxides are
unstable and will undergo homolytic and heterolytic cleavage to form various types of free
radicals (L0- and OOH). These free radicals will further propagate the chain reaction.
During the termination step, free radicals are eliminated by formation of nonradical

products with other free radicals.

2. Role of iron in lipid oxidation

The initiation reaction in lipid oxidation is not only endothermic (Evans and Uri,
1949), but also a rate-determining step. The direct reaction of lipids with oxygen is spin-
forbidden because the ground state of lipids is of singlet multiplicity whereas that of
oxygen is of triplet multiplicity. Therefore, lipid oxidation is unlikely to be a major
reaction in meat unless catalysts or other factors are involved. However, this spin-
forbidden barrier does not apply to reactions with single electrons, hydrogen atoms or
other atoms or molecules containing unpaired electrons such as transition metals or free
radicals. Iron is the most important and abundant catalyst in biological systems.
Biological oxidation is due almost exclusively to metal promoted reactions (Aisen and
Liskowsky, 1980; Harrison and Hoare, 1980; Kanner, 1994). Iron can have two or more
valences. Therefore, the iron has a range of accessible oxidation states which enables

itself to transfer electrons.

8

Mechanisms of iron-catalyzed lipid oxidation depend on the presence or absence of
preformed lipid hydroperoxides. In the presence of hydroperoxides, iron catalyzes the
decomposition of preformed hydroperoxides to form LO- radicals (F enton reaction). The
decomposition of these hydroperoxides increases the rate of chain re-initiation or
propagation. This reaction is referred to as LOOH-dependent lipid oxidation (Gardner,
1989). In the absence of preformed LOOH, or when the concentration of LOOH is very
low, hydrogen peroxide promotes the Fenton reaction (Fee, 1982)and hydroxyl radical
formation (superoxide driven-Fenton reaction). Iron can also promote lipid oxidation
either by directly initiating lipid oxidation (Uri, 1961) or by metal autooxidation (Smith
and Dunkley, 1962)which produces reactive oxygen species such as Of/HOzo and H202.
These reactions are summarized below:

LOOH-dependent lipid oxidation;

Fe” + LOOH —+ Fe” + LO- + OH
LOO + L—eLH + R0

superoxide-driven Fenton reaction;

02- + F63+ H 02 + F62+
202_'i’21'l+ —’ 2H202 +02
Fe 2* + H202 —> Fe3+ + 00H + OH

direct initiation;
Fe“ + LH—> Fe” + L- + If
metal iron autooxidation;
H+
Fe 2+ + 02—' Fe 3+ + 02’ ——> H020

02—/ H02. —’ H202 + 02

(L, LH = lipid; LOOH = lipid hydroperoxide; LOO = alkoxy radical; R0 = lipid free radical
02" = superoxide radical; OOH = hydroxyl radical; F e2+ = ferrous iron; Fe3+ = ferric iron).

9

One particular reason why iron is so important in lipid oxidation is that it acts as a
catalyst in hydroxyl radical production (the superoxide driven-Fenton reaction). The
hydroxyl radical is believed to be necessary for the initiation of lipid oxidation (F ong et al.,
1973; Girotti and Thomas, 1984). The hydroxyl radical is the most potent oxidant that
can be formed from oxygen (E° =1.6 V). It is capable of oxidizing lipids and any other
biological molecules. However, some researchers (Samokyszyn et al., 1989; Minotti and
Aust, 1989; Gutteridge and Halliwell, 1990) doubt that the hydroxyl radical is a primary
initiator of lipid oxidation because it is a short-lived radical with diffusion-limited activity.
As such, the hydroxyl radical will not diffuse into the hydrophobic interior of a
phospholipid membrane to initiate lipid oxidation. Evidence such as the lack of correlation
between hydroxyl radical production and the rate of lipid oxidation (Minotti and Aust,
1987b; Braughler et al., 1986) indicates that the hydroxyl radical is not a primary initiator
of lipid oxidation. The addition of hydroxyl radical scavengers such as catalase (to
remove hydrogen peroxide and block hydroxyl radical formation) also rarely inhibits lipid
oxidation (Minotti and Aust, 1987a). However, supporters of the hydroxyl radical theory
argue that the generation of the hydroxyl radical could be site speciﬁc, such that the
damaging radical species would be generated in the proximity of the polyunsaturated fatty
acids (Schaich and Borg, 1988; Fukuzawa and Fujii, 1992). Schaich and Borg (1988)
have found that iron and hydrogen peroxide can partition from water to lipid. Therefore,
the Fenton reaction may take place in the lipid.

Other forms or complexes of iron have been suggested as catalysts of lipid
oxidation. These include iron-oxygen complexes, hypervalent iron (ferryl, perferryl iron,

porphyrin cation radical) and iron-containing proteins such as heme protein or

10
lipoxygenase (Kanner et al., 1987; Minotti and Aust, 1992; Kanner, 1994). The form of

iron is involved in lipid oxidation has not been fully determined. However, it is generally
agreed that iron plays a vital role in the initiation of lipid oxidation. Some of the more
commonly proposed mechanisms will be discussed.

Aust and colleagues (Samokyszyn et al., 1989; Minotti and Aust, 1989) proposed
that the iron-oxygen complexes (F ezi-Oz-Fey) are responsible for the initiation of lipid
oxidation. They studied the effect of different ratios of Fe” and Fe3+ on lipid oxidation in
liposomes (Minotti and Aust, 1987c). The rate of lipid oxidation was greatest when
approximately 50% of Fe2+ had been oxidized to Fe”. Lipid oxidation did not occur when
all the iron remained in the reduced form or when all the Fe2+ had oxidized to Fe“. They
observed that maximum lipid oxidation occurs when Fe 2+ and Fe 3+ are present in a 1:1
ratio and postulated that the function of the superoxide radical and hydrogen peroxide is
not to form hydroxyl radicals but rather to promote Fe3+ reduction or Fe2+ oxidation from
which the appropriate Fe2+/ Fe3+ initiating species originated.

Another active form of iron is hypervalent iron (iron valence of +4 to +6). These
hypervalent iron complexes include ferryl (FeO)2+, perferryl (F e0)3 i and porphyrin cation
radicals (heme-associated ferryl species). The interaction of hydrogen peroxide with
metmyoglobin or methemoglobin leads to the generation of active forms of heme proteins
(porphyrin cation radical)(Rhee, 1988). It was reported that porphyrin cation radical can
initiate membrane lipid oxidation (Kanner et al., 1991a; Kanner and Hare], 1985). The
proposed mechanism is as follows:

P*-Fe*“=o + LH -> P-Fe*‘=o + L- + PF

1.0 'i' 02 '9 L00.

ll

LOOO + LH 9 LOOH + LO

P-Fe*‘=o + LOOH -> P-Fei3-LOOH -) P*-Fe*‘=o +LOH

Although ferryl or perferryl iron complexes also contain hypervalent iron, they may
not be as signiﬁcant as the porphyrin cation radicals in lipid oxidation. This is because of
the fact that hypervalent iron complexes require macromolecular complexes or a very high
pH for formation and stabilization. Thus, hypervalent iron should not be a signiﬁcant
reactant since, in complex biological systems, decay of hypervalent states to Fe3+ would be
instantaneous, eliminating any enhanced catalytic capability.

Iron-containing proteins such as heme protein or lipoxygenase also play an
important role in lipid oxidation. Watts and Peng (1947) examined the prooxidant effect
of hog muscle extract and concluded that myoglobin and hemoglobin were the catalysts
responsible for fat oxidation. Tappel (1952) reported heme compounds such as
hemoglobin, myoglobin and cytochromes were the dominant catalysts of lipid oxidation in
meat systems. He proposed that heme iron catalyzed the decomposition of lipid
hydroperoxides into free radicals which propagated the free radical chain reaction.

LOOH + Heme-Fe2+ -) Heme-Fe2+ + LOO + OOH

It has also been suggested that heme catalysis may involve an indirect mechanism
in which the hydroxyl radical is produced and subsequently initiates lipid oxidation
reactions. The autooxidation of ferrous heme iron produced the superoxide radical which
was reduced to hydrogen peroxide in the meat system (Wallace et al., 1982; Sadrzadeh et
al., 1984). The hydrogen peroxide was then decomposed by iron to hydroxyl radicals and

subsequently promoted lipid oxidation. It is also possible that in systems which contain

12
hydrogen peroxide, the hydroxyl radical attacks hemoglobin and releases iron which then

catalyzes lipid oxidation as non-heme iron (Puppo and Halliwell, 1988).

Lipoxygenase is widely distributed in many plant, animal and ﬁsh tissues. It
catalyzes the insertion of oxygen into polyunsaturated fatty acids to generate position
speciﬁc hydroperoxides. The hydroperoxides can be further broken-down into several
secondary oxidation products which may promote lipid oxidation. It has been suggested
that lipoxygenase can initiate lipid oxidation in ﬁsh microsomal lipids (German and
Kinsella, 1985). Grossman et al. (1988) suggested that lipoxygenase may contribute to

lipid oxidation in chicken during frozen storage.

The Prooxidant Effect of Salt on Lipid Oxidation in Meat

Sodium chloride is a well recognized prooxidant in meat. In early studies, NaCl
was reported to act either as an antioxidant (Chang and Watts, 1950) or as a prooxidant
(Ellis et al., 1968). One possible cause of the dual role is the presence of impurities in the
salt. Impurities such as trace metals (e. g., iron and copper) or sodium nitrate may act as
prooxidants and antioxidants, respectively (Salih, 1986). Salih (1986) compared the
prooxidant effect of rock salt and calcium-magnesium free salt in turkey. He reported that
turkey treated with rock salt was more oxidized than turkey treated with calcium-
magnesium salt. He suggested that the impurities and the salt itself both could be active
prooxidants in ground turkey. However, Olson and Rust (1973) compared the effect of
low prooxidant salt (salt containing a low amounts of heavy metals: copper 0.1 ug/g salt
and iron 0.4 ug/g salt) and regular ﬂake salt (copper 1.5 ug/g salt and iron 1.5 ug/g salt)

on oxidative rancidity development in dry cured hams. They reported no differences in

13
oxidative rancidity between low prooxidant salt and regular fake salt in dry cured ham.

Subsequent studies by King and Earl (1988), Kanner et al. (1991b) and Ahn et al. (1993a)
using reagent grade NaCl demonstrated that salt itself was the prooxidant effect in ground
turkey. The prooxidant effect was also concentration-dependent. For example, minced
turkey or ground beef treated with increasing NaCl concentrations have higher lipid
oxidation than the meat treated with low NaCl concentrations. (Kanner et al., 1991b;
Torres et al., 1988).

It has also been suggested that NaCl itself is not a prooxidant but that it promotes
the activity of unsaturated fat oxidases (lipoxygenase) present in meat and ﬁsh. Lea
(193 7) observed that pork fat when treated with muscle juice and NaCl had higher
peroxide values than pork fat treated with muscle juice alone. He suggested that pork
muscle juice contains an enzyme (lipoxygenase) which accelerates the oxidation of pork
fat, an effect which can be enhanced by adding NaCl.

Salts other than NaCl, such as KCl and CaClz, can also promote lipid oxidation
(Rhee et al., 1983; King and Bosch, 1990; Ahn et al., 1993a). The prooxidant effect
varied among different salts (Rhee et al., 1983; King and Bosch, 1990). The consensus in
the literature has indicated that signiﬁcant diﬁ‘erences exist in the prooxidant effect of
different salts. Potassium chloride had almost no effect on rancidity development in raw
ground pork during frozen storage (Chang and Watts, 1949, 1950; Watts and Peng,
1947). Rhee et al. (1983) reported that KCl promoted lipid oxidation in raw ground pork
during refrigerated and frozen storage but had no prooxidant effect on cooked ground
pork. King and Bosch (1990) reported that KCl promoted lipid oxidation in dark turkey

meat. Under different storage temperatures (refrigerated and frozen storage), the order of

14
the prooxidant effect was diﬁ‘erent for MgClz and NaCl. For example, the prooxidant

eﬂ‘ect of MgClz was higher than NaCl in cooked ground pork stored for 4 days at 4 °C.
However, the prooxidant effect of MgClz was lower than NaCl during 2 months frozen
storage at -20 °C (Rhee et al., 1983).

There are two factors which may attribute to the lack of consensus on the
prooxidant effects of various salts on lipid oxidation. First, in most studies, the prooxidant
effects were compared at similar concentrations (w/w). However, in some studies, they
were compared at similar molarities. Because of the differences in molecular weights of
the salts, the same weight percentage will have different molarities. For example, for 0.15
M NaCl, KCl, NaBr and KBr, the respective weight percentages of the salts are 0.87%,
1.12%, 1.54 % and 1.79%, respectively. The second reason is that the data in the
literature have been very difficult to compare. Experiments have been performed under
different conditions, and it is difficult to extrapolate data from one set of studies to other
conditions as we discussed above. Different salts, salt concentrations, meat species and

processing/storage conditions were used, thus making comparisons difﬁcult.

Possible Mechanisms for the Prooxidant Effect of NaCl
It has been suggested that the prooxidant effect of NaCl may be related to: 1. the
release of iron from meat (Kanner et al., 1991b; Osinchak et al., 1992); 2. heme pigment
oxidation (Wallace et al., 1982); 3. water activity (Chang and Watts, 1950); 4. the physical
state of the meat (Shomer et al., 1987) and 5. enzyme activity (Lee et al., 1996). It is not
known whether the prooxidant effect of NaCl involves more than one mechanism or

combination of mechanisms which are dependent on each other.

15

1. Sodium chloride and free iron ion

Although NaCl is a well known prooxidant in meat, it does not promote lipid
oxidation in model systems such as emulsions of methyl linoleate and linoleic acid
(Mabrouk and Dugan, 1968) or water-washed muscle ﬁbers (Kanner et al., 1991b).
Mabrouk and Dugan (1968) reported that lipid oxidation in aqueous emulsions of methyl
linoleate and linoleic acid did not occur regardless of the concentration of the dissolved
NaCl. Kanner et al. (1991b) reported similar results in a model system containing water-
washed muscle ﬁbers. Osinchak et al. (1992) reported that lipid oxidation of
phosphatidylcholine liposomes were suppressed by increasing the concentration of
dissolved NaCl in the liposome system. Sodium chloride also inhibits reactions which lead
to lipid oxidation. Harel (1994) indicated that NaCl inhibited ascorbic acid oxidation and
the reactions that produce superoxide radicals, hydrogen peroxide and the hydroxyl
radical, species that are reported to be initiators of lipid oxidation. It is not clear why
NaCl inﬂuences lipid oxidation differently in meat and model systems.

Ellis et a1. (1968) suggested that NaCl may activate a component in lean meat
which results in a change of the oxidative characteristics of adipose tissue. It is also
possible that other compounds which are present in meat systems but not in model systems
may react with NaCl and promote lipid oxidation. Recently, Kanner et al. (1991b) and
Osinchak et al. (1992) indicated that the prooxidant effect of NaCl in meat may be related
to its ability to release free iron. Kanner et al. (1991b) proposed that the prooxidant
effect of NaCl is derived from its ability to displace iron ions from binding to protein
macromolecules. They proposed that free iron ions are major catalysts in meat. Most of

the free iron ions (chelatable iron) are in the cytosol fraction and are probably chelated by

16
proteins. When NaCl is added to the meat, the protein-iron interaction is disrupted. More

iron ions will be released to promote lipid oxidation. Kanner et al. (1991b) found that the
addition of iron salts to minced turkey only slightly increased the rate of lipid oxidation.
However, in the presence of NaCl, lipid oxidation increased approximately 3-5 times
compared to the control (turkey + iron ions). Kanner et al. (1991b) suggested that a large
portion of the added iron ions interacted with proteins in the meat. This interaction
prevented iron ions from aﬂ‘ecting membranous lipids and acting as catalysts of lipid
oxidation. Sodium chloride can disrupt the interaction between the iron ions and proteins.
Therefore, more free iron is available to interact with the lipid fraction and to enhance lipid
oxidation. This hypothesis is further supported by the observation that 3 NaCl solution
will extract more free iron ions than water during the preparation of water-washed muscle
ﬁbers (Kanner et al., 1991b). Sodium chloride may break the protein-iron interaction.
Thus, more iron ions will be released and fewer free iron ions will remain in the water-
washed muscle ﬁbers.

Kanner et al. (199 lb) reported that ethylenediaminetetraacetic acid (EDTA) and
ceruloplasmin (ferroxidase) can suppress NaCl-promoted oxidation in minced turkey meat.
EDTA, a metal chelator, will chelate metal ions and reduce the free metal ion
concentrations in meat. Ceruloplamin is an enzyme which works speciﬁcally on the
oxidation of ferrous iron to ferric iron. Because the ﬁmctions of EDTA and ceruloplasmin
are closely related with iron ions, it was postulated that iron ions are involved in NaCl-
promoted lipid oxidation (Kanner et al., 1991b).

Osinchak et al. (1992) used a different model system (phosphatidylcholine

liposomes and mackerel press juice) with a similar approach to study the prooxidant effect

17
of NaCl. They conﬁrmed the results of Kanner et al. (1991b) that the prooxidant eﬂ‘ect of

NaCl involved iron ions. They further suggested that part of the iron may be released
from the high molecular weight ﬁaction (>10 kilokalton) of the mackerel muscle press
juice. They reported that using NaCl and the high molecular weight fraction (>10
kilodalton) of mackerel press juice as a prooxidant increased the amount of lipid oxidation
by 7-8 times when compared to that generated by the high molecular weight ﬁ'action (>10
kilodalton) of mackerel press juice alone. They suggested that this was due to iron ions
being released ﬁ'om the high molecular weight fraction (>10 kilodalton) of the mackerel
press juice. The effectiveness of NaCl to displace iron may be linked to the high
stabilization constant of chlorine (Kanner et al., 1991b). Although Kanner et al. (1991b)
and Osinchak et al. (1992) proposed that the mechanism for the prooxidant effect of NaCl
is to increase the availability of non-heme iron in meat, no data have been presented to

prove NaCl directly increases the non-heme iron concentrations in meat.

2. Anion-promoted autooxidation

In addition to lipid oxidation, a problem encountered on adding NaCl to muscle is
heme pigment discoloration (Andersen and Skibsted, 1991; Andersen et al., 1990; Asghar
et al., 1990). Lipid and heme pigment oxidation are closely related. The possible
chemical coupling mechanism between heme pigment oxidation and lipid oxidation in meat
is not completely understood. It has been suggested that pigment oxidation will produce
superoxide radicals and initiate lipid oxidation (Andersen et al., 1990; Akamittath et al.,
1990). On the other hand, lipid oxidation also causes pigment discoloration. O’Grady et

al. (1996) reported that stimulation of lipid oxidation with FeCl3/ascorbate led to an

18

increase in both lipid and oxymyoglobin oxidation in a microsome-enriched muscle
ﬁaction model system. Oxymyoglobin oxidation did not occur when the lipid was
removed ﬁom the microsome-enriched muscle fraction containing the prooxidant

(F eCls/ascorbate).
Wallace et al. (1982) indicated that certain anions (Cl’, CN, N3-) either increase

the rate of oxymyoglobin oxidation (anion-promoted autooxidation) by binding to heme
iron and changing its redox potential in favor of oxidation of Fe2+ to Fe”, or by
preferentially stabilizing the Fe3+ oxidation state relative to the Fe2+ state of heme iron.
This anion-promoted autooxidation process not only leads to the formation of
metmyoglobin but also to the formation of superoxide free radicals (Wallace et al., 1982),
which is subsequently converted to hydrogen peroxide. These species are highly reactive
and are capable of reacting directly either with lipid compounds in muscle, or with
metmyoglobin to produce an activated molecule, activated metmyoglobin (MMb-Hzoz)

that can initiate lipid oxidation (Rhee, 1988; Hare] and Kanner, 1985).

3. Water activity

The effect of water activity (a..) on the rate of lipid oxidation is well documented
(Berends, 1993; Gopala-krishna and Prabhakar, 1992). As 21., increases from 0 to about
0.3-0.4, lipid oxidation decreases. When aw increases from 03-04 to 0.75-0.85, lipid
oxidation increases. Lipid oxidation decreases again when aw is over 0.75 (Cheah and
Ledward, 1995; Karel and Yong, 1981). At low a.., lipid free radicals are soluble in the oil
fraction which allows them to diffuse longer distances and spread the reaction

(Kanner,1994). When aw increases from 0 to 0.3-0.4, water molecules lower the

l9
eﬁ‘ectiveness of metal catalysts such as copper and iron, form hydrogen bonds with

hydroperoxides and retard hydroperoxide decomposition. Water reduces oxygen diffusion
by forming a barrier for oxygen over the lipid surface and lipid oxidation decreases. When
a. increases from 0.4 to 075-085, the possible liberation of transition metals cause lipid
oxidation. When a. is over 0.85, the reactants are diluted and lipid oxidation decreases.
The effect of NaCl on lipid oxidation also depends on the amount of moisture
present in the system (Chang and Watts, 1950). Salt itself is probably not a prooxidant
but rather it shifts the aw of the food to an intermediate level where lipid oxidation occurs

at a more rapid rate.

4. Physical state of meat

Addition of NaCl to meat causes extraction of the myoﬁbrillar proteins and results
in a less intact muscle microstructure (Ofstad et al., 1995; Velinov et al., 1990). Kanner et
al. (1991b) reported that increasing the concentration of NaCl enhanced lipid oxidation in
raw minced muscle, especially after a freeze-thaw cycle. They suggested that NaCl and
freeze thaw cycles may cause ﬁssion of the intracellular compounds and the destruction of
the cell structure which further enhances lipid oxidation (Shomer et al., 1987). “Sodium
chloride may affect the physical state of meat in such a way that hemoglobin or other
catalysts would be brought into closer contact with the fat” (Chang and Watts, 1950).
Ahn et al. (1993b) reported that turkey breast patties which contained hemoglobin and
NaCl together were more oxidized than patties which contained only hemoglobin or NaCl
alone. Sodium chloride may have caused a physical change in the liposome structure due

to osmotic pressure differences between the inside and outside of the bilayer (Arnold et

20

al., 1991). The hydrophobic interior of a phospholipid membrane may be more accessible
to prooxidants such as non-heme iron. Thus, more lipid is exposed to the oxidation

process.

5. Enzyme activity

There are many enzymes which exist in meat that will ease the oxidative stress by
converting the superoxide radical and hydrogen peroxide into less active initiating species.
It has been suggested that changing the activity of these enzymes will affect lipid
oxidation. Lee et al. (1996) demonstrated that NaCl could accelerate oxidation by altering
the activity of catalase, glutathione peroxidase and superoxide dismutase antioxidant
enzymes. Thus, more superoxide radicals or hydrogen peroxide are present in meat which
may interact with meat components and form active species such as the porphyrin cation
radical. On the other hand, there are some enzymes existing in meat such as lipoxygenase
which will increase the oxidative stress by converting polyunsaturated fatty acids into lipid
hydroperoxides. Hamberg and Gerwick (1993) reported that the enzyme activity of 12-
lipoxygenase isolated from red marine algae by gel ﬁltration was weak (about 3 % the
amounts of total 12-lipoxygenase products formed from arachidonic acid) ; however,
addition of 0.8-1 M NaCl to a desalted enzymes preparation increased the activity 20-fold
(about 68% 12-lipoxygenase products). Addition of NaCl in all fractions containing 12-

lipoxygenase resulted in a pronounced stimulation of 12-lipoxygenase activity.

21

Non-heme Iron Measurement

Accurate determination of heme and non-heme iron concentrations in foods is
important not only for the assessment of iron bioavailability (Monsen, 1988; Davis et al.,
1992) but also for determining the role of iron in storage stability of foods (Miller et al.,
1994 a,b; Han et al., 1995). Igene et al. (1979) reported that the major prooxidant in
cooked meat was not myoglobin but ﬁee iron which was released from heme pigments
during heating. Kanner et al. (1991b) and Osinchak et al. (1992) suggested that the
prooxidant effect of NaCl may be related to free iron ions in meat. To further study the
role of iron and the prooxidant effects of NaCl, methods accurately determining iron in
meats are necessary.

The Schricker (Schricker et al., 1982), modiﬁed Schricker (Rhee and Ziprin,
1987), Igene (Igene et al., 1979), and the ferrozine method (Carter, 1971) are commonly
used assays to measure non-heme iron concentrations in meat. In general, these methods
consist of sample preparation and iron quantitation steps. In the sample preparation step,
non-heme iron is usually extracted either by an aqueous solution of a metal chelator such
as ethylenediaminetetraacetic acid (Igene et al., 1979), citrate-phosphate (Carter, 1971)
and pyrophosphate-trichloroacetic acid (TCA) (F 0y et al., 1967) or by solubilization in a
strong acid solution (hydrochloric-trichloroacetic acid, Rhee and Ziprin, 1987). The
extracts are then separated into soluble (non-heme) and insoluble (heme iron) ﬁactions by
the addition of a trichloroacetic acid solution. In the sample quantitation procedure, iron
is either directly quantitated by atomic absorption spectroscopy (Igene et al., 1979) or
reacted with a metal chelator such as ferrozine (Carter, 1971) or dipyridyl (Awt, 1970;

Narasinga Rao and Prabhavathi, 1978) to produce a color chromogen, which is then

22
quantitated by measuring its absorption. Although these methods have been used to

measure the non-heme iron concentration in meat, they were originally used for
quantitating total iron concentrations. There are no differences within the quantitation
procedures for non-heme and total iron within each method. The only differences are in
the preparation steps for extracting non-heme and total iron fractions.

These non-heme iron measurement methods have been evaluated by various
researchers (Chen, et al., 1984; Rhee and Ziprin, 1987; Ahn et al., 1993c; Carpenter and
Clark, 1995). However, some issues need to be further addressed which include the lack
of consensus in terminology needed to describe the form of the iron and the variation in

iron values from the same sample obtained using non-heme iron measurement methods.

1. Consensus in terminology

Non-heme iron (Igene et al., 1979; Schricker et al., 1982; Chen et al., 1984; Rhee
and Ziprin, 1987; Ahn et al., 1993c; Carpenter and Clark, 1995), ﬁee iron ions, chelatable
iron ions (Kanner et al., 1991b), ionic iron, complex iron, soluble iron (Lee and
Clydesdale, 1979), and diffusate iron or low molecular iron (Hazell, 1982; Love and
Pearson, 1974) are commonly used to describe the forrnsof iron in meat system. These
terms are often used interchangeably in papers and different terms have been used from
paper to paper by the same researchers. Because of the inconsistencies in the terminology
used (Han et al., 1995; Rhee and Ziprin, 1987), it is not clear whether “non-heme iron”,
“free iron” or “chelatable iron ion” describe the same iron form. For example, researchers
(Rhee and Ziprin, 1987; Ahn et al., 1993c) classiﬁed total iron ion into heme and non-

heme iron. Hazell (1982) classiﬁed total iron into water-insoluble and water-soluble iron.

23
The water-soluble iron was further separated into ferritin, hemoglobin, myoglobin, low

molecule weight iron (<12 kilodalton) or diﬂirsate iron. Lee and Clydesdale (1979)
divided total iron into elemental iron, insoluble iron, complexed and ionic iron.

On the other hand, it is unclear whether the methods used to measure non-heme
iron are capable of differentiating between the various forms of iron. Schricker et al.
(1982) deﬁned non-heme iron as "all of the iron ions not in the heme form which includes
the storage iron ions such as ferritin, transferrin and others ". The Schricker and the
modiﬁed Schricker methods both are capable of recovering added ferritin (Ahn et al.,
19930). The ferrozine method can recover most of the ferritin present (Ahn et al., 1993c).
However, the ability of the Igene (Igene et al.,1979) method to extract iron from ferritin
has not yet been examined. Rhee and Ziprin (1987) suggested that the method of Igene
(Igene et al.,1979) may underestimate the non-heme iron concentration because of
incomplete extraction/recovery of non-heme iron from storage proteins (ferritin,

transferrin and others).

2. Differences in iron values from the same sample between non-heme iron
measurement methods

Research groups have reported differences in iron values when analyzing the same
sample using different methods. For example, Ahn et al. (1993c) reported 3.6, 6.9 and 6.7
ug Fe/g of non-heme iron in turkey leg muscle using the ferrozine (Carter, 1971),
Schricker (Schricker et al., 1982) and modiﬁed Schricker assays (Rhee and Ziprin, 1987),

respectively. Rhee and Ziprin (1987) reported that the modiﬁed Schricker’s method gave

24
higher non-heme iron values compared to Igene (Igene et al.,l979) methodology. It is not

clear why diﬂ‘erent methods give different non-heme iron values for the same sample.
Chen et al. (1984) and Ahn et al. (1993c) suggested that the Schricker or modiﬁed
Schricker method overestimates the actual non-heme iron because iron may be released
from the heme pigments during the strong acidic extraction conditions and during heating.

The Schricker method was modiﬁed by Rhee and Ziprin (1987) because of iron
release from the heme pigments during extraction (Rhee and Ziprin, 1987; Ahn et al.,
1993c; Chen et al., 1984) and nonenzymatic browning. In the modiﬁed Schricker method,
sodium nitrite was added to prevent the release of iron from heme pigments. A second
blank was also applied to correct for higher iron values because of nonenzymatic browning
reactions. However, the modiﬁed Schricker method still gave higher iron values than the
ferrozine and Igene methods. It is not clear whether other factors exist which inﬂate non-
heme iron values in the modiﬁed Schricker method, or whether the method extracts more
non-heme iron from the sample when compared to the other non-heme iron methods.

To verify whether the different iron values obtained using the non-heme iron
measurement methods are due to the extraction procedures or some factors interfering
with the non-heme iron quantitation method, it is necessary to measure the total, non-
heme and heme iron in the same sample and compare values obtained using each method.
However, this has not been done. Regarding the mechanisms involved for the prooxidant
effect of NaCl, it is difﬁcult to provide precise data which relates iron release after
subsequent NaCl addition to lipid oxidation until differences in the commonly utilized
methods can be elucidated. Based on the information available, it cannot be determined

which explanation is valid. In many studies, only non-heme iron is measured, or one of

25
the values (non-heme or heme iron) is calculated by taking the difference from total iron

values (Ahn et al., 1993c, Rhee and Ziprin, 1987). Each fraction should be measured

using a precise analytical method.

26

References

Ahn, D.U., Wolfe, F.H. and Sim, J.S. 1993a. Prevention of lipid oxidation in pre-cooked
turkey meat patties with hot packaging and antioxidant combinations. J. Food Sci.
58:283.

Ahn, D.U., Ajuyah, A., Wolfe, F.H. and Sim, J .S. 1993b. Oxygen availability aﬂ‘ects
prooxidant catalyzed lipid oxidation of cooked turkey patties. J. Food Sci. 58:278.

Ahn, D.U., Wolfe, F.H. and Sim, J .S. 1993c. Three methods for determining nonheme
iron in turkey meat. J. Food Sci. 58:288.

Ahn, D.U. and Maurer, A.J. 1989. Eﬂ‘ects of sodium chloride, phosphate, and dextrose
on the heat stability of puriﬁed myoglobin, hemoglobin, and cytochrome c. Poultry
Sci. 6821218.

Aisen, P. and Liskowsky, I. 1980. Iron transport and storage proteins. Ann. Rev.
Biochem. 49:357.

Akamittath, J .G., Brekke, CJ. and Schanus, E.G. 1990. Lipid oxidation and color
stability in restructured meat systems during ﬁozen storage. J. Food Sci. 55: 1513.

Andersen, H.J., Bertelsen, G. and Skibsted, L.H. 1990. Color and color stability of hot
processed frozen minced beef. Results from chemical model experiments tested
under storage conditions. Meat Sci. 28:87.

Andersen, H.J. and Skibsted, L.H. 1991. Oxidative stability of frozen pork patties. Effect
of light and added salt. J. Food Sci. 5621182.

Arnold, A.R., Bascetta, E. and Gunstone, FD. 1991. Effect of sodium chloride on
prooxidant activity of copper (II) in peroxidation of phospholipid liposomes. J.
Food Sci. 56:571.

Asghar, A., Torres, B, Gray, J .I. and Pearson, AM. 1990. Effect of salt on myoglobin
derivatives in the sarcoplasmic extract from pre-and post-rigor beef in the
presence or absence of mitochondria and microsomes. Meat Sci. 27:197.

Awt, RC. 1970. Determination of iron in beer. J. Inst. Brewing. 76:299.
Berends, CL. 1993. Water activity and lipid oxidation on the rate of lipid oxidation at

constant oxygen concentration. MS. Thesis Michigan State Univ., E. Lansing.
MI.

27

Holland, IL. and Gee, G. 1946. Kinetic studies in the cherrristry of rubber and related
materials: 111 therrnochemistry and mechanisms of oleﬁn oxidation. Trans.
Faraday Soc. 42:236.

Borg, DC and Schaich, KM. 1988. Fenton reactions in lipid phases. Lipids 23:570.

Braughler, J .M., Duncan, LA. and Chase, KL. 1986. The involvement of iron in lipid
peroxidation: importance of ferric to ferrous ratios in initiation, J. Biol. Chem.
261 : 10282.

Carpenter, CE. and Clark, E. 1995. Evaluation of methods used in meat iron analysis
and iron content of raw and cooked meats. J. Agric. Food Chem. 43: 1824.

Carter, P. 1971. Spectrophotometric determination of serum iron at the submicrogram
level with a new reagent (ferrozine). Anal. Biochem. 40:450.

Chang, I. and Watts, BM. 1949. Antioxidants in the hemoglobin catalyzed oxidation of
unsaturated fats. Food Technol. 3:332.

Chang, I. and Watts, BM. 1950. Some effects of salt and moisture on rancidity of fat.
Food Res. 15:313.

Cheah, PB. and Ledward, DA. 1995. High pressure effects in lipid oxidation. J. Am. Oil
Chem. 72:1059.

Chen, C.C., Pearson, A.M., Gray, 11., Fooladi, M.H. and Ku, PK. 1984. Some factors
inﬂuencing the nonheme iron content of meat and its implications in oxidation. J.
Food Sci. 49:581.

Davis, CD, Malecki, EA. and Greger, J .L. 1992. Interactions among dietary manganese,
heme iron, and nonheme iron in women. Am. J. Clin. Nutr. 56:926.

Decker, EA. and Hultin, HO. 1990. Factors inﬂuencing catalysis of lipid oxidation by the
soluble fraction of mackerel muscle. J. Food Sci. 55:947.

Ellis, R., Currie, G.T., Thornton, F.E., Bollinger, NC. and Gaddis, AM. 1968. Carbonyls
in oxidizing fat. 11. The effect of the pro-oxidant activity of sodium chloride on
pork tissue. J. Food Sci. 33:555.

Evans, MG. and Uri, N. 1949. The dissociation constant of hydrogen peroxide and the
electron afﬁnity of the H02 radical. Trans. Faraday Soc. 45:224.

Fee, J .A 1982. In “Oxidase and Related Systems”, King, T.E., Mason, HS. and Morrison,
M (Eds), Pergamon Press, New York, p101.

28

Fong, K.L., McCay, P.B., Poyer,J.L., Keele, BB. and Misra, H. 1973. Evidence that
peroxidation of lysosomal membranes is initiated by hydroxyl free radicals
produced during ﬂavin enzyme activity. J. Biol. Chem. 24827792.

Foy, LA, Williams, H.L., Cortell, S. and Conrad, ME. 1967. A modiﬁed procedure for
the determination of nonheme iron in tissue. Anal. Biochem. 18:559.

Frankel, EN. 1984. Lipid oxidation: Mechanisms, products and biological signiﬁcances.
J. Am. Oil Chem. Soc. 6121908.

Fukuzawa, K. and Fujii, T. 1992. Peroxide dependent and independent lipid peroxidation:
Site-speciﬁc mechanisms of initiation by chelated iron and inhibition by or-
tocopherol. Lipids 272227 .

Gardner, H.W. 1989. Oxygen radical chemistry of polyunsaturated fatty acids. Free
Radical Biol. Med. 7:65.

German, J. B. and Kinsella, J .E. 1985. Lipid oxidation in ﬁsh tissue. Enzymatic initiation
via lipoxygenase. J. Agric. Food Chem. 33:680.

Girotti, AW. and Thomas, J .P. 1984. Damaging effects of oxygen radicals on rescaled
erythrocyte ghosts. J. Biol. Chem. 259: 1744.

Grossman, S., Bergman, M. and Sklan, D. 1988. Lipoxygenase in chicken muscle. J.
Agric. Food Chem. 36:1268.

Gopala-krishna, AG. and Prabhakar, J .V. 1992. Effect of water activity on secondary
products formation in autoxidizing methyl linoleate. J. Am. Oil Chem. 692 17 8.

Gutteridge, J .M.C. and Halliwell, B. 1990. The measurement and mechanism of lipid
peroxidation in biological systems. TIBS 15:129.

Hamberg, M. and Gerwick, W.H. 1993. Biosynthesis of vicinal dihydroxy fatty acids in the
red alga Gracilariopsis lemaneiformis: Identiﬁcation of a sodium-dependent 12-
lipoxygenase and a hydroperoxide isomerase. Arch. Biochem. Biophys. 305: 1 15.

Han, D., McMillin, K.W., Godber, J .S., Bidner, T.D., Younathan, MT and Hart, LT.
1995. Lipid stability of beef model systems with heating and iron fractions. J. Food
Sci. 60:599.

Harel, S. 1994. Oxidation of ascorbic acid and metal ions as affected by NaCl. J. Agric.
Food Chem. 4222402.

Harel, S. and Kanner, J. 1985. Muscle membranal lipid peroxidation initiated by H202-
activated metmyoglobin. J. Agric. Food Chem. 33: 1 188.

29

Harrison, PM. and Hoare, R]. 1980. Metals in Biochemistry, Chapman & Hall. London,
England. p 1.

Hazell, T. 1982. Iron and zinc compounds in the muscle of meats of beef, lamb, pork and
chicken. J. Sci. Food Agric. 33:1049.

Ho, CT. and Chen, Q. 1994. Lipids in food ﬂavors. In “Lipids in Food Flavors”, C.T. Ho
and TC. Hartman (Eds), ACS Symposium Series 558, ACS, Washington, DC,
p. 2-14.

Igene, J .0, King, J .A., Pearson, AM. and Gray, J .I. 1979. Inﬂuence of heme pigments,
nitrite, and non-heme iron on development of warmed-over ﬂavor (W OF) in
cooked meat. J. Agric. Food Chem. 27:838.

Kanner, J. 1994. Oxidative processes in meat and meat products: Quality implications.
Meat Sci. 36:169.

Kanner, J. and Hare], S. 1985. Initiation of membranal lipid peroxidation by activated
metmyoglobin and methemoglobin. Arch. Biochem. Biophys. 2372314.

Kanner, J ., Hare], S. and Granit, R. 1991a. Nitric oxide as an antioxidant. Arch. Biochem.
Biophys. 289:130.

Kanner, J. Hare], S. and J affe. R. 1991b. Lipid peroxidation of muscle food as affected by
NaCl. J. Agic. Food Chem. 39:101.

Kanner, J ., German, J .B. and Kinsella, J .E. 1987. Initiation of lipid peroxidation in
biological systems. CRC Critical Reviews in Food Science and Nutrition. 25:317.

Karel, M. and Yong, S. 1981. “Autoxidation-initiated reaction in food, in “Water Activity:
Inﬂuences on Food Quality.” L.B. Rockland and GP. Stewart (Ed), Academic
Press, New York, NY. p511.

Kijowski, J .M. and Mast, MG. 1988. Effect of sodium chloride and phosphates on the
thermal properties of chicken meat proteins. J. Food Sci. 53:367.

King, A.J. and Bosch, N. 1990. Effect of NaCl and KCl on rancidity of dark turkey meat
heated by microwave. J. Food Sci. 5521549.

King, A.J. and Earl, LA. 1988. Effect of selected sodium and potassium salts on TBA
values of dark meat turkey patties. J. Food Sci. 53:726.

Kubow, S. 1993. Lipid oxidation products in food and atherogenesis. Nutr. Rev. 51:33.

3O

Lea, CH. 193 7. Inﬂuence of tissue oxidases on rancidity. Oxidation of the fat of bacon.
J. Soc. Chem. Ind. 56:376T.

Lee, S.K., Mei, L. and Decker, E. 1996. Effect of NaCl on antioxidant enzyme activity
and lipid oxidation in ground pork. 1996 IFT Abstract. poster 68E-10. Inst. Food
Tech. (New Orleans, LA).

Lee, K. and Clydesdale, EM. 1979. Quantitative determination of the elemental, ferrous,
ferric, soluble, and complexed iron in foods. J. Food Sci. 44:549.

Love, J .D. and Pearson, AM. 1974. Metmyoglobin and non-heme iron as prooxidants in
cooked meat. J. Agric. Food Chem. 22:1032.

Mabrouk, AF. and Dugan Jr, LR. 1968. A kinetic study of the autoxidation of methyl
linoleate and linoleic acid emulsions in the presence of sodium chloride. J. Am.
Oil Chem. Soc. 37:486.

Marshall, A. and Elswyk, M.V. 1995. Dietary polyunsaturated fat. Is it beneﬁcial? Nutr.
Today. 30:207.

McCord, J .M. 1994. Free radicals and prooxidants in health and nutrition. Food Technol.
48:106.

Miller, D.K., Smith, V.L., Kanner, J., Miller, DD. and Lawless, HT. 1994a. Lipid
oxidation and warmed-over aroma in cooked ground pork from swine fed
increasing levels of iron. J. Food Sci. 59:751.

Miller, D.K., Gomez-Basauri, J.V., Smith, V.L., Kanner, J. and Miller, D.D. 1994b.
Dietary iron in swine rations affects nonheme iron and TBARS in pork skeletal
muscles. J. Food Sci. 59:747.

Minotti, G. and Aust, S.D. 1987a. The role of iron in the initiation of lipid peroxidation.
Chem. Phys. Lipids 44:191.

Minotti, G. and Aust, S.D., 1987b. An investigation into themechanism of citrate-Fe”-
dependent lipid peroxidation. Free Rad. Biol. Med. 3:279.

Minotti, G. and Aust, S.D., 1987c. The requirement for iron (III) in the initiation of lipid
peroxidation by iron (11) and hydrogen peroxide. J. Biol. Chem. 262: 1098.

Minotti, G. and Aust, SD. 1989. The role of iron in oxygen radical mediated lipid
peroxidation. Chem. Biol. Interactions 71:1.

Minotti, G. and Aust, SD. 1992. Redox cycling of iron and lipid peroxidation. Lipids
27:219.

31

Monsen, ER. 1988. Iron nutrition and absorption2dietary factors which impact iron
bioavailability. J. Amer. Diet. Assoc. 88:786.

Narasinga Rao, BS. and Prabhavathi, T. 197 8. An in vitro method for predicting the
bioavailability of iron from foods. Am. J. Clin. Nutr. 31: 169.

Oellingrath, I.M. 1988. Heat degradation of heme in methemoglobin and metmyoglobin
model systems measured by reversed-phase ion-pair high performance liquid
chromatography. J. Food Sci. 53:40.

Ofstad, R., Kidman, S., Myklebust, R., Olsen, KL. and Herrnansson, AM. 1995. Liquid-
holding capacity and structural changes in comminuted salmon (Salmo Salar)
muscle as inﬂuenced by pH, salt and temperature. Lebens WissenTechnol. 28:329.

O’Grady, M.N., Monahan, F.J., Mooney, M.T., Butler, F., Buckley, D.J., Kerry, J ., Allen,
P. and Keane, MG. 1996. Inhibition of oxymyoglobin oxidation by vitamin E.
In"meat for the consumer” published by MATFORSK, Norwegian Food
Research Institute. p. 100.

Olson, DC. and Rust, RE. 1973. Oxidative rancidity in dry-cured hams: effect of low
pro-oxidant and antioxidant salt formulation. J. Food Sci. 38:251.

Osinchak, J.E., Hultin, H.O. Zajicek, O.T., Kelleher, SD. and Huang, CH. 1992. Effect
of NaCl on catalysis of lipid oxidation by the soluble fraction of ﬁsh muscle. Free
Radical Biol. Med. 12:35.

Puppo, A. and Halliwell, B. 1988. Formation of hydroxyl radicals from hydrogen peroxide
in the presence of iron. Is hemoglobin a biological Fenton reagent? Biochem. J.
249: 185.

Rhee, KS. 1988. Enzymic and nonenzymic catalysis of lipid oxidation in muscle foods.
Food Technol. 42(6): 127.

Rhee, K.S., Smith, GO and Terrell, RN. 1983. Effect of reduction and replacement of
sodium chloride on rancidity development in raw and cooked ground pork. J.
Food Prot. 46:578.

Rhee, KS. and Ziprin, Y.A. 1987. Modiﬁcation of the Schricker nonheme iron method to
minimize pigment effects for red meats. J. Food Sci. 5221174.

Sadrzadeh, S.M.H., Graf, E., Panter, S.S., Hallaway, PE. and Eaton, J.W. 1984.
Hemoglobin: a biologic Fenton reagent. J. Biol. Chem. 259: 14354.

32

Salih, A.M. 1986. Lipid stability in turkey meat as inﬂuenced by cooking, reﬁigerated
and frozen storage, salt, metal cations and antioxidant. PhD. thesis, Michigan
State Univ., East Lansing, MI.

Samokyszyn, V.M., Miller, D.M., Reif, D.W. and Aust, SD. 1989. Inhibition of
superoxide and ferritin-dependent lipid peroxidation by ceruloplasmin. J. Biol.
Chem. 264:21.

Schricker, B.R., Miller, DD. and Stouffer, J .R. 1982. Measurement and content of
nonheme and total iron in muscle. J. Food Sci. 47:740.

Shomer, 1., Weinberg, 2G. and Vasilever, R. 1987. Structural binding properties of silver
carp muscle aﬂ‘ected by NaCl and CaClz treatments. Food rnicrostruct. 6:199.

Smith, G]. and Dunkley, W.L. 1962. Initiation of lipid peroxidation by a reduced metal
ion. Arch. Biochem. Biophys. 98:46.

Tappel, AL. 1952. Linoleate oxidation catalyzed by hog muscle and adipose tissue
extract. Food Res. 17:550.

Torres, E., Pearson, A.M., Gray, 11., Booren, A.M. and Shimokomaki, M. 1988. Effect
of salt on oxidative changes in pre- and post-rigor ground beef. Meat Sci. 23:51.

Uri, N. 1961. In “Autoxidation and antioxidants” Lundberg, W.O. (ed), Wiley-
Interscience, New York, p55.

Velinov, P.D., Zhikov, M.V. and Cassens, KG. 1990. The effect of tumbling, sodium
chloride and polyphosphates on the microstructure and appearance of whole-
muscle processed meats. Food Mircrostruct. 9291.

Wallace, W.J., Houtchens, R.A., Maxwell, J .C. and Caughey, W.S. 1982. Mechanism of
autoxidation for hemoglobins and myoglobins. J. Biol. Chem. 257:4966.

Watts, HM. and Peng, D. 1947. Lipoxidase activity of hog hemoglobin and muscle
extract. J. Biol. Chem. 1792441.

CHAPTER I

A COMPARISON OF THREE ANALYTICAL PROCEDURES FOR
DETERMINING NON-HEME IRON CONCENTRATION IN GROUND PORK

Abstract

Meat samples containing four concentrations of iron (0, 3, 6 and 9 ug F e/g meat) were
prepared by mixing ground pork with 10% of its weight of a solution containing 0, 33, 66 and 99
ug F e/ml ferrous chloride and forming into patties. The modiﬁed Schricker, Igene, and the
ferrozine procedures were used to quantitate the non-heme, heme and total iron concentrations.
The modiﬁed Schricker method realized greater overall recoveries of added iron than the ferrozine
and Igene methods in all pork samples. For similar pork samples, non-heme and heme iron
distribution was different when evaluated by the different analytical methods. Greater non-heme
iron (3.9 ug/g and 5.1 ug Fe/raw meat; 5.1 ug Fe/g cooked meat) and smaller heme iron
concentrations (8.8 ug F e/g raw meat; 7.9 ug Fe/g cooked meat) were found in the control patties
(i.e., no addition of iron) using the modiﬁed Schricker method. On the other hand, the Igene
method measured the smallest non-heme iron concentrations (1.1 ug F e/g raw meat; 2.2 ug F e/g
cooked meat) and the highest heme iron concentrations (10.1 ug Fe/g raw meat; 10.4 ug Fe/g
cooked meat). The data indicate that the procedures used for separating the non-heme and heme
iron account for the differences in the iron concentration observed in the samples. The quantitation

steps yield similar values.

33

34
Introduction

The accurate determination of heme and non-heme iron concentrations in foods is
important not only for the assessment of iron bioavailability (Monsen, 1988; Davis et al.,
1992) but also for the role of iron in the storage stability of foods. It was well established
by Igene et al. (1979) that the major prooxidant in cooked meat was not myoglobin but
free iron which was released from the heme pigment during heating. It had also been
suggested that non-heme iron is essential for the prooxidant effect of NaCl in meat
(Kanner et al., 1991b; Osinchak et al., 1992). To further study the role of iron forms in
nutrition and lipid oxidation, an accurate non-heme iron measurement method is
necessary.

The Schricker (Schricker et al., 1982), modiﬁed Schricker (Rhee and Ziprin,
1987), Igene et al. (1979), and ferrozine methods (Carter, 1979) are commonly used
assays to measure the non-heme iron concentration in meat. In general, each method can
be divided into sample preparation and iron quantitation steps. In the sample preparation
step, non-heme iron is extracted from the sample by a metal chelator solution such as
EDTA (Igene method) or phosphate/citrate (ferrozine method) or by acid solution
(modiﬁed Schricker method). The extracted irons can be directly quantitated by atomic
absorption spectroscopy or by forming a color complex with metal chelator (ferrozine or
bathophenanthroline) followed by quantitation by spectrophotometery. Although these
methods are widely used to measure non-heme iron in meat, they were originally used for
quantitating total iron in water. The quantitation procedure (either atomic absorption or

colorimetry) is the same for non-heme and total iron for each method. The only difference

35

in the methods is in the preparation of the samples for non-heme and total iron
quantitation.

Non-heme iron analysis measurement has been evaluated by various researchers
(Rhee and Ziprin,1987; Ahn et al., 1993; Carpenter and Clark, 1995) by adding iron to the
meat system and several questions remain unanswered. Ahn et al. (1993) reported 3.6,
6.9 and 6.7 ug non-heme iron/g meat in raw turkey leg using the ferrozine, Schricker and
modiﬁed Schricker assays, respectively. Rhee and Ziprin (1987) also reported that the
modiﬁed Schricker method produced higher non-heme iron concentrations (3.27 ug F e/g
meat) than the Igene (Igene et al., 1979) method (1.16 ug Fe/g meat) in the beef
semimembranosus muscle. It is not clear why different non-heme iron concentrations are
reported for the same sample when different assays are used.

Sample preparation and quantitation procedures are the two variables that may
contribute to the differences in non-heme iron concentrations attained by the various
methods. Ahn et al. (1993) suggested that the modiﬁed Schricker method may
overestimate non-heme iron concentrations because iron ions were released ﬁ'om the heme
proteins during sample preparation. In contrast, Rhee and Ziprin (1987) suggested that
the Igene method might underestimate the non-heme iron concentrations in meat due to its
inability to extract all of the non-heme iron in meat. No evidence is available to verify that
the difference in non-heme iron concentrations is due to different sample preparation
procedures.

The Schricker assay generally detects greater non-heme iron concentrations than
the ferrozine or Igene method. It was postulated by Rhee and Ziprin (1987) that the

higher iron concentrations are due to the release of iron from heme pigments and

36

nonenzymatic browning when using the Schricker method. The nonenzymatic browning
reaction produced a color complex which increased the apparent iron-chelator color
complex absorption, and therefore, overestimated the non-heme iron concentration. They
later modiﬁed the Schricker method by adding sodium nitrite to prevent iron release ﬁom
heme pigments and applying a second blank to correct for the greater iron concentrations
obtained because of nonenzymatic browning. Because the modiﬁed Schricker method still
gives greater non-heme iron concentrations than the ferrozine and Igene methods, it is not
clear whether other factors exist which will interfere with the iron quantitation step of the
modiﬁed Schricker method.

The percent iron recovered when a known amount is added to meat is a common
technique used to evaluate non-heme iron measurement methodology. However, iron
recovery in the Igene method has not been examined and compared to the other methods.
Whether the variability in non-heme iron concentrations for the same sample when
analyzed by different methods is due to sample preparation or quantitation procedures has
not been examined. The objective of this study was to evaluate the iron recovery of the
Igene method and to determine whether higher non-heme iron concentrations obtained
using the modiﬁed Schricker method are due to differences in sample preparation or

differences in the quantitation procedure.

37
Materials and Methods

Materials

Three fresh boneless pork legs were obtained within 24 hr of slaughter from the
Michigan State University Meat Laboratory. Ferrozine (3 -(2-pyridyl)-5,6-bis (4-phenyl
sulfonic acid)-1,2,4-triazine), bathophenanthroline disulfonic acid (sodium salt), and
thioglycolic acid (96-99%) were purchased from Sigma Cherrrical Co. (St. Louis, MO).
Trichloroacetic acid, ammonium acetate, nitric acid, ethylenediaminetetraacetic acid
(EDTA, disodium salt) and perchloric acid were obtained from J .T. Baker Inc.
(Phillipsburg, NJ). Sodium nitrite and hydrochloric acid were purchased from
Mallinckrodt Inc. (Paris, KY). Neocuproine was obtained from Fisher Scientific Company
(Fair Lawn, NJ). All chemicals were reagent grade. All glassware was immersed in 4 N
hydrochloric acid solution overnight to remove any trace iron and then rinsed with distilled
water to remove hydrochloric acid residues. Disposable plastic test tubes (Baxter
Healthcare Corp., Romulus, MI) were used to prevent possible iron contamination during

the analyses.

Sample preparation

The sample preparation procedures are illustrated in Figure 1-1. All visible fat was
removed and the pork was cut into 1 cm x 1 cm cubes, overwrapped in a plastic bag and
frozen at -26 °C overnight. The frozen pork cubes were placed in a cheesecloth and
fractured using a hammer. The fractured pork was powdered by mixing with dry ice and

homogenizing in a blender (Tekmar Company, Cincinnati, OH). Connective tissue was

38

Pork (legs)
Trim visible
+— fat and cut into
small cubes
Frozen (-26 °C)
ove ’ght
«— + dry ice
Homogenize
in blender

screen through
i ‘ a sieve (0.16 X 0.16 mm)

Store at -26 °C

 

overnight
Weigh riieat powder (100g)
add different cone.
‘ FeClz solution
V

Mix 3.0 min. in beaker

Cook

 

 

immerse in 83 °C water bath

to internal
temperature 70 °C

0, 3, 6 and 9
ug/g eat

 

 

 

Raw
0, 3, 6, 9
ug/g meat
Total iron analysis

 

 

heme ironlanalysis

Non-hemi iron analysis

Figure 1-1 Sample preparation scheme for comparing modiﬁed Schricker, ferrozine and Igene

methodology

Ferrozine
method

4 g meat
+4 g water

+ 4 ml EDTA

0
homo genize

 

centrifuged at
3,000xG for
10 min.

Transfer the
supernatant to
a test tube

 

measiired
non-heme iron
using atomic
absorption
spectroscopy

39

ferrozine
method

3 g meat + 9 ml
citrate-phosphate
buffer

1

homogenize

+4 ml
ascorbic acid

incubate at
room temperature
for 15 min

+8ml
TCA

centrifuge at
3,000xG for
10 nrin.

uusiers m1

supernatant to
test tube

+ 2 ml buffer
+ 0.5 m1 ferrozine
solution

 

V

measure absorbance at
563 nm

modiﬁed
Schricker
method

5 gmeat

+ 5 ml water
+ 0.4 ml NaNOz
solution

equilibrate for
30 min

+ 15 ml acid
mixture
(HC1+TCA)

incubated in a water
bath at 65 °C for
20 min.

Cooled at
cold water

Allow the tube to settle
until the supernatant
become clear

1 ml acidic liquid
+ 5 m1 bathophenanthroline
solution

1

measure absorbance at
540 nm

Figure 1-2 Flow diagram for different non-heme iron measurement procedures

40

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Meat Sample
a d a id a d
d1 on dig on dig 'on
total total total
iron ., iron V iron modiﬁed
Igene ferrozine Schricker
sample sample sample
preparation preparation preparation
procedure procedure procedure
precipitate precipitate precipitate
a id a id a id
dig tion dige 'on dig tion
heme iron heme iron heme iron
fraction fraction fraction
supeTatant mm+Mt supenftant
non-heme non-heme non-heme
iron iron iron
fraction fraction fraction
\ similar quantiﬁcation
,, procedure were used to

 

quantiﬁcation

by Igene, ferrozine,
and modiﬁed
Schricker methods

measured total, heme,

and non-heme iron fractions
prepared by ferrozine and
modiﬁed Schricker methods

Figure 1-3 Flow chart for the evaluation of different iron quantitation procedures

41
removed by passing the powdered meat though a sieve (0.16 x 0.16 mm). The powdered

tissues were stored overnight at -26 °C to permit evaporation of the remaining dry ice.
Four treatments of iron (0, 3, 6 and 9 ug F e/g meat) were added by hand-mixing 100g
pork with 10 ml ferrous chloride solution (0, 33, 66 and 99 ug F e/g meat) for 3 min with
spatula in a 250 ml beaker. For the control samples (i.e., no added iron), 10 ml distilled
water were added to the powdered meat and mixed in a similar fashion. Cooked samples
were heated to an internal temperature of 70 °C in a water bath (temperature = 83 i 2 °C).

For the quantitation of total iron, the samples were prepared using nitric acid and
perchloric acid digestion (Igene et al., 1979). In the preparation step for non-heme iron
fractions, samples were separated into supernatant (non-heme samples) according to Igene
et al. (1979), ferrozine (Carter, 1971), and modiﬁed Schricker (Rhee and Ziprin, 1987)
procedures and precipitate fractions (Figure 1-2). The recovery is calculated by
[(measured non-heme iron in sample - non-heme iron in control) x 100%]/added non-
heme iron. The precipitate was further digested by nitric acid and perchloric acid and
identiﬁed as the heme iron fractions (Figure 1-3). Aliquots of the digests or extracts for
total iron, heme iron and non-heme iron were quantitated by the Igene (atomic

absorption), ferrozine and modiﬁed Schricker (colorrnetic) methods (Figure 1-3).

Data analysis

The experiments were designed as a four factor (quantitation method x preparation
method x treatments x replication) randomized complete block design with balanced data
(Gill, 1978). Three replicates were carried out in this study. Means, standard errors, sum

of squares and mean square errors were calculated using the MSTAT-C microcomputer

42
statistical program (Michigan State University, 1989). The student t-test was used to

make contrasts between treatments (Gill, 197 8).

43

Results and Discussion

Recovery of added iron using three methods

Recovery data for added iron to raw and cooked samples using the three non-heme
iron methods are presented in Figures 1-4 and 1-5, respectively. For the raw and cooked
control samples, the modiﬁed Schricker realized greater (p<0.05) non-heme iron
concentrations than the Igene method. This is consistent with iron data for beef that was
reported by Rhee and Ziprin (1987). Greater non-heme iron concentrations were obtained
by the modiﬁed Schricker method compared to the ferrozine method. These are
consistent with turkey data reported by Ahn et al. (1993). These investigations
demonstrated higher non-heme iron concentrations with the modiﬁed Schricker method
than with the ferrozine method. A similar trend was found for the various iron treatments
(Figures 1-4 and 1-5). The order of non-heme iron recovery for the methods was
modiﬁed Schricker > ferrozine > Igene. These data suggest that a substantial portion of
the added iron is not completely recovered by the Igene method. The recovery rate
decreased signiﬁcantly (p<0.05) when the levels of added iron was increased. It appears
that the capacity for this method to extract non-heme iron may be exhausted as more iron
is added. The lower recoveries may also be due to the added iron binding to protein
macromolecules in meat systems (Kanner et al., 1991) and extraction procedures which
are not strong enough to separate the iron-protein macromolecular interactions due to the
low pH (pH = 5.6) in the extraction procedure. These results agree with Rhee and
Ziprin’s (1987) hypothesis which suggested that the Igene method underestimates the non-

heme iron concentration of a meat sample.

44

The recovery of added iron using the ferrozine method is consistent with
recoveries reported by Ahn et al. (1993) for cooked samples (100%). However, the
recovery from raw samples is about 10% lower than the data of Ahn et al. (1993). The
difference in iron recovery might be due to the differences in muscle species and sample
preparation. To conﬁrm this hypothesis, turkey was ground (instead of powdered) as
described by Ahn et al. (1993) and the experiment with added iron was repeated. The
percent recovery of added iron in the raw sample was similar to data (100%) previously
reported by Ahn et a1. (1993). Ahn’s investigations suggested that because both the
ferrozine and modiﬁed Schricker methods were able to completely recover the added iron
from raw ground turkey, the differences in non-heme iron concentrations were not due to
recovery. However, the data in the present study with indicates that the ferrozine method
produced a different recovery (about 90%) when raw ground pork or different sample
preparation procedures are used (Figures 1-4 and 1-5). The recovery of non-heme iron

may be affected by species as well as physical sample preparation procedures.

Evaluation of three sample quantitation procedures

Both raw and cooked samples were separated into non-heme and heme fractions
using methods described by Igene et al. (1979), Carter (1971) and Rhee and Ziprin
(1987). The total, non-heme and heme iron fractions were quantitated using the Igene,
ferrozine and modiﬁed Schricker methods (Tables 1-1 and 1-2). There were no signiﬁcant

differences (p>0.05) in total, non-heme and heme iron concentrations quantitated by the

45

 

Igene Ferrozine Modified
Schricker

 

[:13 ug Felg meat I6 ug Felg meat I9 ug Felg meat

 

Figure 1-4. The recovery of added iron from raw ground pork using three
procedures for determining non-heme iron (recovery with different letters
within the same quantitation procedure are signiﬁcant, p<0.05)

46

 

Igene Ferrozine Modified
Schricker

 

Dilug Felgmeat .8ugFe/gmeat .9ug Felmeat

 

Figure 1-5. The recovery of added iron from cooked ground pork using three
procedures for determining non-heme iron (recovery with different letters
within the same quantitation procedure are signiﬁcant, p<0.05)

47

Table 1-1. Comparison of iron quantitation procedures of raw pork extracted using the

 

 

Igene methodl.
modiﬁed
Fraction Treatment Igenez’3 ferrozine Schricker
Total iron pork (control) 10210.4 11.0109 10811.3
(ug Fe/g meat) pork (3ug Fe/g meat) 13611.7 14211.1 14011.7
pork (6ug Fe/g meat) 15.9108 17.5109 16911.8
pork (9ug Fe/g meat) 18811.1 20.1101 20011.6
Non-heme iron pork (control) 1.1102 1.3107 1.3101
(ug Fe/g meat) pork (3ug Fe/g meat) 2.9105 3.2113 3.3106
pork (6ug Felg meat) 3.8103 4.0115 3911.0
pork (9ug Fe/g meat) 4.2103 4411.3 4.7105
Heme iron pork (control) 10110.7 10211.2 9811.4
(ug Fe/g meat) pork (3ug Fe/g meat) 12811.4 12.5109 11911.3
pork (6ug Fe/g meat) 15.2114 15011.1 14211.6
pork (9ug Felg meat) 17.4120 17.8109 16711.6

 

1. Sample separation into heme, non-heme and total iron using the Igene method and
quantiﬁcation of each fraction using the Igene, ferrozine, and modiﬁed Schricker

methods

2. Mean 1 standard deviation (n =3).

w

treatment in total, non-heme and heme iron.

. There is no signiﬁcant difference (p<0.05) between quantitation procedures within each

48

Table 1-2. Comparison of iron quantitation procedures of cooked1 pork extracted using

 

 

the Igene methodz.
modiﬁed
Fraction Treatment Igene3'4 ferrozine Schricker
Total iron pork (control) 11.4125 12411.8 14.7128
(ug Felg meat) pork (3ug Felg meat) 15.6121 15.6103 18.2123
pork (6ug Fe/g meat) 18.3126 19.3124 20.8128
pork (9ug Fe/g meat) 20913.1 22011.3 23.9129
Non-heme iron pork (control) 2210.6 2111.0 2310.1
(ug Felg meat) pork (3ug Felg meat) 3.4109 3.4109 3.2101
pork (6ug Felg meat) 4211.3 3.911.] 3.6101
pork (9ug Felg meat) 4.8116 4311.3 4.2103
Heme iron pork (control) 10412.0 9411.8 10411.2
(ug Fe/g meat) pork (3ug Fe/g meat) 12411.7 12811.5 11.7114
pork (6ug Fe/g meat) 15.6109 15.9107 15211.7
pork (9ug Fe/g meat) 17.8118 18311.6 16.6120

 

l. The cooked samples were heated to an internal temperature of 70°C in a 83°C water

bath.

2. Sample separation into heme, non-heme and total iron using the Igene method and
quantiﬁcation of each fraction using the Igene, ferrozine, and modiﬁed Schricker

methods.

0)

Mean 1 standard deviation (n =3).

4. There is no signiﬁcant diﬁ‘erence (p>0.05) between quantitation procedures within each
treatment in total, non-heme and heme iron.

49
three methods for raw and cooked samples. These data agree with those of Carpenter and

Clark (1995) who also reported there were no differences in iron concentrations obtained
using ferrozine (colorrnetric) and atomic absorption spectroscopy quantitation methods.

Similar results were obtained for both raw and cooked samples prepared using the
modiﬁed Schricker sample preparation procedures (Tables 1-5 and 1-6). However, non-
heme iron concentrations of fractions prepared using the ferrozine procedure (Tables 1-3
and 1-4) were signiﬁcantly (p<0.05) smaller when quantitated by the modiﬁed Schricker
and Igene procedures than by the ferrozine method. This may be because a citrate-
phosphate buffer used in the ferrozine procedure interferes with the modiﬁed Schricker
method and because the ferrozine procedure does not use an acid digestion step. In the
iron quantitation step a metal chelator (bathophenanthroline) was used to chelate metal
and form a color complex in the modiﬁed Schricker method. Both citrate and phosphate
are metal chelators which may compete with the bathophenanthroline for metal.
Therefore, the sample preparation step may have competing chelating reactions which
cause interference. Total and heme iron data indicate that the quantitation procedures are
not responsible for the differences in non-heme iron concentrations reported for these

different methods.

Evaluation of three sample preparation procedures
Data from samples, which were separated into heme and non-heme fractions and
quantitated as described by Igene et al. (1979), Carter (1971) and Rhee and Ziprin (1987),

are presented in Table 1-7. The non-heme iron concentrations in control

50

Table 1-3. Comparison of iron quantitation procedures of raw pork extracted using the
ferrozine methodl.

 

 

modiﬁed
Fraction Treatment Igene” ferrozine Schricker
Total iron pork (control) 10.2104 11.0109 10811.3
(ug Fe/g meat) pork (3ug Fe/g meat) 13611.7 14211.1 14.0117
pork (6ug Felg meat) 15.9108 17.5109 16911.8
pork (9ug Fe/g meat) 18811.1 20.1101 20011.6
Non-heme iron4 pork (control) 1.0104A 2610.3B 1.1105A
(ug Felg meat) pork (3ug Felg meat) 2.3103A 5310.2B 2,510.5A
pork (6ug Fe/g meat) 3,310.4A 8.21033 3.8104"
pork (9ug Fe/g meat) 4510.5A 10410.4B 5,010.4A
Heme iron pork (control) 8.311.] 10.3107 11.5123
(ug Fe/g meat) pork (3ug Fe/g meat) 8511.6 10211.1 10711.1
pork (6ug Fe/g meat) 9011.2 11.7115 11.7115
pork (9ug Fe/g meat) 9611.6 11411.2 12211.2

 

1. Sample separation into heme, non-heme and total iron using the ferrozine method and
quantiﬁcation of each fraction using the Igene, ferrozine, and modiﬁed Schricker

methods

2. Mean 1 standard deviation (n =3).
3. There are no signiﬁcant differences Q9005) between quantitation procedures within

each treatment in total and heme iron.

4. Non-heme iron concentrations in each row with the same capital letter superscript are
not signiﬁcantly different (p<0.05).

Table 1-4. Comparison of iron quantitation procedures of cooked1 pork extracted using
the ferrozine methodz.

 

 

modiﬁed
Fraction Treatment Igene 3’4 ferrozine Schricker
Total iron pork (control) 11.4125 12411.8 14.7128
(ug Fe/g meat) pork (3ug Fe/g meat) 15.6121 15.6103 18.2123
pork (6ug Fe/g meat) 18.3126 19.3124 20.8128
pork (9ug Fe/g meat) 20913.1 22011.3 23.9129
Non-heme irons pork (control) 2310.9A 4.0103B 3.111.7A
(ug Fe/g meat) pork (3ug Felg meat) 4.2112A 7.1105B 4,911.1A
pork (6ug Felg meat) 6.0107" 9,910.9B 6.6120"
pork (9ug Fe/g meat) 7,511.3A 12.4109B 8,112.8A
Heme iron pork (control) 8.4122 10.7107 11.7111
(ug Fe/g meat) pork (3ug Fe/g meat) 9.0128 10.6107 12011.1
pork (6ug Fe/g meat) 9711.9 11.9103 13.1104
pork (9ug Fe/g meat) 9.2124 11.9103 13111.1

 

1. The cooked samples were heated to an internal temperature of 70°C in a 83°C water

bath.

2. Sample separation into heme, non-heme and total iron using the ferrozine method and
quantiﬁcation of each fraction using the Igene, ferrozine, and modiﬁed Schricker

methods

3. Mean 1 standard deviation (n =3).
4. There are no signiﬁcant differences (p>0.05) between quantitation procedures within

each treatment in total and heme iron.

5. Non-heme iron concentrations in each row with the same capital letter superscript are
not signiﬁcantly different (p<0.05).

52

Table 1-5.Comparison of iron quantitation procedures of raw pork extracted using the

 

 

modiﬁed Schricker method‘.

modiﬁed

Fraction Treatment Igenez'3 ferrozine Schricker
Total iron pork (control) 10210.4 11.0109 10.8113
(ug Fe/g meat) pork (3ug Felg meat) 13.6117 14.2111 14.0117
pork (6ug Fe/g meat) 15.9108 17510.9 16.9118
pork (9ug Fe/g meat) 18.8111 20110.1 20011.6
Non-heme iron pork (control) 4.2103 4.8105 3.9103
(ug Felg meat) pork (3ug Fe/g meat) 7.1111 8.0114 7.4113
pork (6ug Fe/g meat) 10010.3 10611.1 10.0104
pork (9ug Fe/g meat) 12810.6 14.3114 13.4111
Heme iron pork (control) 6510.7 8511.2 8.8121
(ug Fe/g meat) pork (3ug Fe/g meat) 6.7126 8.6115 8.7118
pork (6ug Fe/g meat) 7.3121 8.7113 8.9116

pork (9ug Fe/g meat) 7.8113 9.2113 9.8110

 

1. Sample separation into heme, non-heme and total iron using the modiﬁed Schricker
method and quantiﬁcation of each fraction using the Igene, ferrozine, and modiﬁed
Schricker methods

2. Mean 1 standard deviation (n =3).

5”

each treatment in total, non-heme and heme iron.

There are no signiﬁcant differences (p>0.05) between quantitation procedures within

53

Table 1-6.Comparison of iron quantitation procedures of cooked1 pork extracted using

the modiﬁed Schricker methodz.

 

 

modiﬁed
Fraction Treatment Igene3’4 ferrozine Schricker
Total iron pork (control) 11.4125 12411.8 14.7128
(ug Felg meat) pork (3ug Felg meat) 15.6121 15.6103 18.2123
pork (6ug Fe/g meat) 18.3126 19.3124 20.8128
pork (9ug Fe/g meat) 20913.1 22011.3 23.9129
Non-heme iron pork (control) 5.3109 6.4114 5.1104
(ug Felg meat) pork (3ug Felg meat) 8.0110 9.4117 8.7102
pork (6ug Fe/g meat) 10910.8 12511.9 12.4106
pork (9ug Fe/g meat) 13.3108 16.0118 15.2110
Heme iron pork (control) 7.1122 8.6113 7.9112
(ug Fe/g meat) pork (3ug Fe/g meat) 8.1115 9,111.6 8,411.2
pork (6ug Fe/g meat) 7.6121 9.3119 8.9120
pork (9ug Fe/g meat) 8.8118 10211.6 9.5115

 

1. The cooked samples were heated to an internal temperature of 70°C in a 83°C water

bath.

2. Sample separation into heme, non-heme and total iron using the ferrozine method and
quantiﬁcation of each fraction using the Igene, ferrozine, and modiﬁed Schricker

methods
. Mean 1 standard deviation (n =3).
There are no signiﬁcant differences (p>0.05) between quantitation procedures within

h)

each treatment in total, non-heme and heme iron.

54

Table 1-7. Heme, non-heme and total iron concentrations in the raw pork prepared
following by Igene, ferrozine and modiﬁed Schricker methods.

 

 

 

modiﬁed
Fraction Treatment Igenel ferrozine Schricker
Total iron pork (control) 10.2104 11.0109 10811.3
(ug Fe/g meat) pork (3ug Fe/g meat) 13.6117 14.2111 14.0117
pork (6ug Fe/g meat) 15.9108 17510.9 16.9118
pork (9ug Fe/g meat) 18.8111 20.1101 20.0116
Non-heme pork (control) 1.1 10.2"A 2.6103” 3 .9103.“C
iron (ug Fe/g pork (3ug Fe/g meat) 2.9105“ 5.3102LB 7.411.3"’C
meat) pork (6ug Felg meat) 3.810.3"’°’A 3.210.313B 10.11104“c
pork (9ug Fe/g meat) 42:03:"A 10.410.4‘3 13,411.11”C
Heme ironz’3 pork (control) 10.1107“ 10310.7LA 8.812.1“’B
(ug Fe/g meat) pork (3ug Fe/g meat) 12811.4"A 10211.1LA 8.711 .8“’13
pork (6ug Fe/g meat) 15.21145A 11.7115LB 8.911.6‘LC
pork (9ug Fe/g meat) 17.41201A 11.4112LB 9.8110“
1. Mean 1 standard deviation (n =3).
2. Treatments with different superscript lower case letters within the same column and

iron fraction are signiﬁcantly different (p<0.05).
3. Treatments in each row with the same capital letter superscript are not signiﬁcantly
different (p<0.05).

55
samples prepared and quantitated by the Igene, ferrozine and modiﬁed Schricker method

are 1.1, 2.6 and 3.9 ug Fe/g meat, respectively. Since there are no differences in the
quantitation steps (colormetric and atomic absorption) of the three non-heme iron
methods, the difference in the non-heme iron concentrations in the control samples are due
to differences in the preparation of the samples. The modiﬁed Schricker procedure
extracts more non-heme iron than does the ferrozine and the Igene methods. The same
trend was also found for the samples containing the three levels of added iron. Data for
the cooked pork samples, separated and quantitated as described above, are presented in
Table 1-8. These data exhibited the same trend as the raw samples. The modiﬁed
Schricker sample preparation method gave higher non-heme iron concentrations than the
ferrozine and the Igene methods.

The heme iron concentrations have an inverse relationship when compared to non-
heme iron (Tables 1-7 and 1-8). The trend for all heme samples is: Igene > ferrozine >
modiﬁed Schricker methodology. The modiﬁed Schricker method, which gave the highest
non-heme iron value, produced no signiﬁcant differences (p>0.05) in heme iron
concentration in all the iron addition treatments (Table 1-7). On the other hand, the Igene
method realized the smallest non-heme iron recovery and signiﬁcantly different (p<0.05)
heme iron concentrations among the various treatments. The heme iron data (Table 17)
indicates that the ferrozine and the Schricker methods extract more added iron. This
statement can be made because no signiﬁcant differences (p>0.05) were found in the heme
iron fractions. The data for cooked pork, separated and quantitated as described above,

are presented in Table 1-8. Similar trends as in raw samples

56

Table 1-8. Heme, non-heme and total iron concentrations in the cooked1 pork prepared

following by Igene, ferrozine and modiﬁed Schricker methods

 

 

modiﬁed
Fraction Treatment Igene2 ferrozine Schricker
Total iron pork (control) 11.4125 12411.8 14.7128
(ug Felg meat) pork (3ug Fe/g meat) 15.6121 15.6103 18.2123
pork (6ug Fe/g meat) 18.3126 19312.4 20812.8
pork (9ug Fe/g meat) 20.9131 22.0113 23912.9
Non-heme pork (control) 2210.6M 4.0103"B 5.1104”:
iron (ug Felg pork (3ug Fe/g meat) 3.4109"A 7. 110.51B 8.7102“:
meat) pork (6ug Fe/g meat) 4,211.3”A 9.914191B 12.4_+_—o.t5""~C
pork (9ug Fe/g meat) 4.8116ILA 12.41091 3 15.211 .o“~C
Heme iron3'4 pork (control) 10412.0M 10.71071"A 7,911.2“B
(ug Felg meat) pork (3ug Fe/g meat) 12411.7“A 10610.7"J3 8.411 .2°’C
pork (6ug Fe/g meat) 15.6109“ 11,910.31?B 8.912.0°’C
pork (9ug Fe/g meat) 17.8118“ 11910.3“3 9,511.5“B

 

l. The cooked samples were heated to an internal temperature of 70°C in a 83°C water

bath.

2. Mean 1 standard deviation (n =3).

Lo)

are signiﬁcant (p<0.05).
4. Treatments in each row with the same capital letter superscript are not signiﬁcantly
different (p<0.05).

Treatments with different superscript letters within the same column and iron ﬁactions

57

were found. The Igene method gave the lowest non-heme iron concentrations and the
highest heme iron concentrations among the non-heme iron measurement methods. The
modiﬁed Schricker procedure produced the highest non-heme iron concentrations and the
lowest heme iron concentrations compared to the other two non-heme iron measurement
methods.

Another question which needs to be addressed is “which method will give non-
heme iron values which are closest to the actual concentration in meat?”. The Igene
method should not be used because of its poor iron recovery. For the ferrozine method,
recovery of the added iron ranged from about 90-100% depending on which specie or
preparation step (grinding or powdering) was used. Ahn et al. (1993) reported that the
recovery of non-heme iron from the iron-containing proteins such as ferritin is about 80-
87% in ground turkey. Based on Ahn’s data, the ferrozine method may underestimate the
actual non-heme iron concentration in meat. For the modiﬁed Schricker method, Ahn et
al. (1993) suggested that the modiﬁed Schricker method overestimates the actual non-
heme iron present in the sample. This is most likely due to iron released from the heme
pigments as a result of the strong acidic conditions and subsequent heating during
extraction. However, there are no data to support the above hypothesis. On the other
hand, Rhee and Ziprin (1987) presented data which indicate that sodium nitrite applied to
the modiﬁed Schricker method can effectively prevent the release of the iron ﬁom heme
proteins even under acidic and heating conditions. Small amounts of iron detected in the
non-heme fraction were released from the heme fraction using this extraction procedure.
Based on iron recovery and data reported by Rhee and Ziprin (1987), the modiﬁed

Schricker method will minimally overestimate the actual non-heme iron in meat because

58

sodium nitrite can effectively prevent the release of iron ﬁom the heme proteins. The

modiﬁed Schricker sample preparation method is the most effective way to completely
extract non-heme iron from meat samples and was used in subsequent studies. Atomic
absorption spectroscopy will be used for iron quantitation because it is rapid and it can

handle many samples.

59

Summary and Conclusions

The sample preparation step, not the iron quantitation step, contributes to the
diﬂ‘erences in iron concentrations realized by the three methods under evaluation. The
modiﬁed Schricker method produced higher values for non-heme iron because it extracted
more iron ﬁom the sample than the ferrozine and Igene procedures. The Igene procedure
was not an effective procedure to extract the non-heme iron. These data indicate that the
ferrozine method may not completely recover added iron from either pork or turkey
samples. The modiﬁed Schricker method is the most effective procedure to completely

extract non-heme iron from meat.

60

References

Ahn, D.U., Wolfe, F.H. and Sim, J .S. 1993. Three methods for determining nonheme iron
in turkey meat. J. Food Sci. 58:288.

Carpenter, CE. and Clark, E. 1995. Evaluation of methods used in meat iron analysis
and iron content of raw and cooked meats. J. Agric. Food Chem. 43: 1824.

Carter, P. 1971. Spectrophotometric determination of serum iron at the submicrogram
level with a new reagent (ferrozine). Anal. Biochem. 40:450.

Chen, C.C., Pearson, A.M., Gray, J .I., Fooladi, M.H. and Ku, PK. 1984. Some factors
inﬂuencing the nonheme iron content of meat and its implications in oxidation. J.
Food Sci. 492581.

Davis, CD, Malecki, EA. and Greger, J .L. 1992. Interactions among dietary manganese,
heme iron, and nonheme iron in women. Am. J. Clin. Nutr. 56:926.

Gill, J .L. 1978. “Design and Analysis of Experiments in the Animal and Medical
Sciences.” Iowa State Univ. Press, Ames, IA.

Igene, J .0, King, J .A., Pearson, A.M. and Gray, J .I. 1979. Inﬂuence of heme pigments,
nitrite, and non-heme iron on development of warmed-over ﬂavor (WOF) in
cooked meat. J. Agric. Food Chem. 27:838.

Kanner, J ., Harel, S. and Jaffe. R. 1991. Lipid peroxidation of muscle food as affected by
NaCl. J. Agic. Food Chem. 39:101.

Monsen, ER. 1988. Iron nutrition and absorptionzdietary factors which impact iron
bioavailability. J. Am Diet. Asso. 88:786.

Osinchak, J.E., Hultin, H.O., Zajicek, O.T., Kelleher, SD. and Huang, CH. 1992. Effect
of NaCl on catalysis of lipid oxidation by the soluble fraction of ﬁsh muscle. Free
Radical Biol. Med. 12:35.

Rhee, KS. and Ziprin, Y.A. 1987. Modiﬁcation of the Schricker nonheme iron method to
minimize pigment effects for red meats. J. Food Sci. 52: 1 174.

Schricker, B.R., Miller, DD. and Stouﬁ‘er, J .R. 1982. Measurement and content of
nonheme and total iron in muscle. J. Food Sci. 47:740.

CHAPTER II

THE EFFECT OF VARIOUS SALTS ON NON-HEME IRON
CONCENTRATIONS AND LIPID OXIDATION IN GROUND PORK

Abstract

The effect of NaCl concentration (0, 8.8, 17.4 and 26.1 mg/g meat) and different
salts (NaCl, KCl, NaBr and KBr) on non-heme iron concentration and lipid oxidation was
studied in raw and cooked ground pork. As NaCl concentration increased from 0 to 26.1
mg/g meat , lipid oxidation increased (p<0.05) in both raw and cooked samples when
monitored by thiobarbituric acid-reactive substances (TBARS) and peroxide values.
Increasing NaCl concentration increased non-heme iron concentrations for raw samples at
day 6 and for the cooked samples immediately after cooking (day 0). Treatments
containing different salts had higher (p<0.05) TBARS and peroxide values than the
control. Different salts (N aCl, NaBr and KBr) also signiﬁcantly (p<0.05) increased non-
heme iron concentrations in raw samples at day 6 of storage. At the same molarity, the
prooxidant effect and non-heme iron concentrations between different salts were not

signiﬁcantly different.

61

62
Introduction

Sodium chloride has been established as a prooxidant in meat (King and Bosch,
1990; Arnold et al., 1991; Kanner et al., 1991; Osinchak et al., 1992; Ahn et al. 1993a,b;).
However, its exact mechanism as a prooxidant is unclear. Recently, researchers have
linked the prooxidant eﬁ‘ect of NaCl to the release of iron ions in meat (Kanner et al.,
1991; Osinchak et al., 1992). Kanner et a1. (1991) proposed that NaCl increases the
release of iron ions and makes them available for lipid oxidation; and they postulated that
when free iron ions were added to ground turkey, a large portion of the added iron
interacted with protein macromolecules. This interaction prevented iron ions ﬁom
reacting with membranous lipids and acting as a catalyst of lipid oxidation. Sodium
chloride interrupts the interaction between iron ions and protein macromolecules.
Therefore, more free iron is available to interact with the lipid fraction and catalyze lipid
oxidation. Osinchak et al. (1992) used a similar approach as Kanner et al. (1991) when
studying the prooxidant effect of NaCl in a model system (phosphatidylcholine liposomes
and mackerel press juice). They conﬁrmed that the prooxidant effect of NaCl involved
iron ions. In order to verify the hypothesis that the prooxidant effect of NaCl in meat
involves iron, it is necessary to establish that NaCl does increase free iron in meat systems.
There are few research data available relative to the effect of NaCl on the non-heme iron
concentration in pork. One factor which adds to the difﬁculty of studying this question is
the inherent difficulty with the non-heme iron measurement method itself as discussed in
chapter I.

Salts other than NaCl, such as KCl and MgClz, have also been shown to promote

lipid oxidation in meat systems (Rhee et al., 1983; Ahn et al., 1993a; King and Bosch,

63
1990). Because of the similarity in the chemical composition of these alkali and alkali-

earth halides (LiCl, NaCl, NaF, KC], NaBr and NaI), researchers have attempted to
investigate their prooxidant mechanisms by examining the effects of the salts and the
eﬁ‘ects of their anions and cations on lipid oxidation in meat systems (Ellis et al., 1968;
Osinchak et al., 1992; Wettasinghe and Shahidi, 1996). Diﬁ‘erent salts have different
prooxidant effect in lipid oxidation in meat (Rhee et al., 1983; King and Bosch, 1990).
However, the consensus in the literature regarding their relative prooxidant eﬁ‘ect in lipid
oxidation indicates that extensive differences exist. For instance, KC] had no effect on
rancidity development in raw ground pork during frozen storage (Chang and Watts, 1949,
1950; Watts and Peng, 1947). However, Rhee et a]. (1983) reported KC] promoted lipid
oxidation in raw ground pork but had no prooxidant effect in cooked ground pork. King
and Bosch (1990) reported that KCl promoted lipid oxidation in ground turkey. Under
speciﬁc storage conditions the prooxidant effect of different salts is also not consistent.
For example, the prooxidant effect of MgC]; is higher than NaC] in cooked ground pork
during reﬁigerated storage. However, the prooxidant effect of Mng is lower than NaCl
during frozen storage (Rhee et al., 1983). It is not understood why these differences
occur as the prooxidant effect of various salts is studied. It is also not clear whether these
salts share the same prooxidant mechanism as NaCl, or if their prooxidant effects on lipid
oxidation are due to their varying ability to release iron ions.

Since the prooxidant effect of NaCl is related to concentration, treatments with
higher NaC] concentrations had greater lipid oxidation (King and Bosch, 1990; Torres et
al., 1988). The ﬁrst objective of this study is to examine the effect of NaC] concentration

on lipid oxidation and non-heme iron release in ground pork. The second objective of this

64

research is to study the effect of different salts on non-heme iron concentrations and lipid

oxidation.

65
Materials and Methods

Pork (legs) from market hogs were obtained from a local meat company within 48
hr of slaughter. The fresh legs were sealed in plastic bags and covered with ice during
transportation. Bathophenanthroline disulfonic acid (sodium salt), thioglycolic acid (96-
99%), NaCl, KC] NaBr and KBr were purchased ﬁ'om Sigma Chemical Co. (St. Louis,
MO). Trichloroacetic acid, ammonium acetate, nitric acid and perchloric acid were
obtained from J .T. Baker Inc. (Phillipsburg, NJ). Sodium nitrite and hydrochloric acid
were purchased from Mallinckrodt Inc.(Paris, KY). All chemicals were reagent grade.
Possible contamination of glassware with trace quantities of iron was eliminated by
immersing in 4 N hydrochloric acid solution overnight and rinsing with distilled water the
following day. Disposable plastic test tubes (Baxter Diagnostics Inc., McGraw Park, IL)

were used to prevent possible iron contamination.

Sample preparation

Bone and skin were removed from the legs immediately after arriving at Michigan
State University. All visible fat was removed and the pork was ground (Hobart, Troy,
OH) twice (through a 9 mm and a 3 mm plate). For the NaCl concentration study, 150g
ground pork and 15 ml of NaCl solutions of varying concentrations (1.65, 3.27 or 4.91
M) were hand mixed using a spoon in a 250 ml beaker for 1.5 min at 4 °C to give target
NaCl concentrations of 8.8, 17.4 and 26.1 mg/g meat (equal to about 0.15, 0.30 and 0.45
M NaCl, respectively). For the control (no NaCl), 15 ml of distilled water were added to
the pork and mixed in a similar manner. For the second study, 150 g ground pork were

mixed with 15 ml of 1.65 M salt solutions (NaCl, KC], NaBr and KBr) to reach a target of

66
8.8 mg salt/g meat (equal to a 0.15 M salt concentration). For the control, 15 ml distilled

water were added to the ground pork and similarly mixed. In the cooked study, 165 g
samples were left in the 250 ml beaker after salt incorporation, covered with aluminum foil
and a thermometer was inserted to the center of the meat to monitor the internal
temperature during cooking. The samples were immersed in a 83 i 2 °C water bath and
cooked to an internal temperature of 70 °C (Igene et al.1979; Apte and Morrissey, 1987).
Samples were stored in a reﬁigerator (4 °C) during the study.

Lipid oxidation was monitored by measuring the development of TBARS
(thiobarbituric acid-reactive substances) and peroxide values. The TBARS measurement
was based on the method of Tarladgis et a1. (1960) as modiﬁed by Cracke] et al. (1988).
Propyl gallate and EDTA were added to the sample in the modiﬁed method (Crackel et
al., 1988) to prevent any ﬁirther lipid oxidation during sample preparation and TBARS
were expressed as mg malonaldehyde/Kg sample. Peroxide values were measured
according to Shantha and Decker (1994). The method is based on the principle of the
rapid peroxide-mediated oxidation of ferrous to ferric iron. The latter, in the presence of
cyanide ion, forms a ferric-cyanide color complex which can be measured using a
spectrophotometer. Total iron ﬁactions were prepared as described by Igene et al.
(1979). This method utilizes a nitric acid and perchloric acid digestion followed by
quantitation utilizing atomic absorption spectroscopy. Non-heme iron was determined
according to the modiﬁed Schricker method (Rhee and Ziprin, 1987) and quantitated
using atomic absorption spectroscopy as discussed in chapter I. Non-heme iron ions were
ﬁrst extracted from the samples using a HCl/TCA acid solution in 65 °C water bath for 20

hr (Figure 1-2). The non-heme iron concentration of the acid extract was directly

67

measured using atomic absorption spectroscopy. Lipid oxidation and non-heme iron were
monitored on days 0, 3 and 6 for raw samples and days 0, l and 2 for cooked samples.
Storage time designated as day 0 represents analyses carried out immediately after the salt
or distilled water was mixed with the meat.
Data analysis

The experiment utilized a split-plot design with repeated measurements (Gill,
1978). Means, standard errors, sum of squares, mean square errors and the least
signiﬁcant difference (LSD) test were calculated using the MSTAT-C microcomputer
statistical program (Michigan State University, 1989). Three replicates were carried out

in each study.

68

Results and Discussion

Effect of NaCl concentrations on TBARS and Peroxide values

The effects of NaCl concentrations on lipid oxidation in the raw and cooked pork
systems were monitored by measuring TBARS and peroxide values (Tables 2-1 and 2-2).
For the raw pork study, the addition of NaCl signiﬁcantly (p<0.05) increased TBARS and
peroxide values during the 6 day reﬁigerated storage (4 °C) period. On the other hand,
the treatment without NaCl (control) produced almost constant values during the storage
period (Table 2-1). At day 0 there were no signiﬁcant differences (p<0.05) in the extent
of lipid oxidation between pork sample treatments monitored by both TBARS and
peroxide values. After 3 days of reﬁigerated storage, the addition of NaCl (8.8, 17.4 and
26.1 mg/g meat ) to the pork resulted in higher TBARS values than the control.
Increasing NaCl concentrations produced greater (p<0.05) TBARS (26.1 mg NaCl/g meat
> 17.4 mg NaCl/g meat > 8.8 mg NaCVg meat > control). These data are consistent
with those reported by Rhee et a]. (1983) and Akamittath et al. (1990) indicating that
NaCl is a prooxidant. Increasing NaCl concentration increased lipid oxidation (King and
Bosch, 1990; Kuo and Ockerman, 1985; Torres et al., 1988). At day 6, the TBARS
values for the 26.1 mg NaCl/g meat treatments were not statistically different from those
for the 17.4 mg NaCl/g meat treatments. However, all values were signiﬁcantly Q)<0.05)
greater than the control. Trends established by peroxide value measurements were similar
to those obtained by measuring TBARS (Table 2-1). Statistical signﬁcance was not the

same for TBARS and peroxide value methods.

69

Table 2-1. The eﬂ‘ect of NaCl concentrations on TBARS and peroxide values in raw
ground pork during reﬁ'igerated storage (4 °C)

 

TBARSl (mg malonaldehyde/kg meat)

 

 

Treatment Day 02’3 Day 3 Day 6
1. pork (control) 0.3“ 0.3“ 0.3“
2. pork (8.8 mg NaCl/g meat) 0.3“ 2.7M3 6.0ho
3. pork (17.4 mg NaCl/g meat) 0.5“ 4.2c8 8.7°°
4. pork (26.1 mg NaCl/g meat) 0.7“ 5.6“ 9.1cc

Peroxide valuesl
(milliequivalents of peroxide values/kg sample)

 

 

Day 0 Day 3 Day 6
1. pork (control) 1.6“ 2.3“ 2.4“
2. pork (8.8 mg NaCl/g meat) 2.4“ 3.2“ 9.5“
3. pork (17.4 mg NaCl/g meat) 3.5"A 4.6“’A 14.5CB
4. pork (26.1 mg NaCl/g meat) 3.1“ 7.2“ 18.8“

 

H
o

TBARS mean square error = 0.72, peroxide values mean square error = 2.12.

Means with different superscript letters within the same column are signiﬁcantly
different (p<0.05).

3. TBARS or peroxide values in each row with the same capital letter superscript are not
signiﬁcantly different (p<0.05).

.N

70
For cooked samples (Table 2-2), similar trends in lipid oxidation between

treatments were observed as was for the raw samples. Increasing NaCl concentrations
signiﬁcantly (p<0.05) increased lipid oxidation as measured by both TBARS and peroxide
values during two days of refrigerated storage. However, the most signiﬁcant diﬁ‘erences
in the lipid oxidation between treatments occurred at the early stage of storage (day 0)
rather than at the later stages of storage. For example, at day 0, NaCl treatments (8.8,
17.4 and 26.1 mg NaCl/g meat ) had signiﬁcantly higher TBARS and peroxide values than
the control. After 1 day of reﬁigerated storage, there was no signiﬁcant difference in lipid
oxidation between 8.8 mg NaCl/g meat treatment and the control monitored by TBARS
and peroxide values. The NaCl concentration effect in TBARS and peroxide values are
either less apparent or not signiﬁcant between cooked treatments when compared to the
raw treatments (Tables 2-1 and 2-2). This phenomenon is consistent with reports by other
researchers that NaCl or other salts promote lipid oxidation in raw samples but may or
may not in cooked samples. For example, Torres et al. (1988) reported no differences in
lipid oxidation in cooked post-rigor beef treated with different NaCl concentrations (0 - 4
%) but documented differences in the raw samples. Ahn et al. (1993b) reported there
were no differences in lipid oxidation in cooked turkey containing 0 and 2 % NaCl.
Several explanations for the differences in conclusions by the investigators regarding the
effect of NaCl on lipid oxidation in cooked meat are possible. First, lipid oxidation is a
free radical reaction and the initiation reaction is the rate-determining step. Once the
reaction is initiated, the importance of the presence of the prooxidant may be reduced.
Cooking can provide the energy to initiate lipid oxidation in meat. Although NaCl has a

prooxidant effect by itself, the NaCl effect is confounded with cooking effects and the

71

Table 2-2. The effect of NaCl concentrations TBARS and peroxide values in cookedl
ground pork during refrigerated storage (4 °C)

 

TBARS2 (mg malonaldehyde/kg meat)

 

 

Treatment Day 03" Day 1 Day 2
1. pork (control) 1.7“ 3.8“ 5.2“
2. pork (8.8 mg NaCl/g meat) 2.3“ 4.4“'3 8.0“
3. pork (17.4 mg NaCl/g meat) 3.2“ 5.8”“ 7.8“
4. pork (28.1 mg NaCl/g meat) 3.3cA 8.3cl3 7.9“

Peroxide values2
(milliequivalents of peroxide values/kg sample)

 

 

Day 0 Day 1 Day 2
1. pork (control) 2.8"A 6.3‘9 9.8“C
2. pork (8.8 mg NaCl/g meat) 3.8“ 8.9‘“ 9.8“
3. pork (17.4 mg NaCl/g meat) 4.8”“ 7.7”“ 9.9“
4. pork (28.1 mg NaCl/g meat) 5.1“ 8.1“ 11.1“

 

l. The cooked samples were heated to an internal temperature of 70 °C in a 83 °C water
bath.
. TBARS mean square error = 0.49, peroxide values mean square error = 1.68.
. Means with different superscript letters within the same column are signiﬁcantly
different (p<0.05).
4. TBARS or peroxide values in each row with the same capital letter superscript are not
signiﬁcantly different (p<0.05).

MN

72
NaCl effect on lipid oxidation is less dramatic for cooked samples (Tables 2-1 and 2-2).

Second, it is possible that heating and NaCl may have similar mechanisms for promoting
lipid oxidation. For example, both NaCl and heating break down the meat microstructure
(Ofstad et al., 1995; Velinov et al., 1990) and release iron ions (Kanner et al., 1991;
Osinchak et al., 1992; Love and Pearson, 1974). Therefore, the prooxidant effect of NaCl
will be less signiﬁcant after heating because the microstmcture has already been broken
down with or without NaCl addition. The exception may be when a high concentration of
NaCl is used. A third possibility is attributed to the experimental error due to differences
in the meat systems or species examined, complexity of the model system, and
concentrations of salt used (Srinivasan and Xiong, 1996).

For example, Wettasinghe and Shahidi (1996) reported that lean pork treated with NaCl at
a concentration of 100 meq/kg sample exhibited a prooxidative effect after day 3 of
refrigerated storage. However, lean pork treated with NaCl at day 1 had signiﬁcantly
(p<0.05) lower lipid oxidation than the control (antioxidant effect). This observation is in
direct conﬂict with the hypothesis that NaCl is a prooxidant. This observation may be due

to experimental error.

Effect of NaCl on non-heme iron concentration

There were no differences in total iron concentrations within raw or cooked
samples (Tables 2-3 and 2-4). This indicates that salt or other materials used in this study
do not contribute to the total iron concentrations in the meat systems. In raw, control
samples (no salt) , non-heme iron concentration did not increase during 6 days of

refrigerated storage (Table 2-3). In raw samples containing salt, non-heme iron

73
concentrations signiﬁcantly increased during the reﬁigerated storage period from day 3 to

6. Decker and Hultin (1990) reported that low-molecular weight fraction (<10 kilodalton)
iron increased signiﬁcantly in mackerel muscle during 7 days storage at 0 °C. This
increase may be attributed to the presence of ascorbate and the superoxide radicals which
were reported to release iron from ferritin (Biemond et al., 1986; Boyer and McCleary,
1987) and hydrogen peroxide which may cause release of heme iron (Rhee et al., 1987)
during storage. Sodium chloride seems to have accelerated this process. Therefore, non-
heme iron concentrations in the NaCl-added treatments increase signiﬁcantly during
storage (Table 2-3). There were no differences in the non-heme iron concentrations
between treatments with or without NaCl on days 0 and 3. This is consistent with data
reported by Ahn et al (1993 c) which indicated that NaCl does not affect the non-heme
iron concentrations in raw turkey meat. On day 6, the NaCl treatments had higher non-
heme iron concentrations than the control. However, only treatments with 17.4 and 26.1
mg NaCl/g meat had non-heme iron concentrations that were signiﬁcantly (p<0.05)
greater than the control.

In the cooked samples (Table 2-4), 17.4 and 26.1 mg NaCl/g meat treatments
signiﬁcantly (p < 0.05) increased non-heme iron concentrations in ground pork at day 0.
This observation is also in agreement with Ahn et al. (1993c) who found that NaCl
increased the amounts of non-heme iron in cooked turkey leg meat and mechanically
deboned turkey meat. However, because the total iron concentration in the cooked
samples also increased, part of the iron increase in cooked samples may be a concentration

effect due to the water or lipid loss during cooking.

74

Table 2-3. Total and non-heme iron concentration of raw ground pork during reﬁigerated

 

 

 

storage (4 °C)

Non—heme ironl Total iron
(ug Fe/g meat) (ug Fe/g meat)

Treatment Day 02‘3 Day 3 Day 6

1. pork (control) 5.0“ 5.3“ 5.7“ 11.2'

2. pork (8.8 mg NaCl/g meat) 4.8“ 5.3“ 72'“ 10.9‘

3. pork (17.4 mg NaCl/g meat) 5.2“ 5.8“ 7.8“ 11.9'

4. pork (28.1 mg NaCl/g meat) 5.9“ 5.9“ 8.3“ 11.4'

 

H

. Non-heme iron mean square error = 0.65.

. Means with different superscript letters within the same column are signiﬁcantly
different (p<0.05).

3. Non-heme iron concentration in each row with the same capital letter superscript are

not signiﬁcantly different (p<0.05).

N

75

Table 2-4. Total iron and non-heme iron concentration of cooked1 ground pork during

 

 

 

reﬁigerated storage (4 °C)
Non-heme iron (ug F e/ g meat)2 Total iron
(ug Fe/g meat)

Treatment Day 03" Day 1 Day 2

1. pork (control) 5.3“ 8.8“ 5.8“ 13.8'
2. pork (8.8 mg NaCl/g meat) 59'“ 7.1“ 8.4'“ 14.7'
3. pork (17.4 mg NaCl/g meat) 7.7“ 8.7“ 7.8“ 13.8'
4. pork (28.1 mg NaCl/g meat) 81“ 7.1“ 7.1'“ 14.0'

 

1. The cooked samples were heated to an internal temperature of 70 °C in a 83 °C water

bath.

Non-heme iron mean square error = 1.90.

Means with different superscript letters within the same column are signiﬁcantly

different (p<0.05).

4. Non-heme iron concentration in each row with the same capital letter superscript are
not signiﬁcantly different (p<0.05).

5"!"

76

The relationship between lipid oxidation, non-heme iron concentration and NaCl
concentration

The data in Tables 2-1 and 2-2 indicate that increasing NaCl concentration will
increase TBARS and peroxide values in raw and cooked samples. As NaCl
concentrations increased in the cooked samples at day 0, the non-heme iron concentration
increased and measures of lipid oxidation increased proportionately. Greater NaCl
concentrations resulted in greater lipid oxidation and non-heme iron concentrations at day
0. In the raw pork study, increasing NaCl concentrations produced signiﬁcantly higher
TBARS and peroxide values than the control after 3 days of refrigerated storage. NaCl
affected non-heme iron concentrations in samples stored for 6 days at 4 °C. The response
of non-heme iron release to increases in NaCl concentrations in raw pork samples was
slower when compared to the response of lipid oxidation to the NaCl concentration effect.
This may be because most of the loosely bound non-heme irons were removed by NaCl at
day 0. It took longer time for NaCl to release the tightly bound non-heme iron from the
meat. Therefore, the effect of NaCl on non-heme iron release was only apparent at day 6.
Since there was good agreement between lipid oxidation and non-heme iron
concentrations in cooked samples at day 0 and in raw samples at day 6 as discussed above,
the hypothesis that the prooxidant mechanism of NaCl is to increase the availability of iron

to catalyze lipid oxidation in meat is valid.

77

The effect of' various salts on lipid oxidation and non-heme iron concentration in
pork

In the raw pork study, there were no diﬂ‘erences in lipid oxidation and non-heme
iron concentrations between various salts treatments at day 0 (Tables 2-5, 2-6). After day
3, lipid oxidation in treatments with the various salts (NaCl, KCl, NaBr and KBr) was
signiﬁcantly (p<0.05) higher than the control when monitored by TBARS. However, only
NaCl and NaBr treatments were signiﬁcantly (p<0.05) higher than the control when
monitored by the peroxide values. The deviation between TBARS and peroxide values
may be because the variation for peroxide values methodology was greater than TBARS
methodology in this study (TBARS mean square error = 0.865 and peroxide values =
2.682). For the corresponding non-heme iron concentrations at day 3, only the 8.8 mg
NaCl/g meat treatment was signiﬁcantly diﬂ‘erent from the control (Table 2-6). The other
three salt treatments had higher, but not signiﬁcant, non-heme iron concentrations than the
control. This may be due to the sensitivity of the non-heme iron measurement method as
discussed above. There were no signiﬁcant differences in TBARS and non-heme iron
concentrations between various salt treatments. At day 6, the various salt treatments had
higher TBARS and peroxide values than the control except the KBr treatment when
monitored by peroxide values (Table 2-5). The non-heme iron concentrations also reﬂect
a similar trend as lipid oxidation except for 8.8 mg KCl/g meat (Table 2-6). There were
no signiﬁcant differences in non-heme iron concentration and lipid oxidation (TBARS and
peroxide values) among various salt treatments. In general, these data demonstrated that

salts (KCl, NaBr and KBr) other than the NaCl can also promote lipid

78

Table 2-5. The effect of different salts on TBARS and peroxide values in raw
ground pork during refrigerated storage (4 °C)

 

TBARSl (mg malonaldehyde/kg meat)

 

 

 

 

Treatment Day 02’3 Day 3 Day 6
1. control 0.3“ 0.3“ 0.3“
2. control (8.8 mg NaCl/g meat) 0.4'A 1.9[,8 3.5"6
3. control (11.2 mg KCl/g meat) 0.4'LA 1.8” 3.4”
4. control (15.5 mg NaBr/g meat) 0.5"A 1.3IDA 3.1be
5. control (17.9 mg KBr/g meat) 0.9'A 1.2M 2.4M3

Peroxide valuesl (milliequivalents of

peroxide values/kg sample)

Day 0 Day 3 Day 6

1. control 2.0“ 2.3“ 3.2“
2. control (8.8 mg NaCl/g meat) 2.8“ 7.3“ 8.9“
3. control (11.2 mg KCl/g meat) 2.2"A 5.9"ch 8.8"c
4. control (15.5 mg KCl/g meat) 1.8"A 6.4mI3 7.2”'3
5. control (17.9 mg KCl/g meat) 1.8"A 3.2M 5.7'IbB

 

1. TBARS mean square error = 0.87, peroxide values mean square = 2.68.

2. Means with different superscript letters within the same column are signiﬁcantly
different (p<0.05).

3. TBARS or peroxide values in each row with the same capital letter superscript are not
signiﬁcantly different (p<0.05).

79

Table 2-6. Total and non-heme iron concentration in raw ground pork during reﬁigerated

 

 

 

 

storage (4 °C)
Non-heme ironl (ug Fe/g meat) Total iron
(ug Felg meat)
Treatment Day 02'3 Day 3 Day 6
1. control 3.8“ 4.1“ 3.9“ 11.8‘
2. control (8.8 mg NaCl/g meat) 3.9“ 5.4“ 6.2“ 11.6‘
3. control (11.2 mg KCl/g meat) 4.7"A 4.9“ 5.2'b 11.3'
4. control (15.5 mg KCl/g meat) 4.0"A 5.2“8 6.6be 11.2'
5. control (17.9 mg KCl/g meat) 4.3“ 5.2““ 8.0“ 11.9'
1. Non-heme iron mean square error = 1.07.
2. Means with different superscript letters within the same column are signiﬁcantly
different (p<0.05).

3. Non-heme iron concentration in each row with the same capital letter superscript are

not signiﬁcantly different (p<0.05).

80

oxidation and increase the non-heme iron concentrations in the meat systems. However,
the diﬂ‘erences in prooxidant activity and non-heme iron concentration were not signiﬁcant
at the same molarity between these various salts. These data are in agreement with
Srinivasan and Xiong (1996) who reported that no signiﬁcant diﬂ‘erences in lipid oxidation
(p > 0.05) were observed between various salts (NaCl, NaBr, Nastr, KCl and LiCl) in
salted beef heart surimi during reﬁigerated storage at low ionic strength (0.1 M).
However, the results are different from Rhee et al. (1983) who reported that NaCl
promoted more extensive lipid oxidation than KCI in meat at the same molarity.
Wettasinghe and Shahidi (1996) also indicated that lean pork treated with NaCl, NaBr,
KCl and KBr at concentrations of 100 and 200 meq/kg sample had diﬂ‘erent degrees of
lipid oxidation during reﬁigerated storage. The prooxidant activity order between these
salts (NaCl, NaBr, KCl and KBr) are not always constant during the storage. The
conﬂicting observations on the role of various salts in lipid oxidation may be due to
differences in the meat systems or species examined, complexity of the model system, and
concentrations of salt used (Srinivasan and Xiong, 1996). For example, Srinivasan and
Xiong (1996) reported when various salts were added at a low concentration (0.1 M) to
buffer-washed surimi, they exhibited only a minimal effect on lipid oxidation. When added
at a much higher concentration (0.6 M), these salts stimulated lipid oxidation to various
extents. When adding salt on a weight percentage basis rather than on a molar basis, King
and Bosch (1990) reported that 2 % NaCl in cooked ground turkey was more oxidized
than turkey containing 2 % KCl. These differences may be attributed to the fact that the
same percentage weight of the various salts will result in different molarities (NaCl > KCl

> NaBr > KBr) because each salt has a different molecular weight.

81
For cooked meat samples, various salt treatments had higher TBARS and peroxide

values than the control at day 0 (Table 2-7). After 1 day of refrigerated storage, the
differences in lipid oxidation between control and various salt treatments were not
signiﬁcant. However, all TBARS values were above 3.5 and all peroxide values were
above 8.0. The salt effect was less signiﬁcant in the cooked samples compared to the raw
samples as was discussed previously. For the corresponding non-heme iron data at day 0,
treatments with various salts had a trend of higher non-heme iron concentrations than the
control (Table 2-8). However, only the KBr treatment was signiﬁcant (p<0.05). After 2
days of reﬁigerated storage, NaCl, KC] and KBr treatments had signiﬁcantly higher non-
heme iron concentrations than control.

The data demonstrating the effect of various salts on lipid oxidation and non-heme
iron concentration in pork are similar to the NaCl data presented above (Tables 2-1, 2-2,
2-3 and 2-4). Various salts will increase lipid oxidation in both raw and cooked samples
when monitored by TBARS and peroxide values. Various salts also signiﬁcantly (p<0.05)
increased the non-heme iron concentrations in raw pork at 6 days of storage but only
immediately after addition in the cooked samples. Because of the similar response of these
various salts (NaCl, KCl, NaBr and KBr) on lipid oxidation and non-heme iron
concentration in the meat systems, it is possible that these salts share the same prooxidant
mechanism as NaCl. The mechanism proposed is that the prooxidant effect of salt is

related to non-heme iron release.

82

Table 2-7. The eﬂ‘ect of different salts on TBARS and peroxide values in cookedl

ground pork during reﬁigerated storage (4 °C)

 

TBARS (mg malonaldehyde/kg meat)2

 

 

Treatment Day 03'4 Day 1 Day 2
1. control 1.9“ 3.6as 4.6'c
2. control (8.8 mg NaCl/g meat) 2.8“ 4.7“ 5.3“
3. control (11.2 mg KCl/g meat) 2.7“ 4.3'8 4.5“
4. control (15.5 mg KCl/g meat) 3.6“ 5.1“ 5.0'8
5. control (17 .9 mg KCl/g meat) 3.4“ 5.1'8 5.3'8

Peroxide values2 (milliequivalents

of peroxide values/kg sample)

 

 

 

Day 0 Day 1 Day 2

1. control 4.1“ 8.2'8 10.2'0

2. control (8.8 mg NaCl/g meat) 5.6“ 9.6''3 10.5'.8

3. control (11.2 mg KCl/g meat) 5.3“ 9.0'8 10.7'6

4. control (15.5 mg KCl/g meat) 6.1“ 8.9'Ila 10.8ac

5. control (17.9 mg KCl/g meat) 5.3“ 9.8“ 10.3“

1. The cooked samples were heated to an internal temperature of 70 °C in a 83 °C water
bath.

2. TBARS mean square error = 0.82, peroxide values mean square error = 0.95.

3. Means with different superscript letters within the same column are signiﬁcantly
different (p<0.05).

4. TBARS or peroxide values in each row with the same capital letter superscript are not

signiﬁcantly different (p<0.05).

83

Table 2-8. Total iron and Non-heme iron concentration in cookedl ground pork during
reﬁigerated storage (4 °C)

 

 

 

Non-heme iron2 (ug Fe/g meat) Total iron
(ug Felg meat)

Treatment Day 03" Day 1 Day 2

1. control 5.0“ 8.4“ 5.8“ 12.6'
2. control (8.8 mg NaCl/g meat) 5.6“A 6.7“ 6.9“ 12.6‘
3. control (11.2 mg KCl/g meat) 6.4"“ 6.0“ 6.6“ 13.3'
4. control (15.5 mg KCl/g meat) 6.1 “A 6.7“ 6.4"“ 12.6'
5. control (17.9 mg KCl/g meat) 7.1“ 6.5“ 6.7“ 12.1‘I

 

p—I

The cooked samples were heated to an internal temperature of 70 °C in a 83 °C water
bath.

2. Non-heme iron mean square error = 1.89.

Means with different superscript letters within the same column are signiﬁcantly
different (p<0.05).

4. Non-heme iron concentration in each row with the same capital letter superscript are
not signiﬁcantly different Q)<0.05).

1.»

84

Summary and Conclusions
The addition of NaCl increased lipid oxidation in both raw and cooked samples as
well as non-heme iron concentrations. The corresponding non-heme iron analyses
indicated that non-heme iron concentration increased signiﬁcantly immediately for cooked
samples and after 6 days of storage for raw samples. Salts (KCl, NaBr and KBr), other
than NaCl, also had similar effects as NaCl with respect to lipid oxidation and non-heme
iron concentration. Based upon data presented, it is possible that the prooxidant effect of

salt is to make more iron available to catalyze lipid oxidation.

85

References

Ahn, D.U., Wolfe, F.H. and Sim, J .S. 1993a. Prevention of lipid oxidation in pre-cooked
turkey meat patties with hot packaging and antioxidant combinations. J. Food Sci.
58:283.

Ahn, D.U., Ajuyah, A., Wolfe, F.H. and Sim, J .S. 1993b. Oxygen availability affects
prooxidant catalyzed lipid oxidation of cooked turkey patties. J. Food Sci. 58:278.

Ahn, D.U., Wolfe, F .H. and Sim, J .S. 1993c. Three methods for determining nonheme
iron in turkey meat. J. Food Sci. 58:288.

Akamittath, J .G., Brekke, OJ. and Schanus, E.G. 1990. Lipid oxidation and color
stability in restructured meat systems during frozen storage. J. Food Sci. 55: 1513.

Apte, S. and Morrissey, PA. 1987. Eﬁ‘ect of haemoglobin and ferritin on lipid oxidation in
raw and cooked muscle systems. Food Chem. 25:127.

Arnold, A.R., Bascetta, E. and Gunstone, FD. 1991. Effect of sodium chloride on
prooxidant activity of copper (II) in peroxidation of phospholipid liposomes. J.
Food Sci. 56:571.

Biemond, P., Swaak, J .G., Beudorff, CM. and Koster, J .F. 1986. Superoxide-dependent
and -independent mechanisms of iron mobilization from ferritin by xanthine
oxidase. Biochem. J. 239: 169.

Boyer, RF. and McCleary, CJ. 1987 . Superoxide ion as a primary reductant in ascorbate-
mediated fenitin iron release. Free Radical Biology & Medicine. 3:389.

Chang, I. and Watts, BM. 1949. Antioxidants in the hemoglobin catalyzed oxidation of
unsaturated fats. Food Technol. 3:332.

Chang, I. and Watts, BM. 1950. Some effects of salt and moisture on rancidity of fat.
Food Res. 15:313.

Crackel, R.L., Gray, J 1., Pearson, A.M., Booren, A.M. and Buckley, DJ. 1988. Some
further observations on the TBARS test as an index of lipid oxidation in meats.
Food Chem. 28:138.

Decker, EA. and Hultin, H0. 1990. Factors inﬂuencing catalysis of lipid oxidation by
the soluble fraction of mackerel muscle. J. Food Sci. 55:947.

Ellis, R., Currie, G.T., Thornton, F.E., Bollinger, NC. and Gaddis, A.M. 1968. Carbonyls
in oxidizing fat. 11. The effect of the pro-oxidant activity of sodium chloride on
pork tissue. J. Food Sci. 33:555.

86

Gill, J.L. 1978. “Design and Analysis of Experiments in the Animal and Medical
Sciences.” Iowa State Univ. Press, Ames, IA.

Igene, J .0, King, J .A., Pearson, A.M. and Gray, 11. 1979. Inﬂuence of Heme pigments,
nitrite, and non-heme iron on development of warmed-over ﬂavor (W OF) in
cooked meat. J. Agric. Food Chem. 27:838.

Kanner, J ., Harel, S. and Jaffe. R. 1991. Lipid peroxidation of Muscle food as affected by
NaCl. J. Agic. Food Chem. 39:101.

King, A]. and Bosch, N. 1990. Eﬁ‘ect of NaCl and KC] on rancidity of dark turkey meat
heated by microwave. J. Food Sci. 55: 1549.

Kuo, J .C. and Ockerman, H.W. 1985. Effects of rigor state, salt level and storage time on
chemical and sensory traits of ﬁozen and freeze-dried ground beef. J. Food Prot.
48:142.

Love, JD. and Pearson, A.M. 1974. Metmyoglobin and non-heme iron as prooxidants in
cooked meat. J. Agric. Food Chem. 22: 1032.

Ofstad, R., Kidman, 8., Myklebust, R., Olsen, KL. and Herrnansson, A.M. 1995. Liquid-
holding capacity and structural changes in comminuted salmon (Salmo Salar)
muscle as inﬂuenced by pH, salt and temperature. Lebens. Wissen. Technol.
28:329.

Osinchak, J .E., Hultin, H.O., Zajicek, O.T., Kelleher, SD. and Huang, CH. 1992. Effect
of NaCl on catalysis of lipid oxidation by the soluble fraction of ﬁsh muscle. Free
Radical Biol. Med. 12:35.

Rhee , K.S., Smith, 6C. and Terrell, RN. 1983. Effect of reduction and replacement of
sodium chloride on rancidity development in raw and cooked ground pork. J. Food
Prot. 46:578.

Rhee, KS. and Ziprin, Y.A. 1987. Modiﬁcation of the Schricker nonheme iron method to
minimize pigment effects for red meats. J. Food Sci. 52: 1 174.

Rhee, K.S., Zibru, Y.A. and Ordonez, G. 1987 . Catalysis of lipid oxidation in raw and
cooked beef by metmyoglobin-hydrogen peroxide, nonheme iron and enzyme
systems. J. Agric. Food Chem. 35:1013.

Shantha, NC. and Decker, EA. 1994. Rapid, sensitive, iron-base Spectrophotometric
methods for determination of peroxide values of food lipids. J. AOAC Int. 77:421.

87

Srinivasan, S. and Xiong, Y.L. 1996. Sodium chloride-mediated lipid oxidation in beef
heart surimi-like material. J. Agric. Food Chem. 44: 1697.

Tarladgis, B.G., Watts, RM. and Younathan, MT. 1960. A distillation method for the
quantitative determination of malonaldehyde in rancid foods. J. Am. Oil Chem.
Soc. 37:44.

Torres, E., Pearson, A.M., Gray, J.I., Booren, A.M. and Shimokomaki, M. 1988. Eﬂ‘ect
of salt on oxidative changes in pre- and post-rigor ground beef. Meat Sci. 23: 151.

Velinov. P.D., Zhikov, M.V. and Cassens, R.G. 1990. The effect of tumbling, sodium
chloride and polyphosphates on the microstructure and appearance of whole-
muscle processed meats. Food Structure. 9:91.

Watts, BM. and Peng, D. 1947. Lipoxidase activity of hog hemoglobin and muscle
extract. J. Biol. Chem. 1792441.

Wettasinghe, M. and Shahidi, F. 1996. Oxidative stability of cooked comminuted lean
pork as affected by alkali and alkali-earth halides. J. Food Sci. 61: 1 160.

CHAPTER III

THE EFFECTS OF CHLORIDE AND BROMIDE SALTS ON N ON-HEME IRON
CONCENTRATION AND LIPID OXIDATION IN A MODEL SYSTEM
CONTAINING VARIOUS FORMS OF IRON

Abstract
The effects of NaCl, KCl, NaBr and KBr salts on non—heme iron release from hemoglobin
and myoglobin and on recovery of added iron were studied in raw and cooked water-washed
muscle ﬁber model systems. Lipid oxidation was monitored by the thiobarbituric acid-reactive
substances (TBARS) method, and iron was determined using atomic absorption spectroscopy. The
addition of various chloride and bromide salts to the model system containing heme-containing
proteins increased TBARS values (1.58 for myoglobin alone, 4.02 for myoglobin + NaCl after 6
days). However, the addition of these salts had no effect on the release of non-heme iron from
hemoglobin and myoglobin in either the raw or cooked samples (7.0 ug F e/g meat for myoglobin
alone, 7.2 ug Fe/g meat for myoglobin + NaCl after 2 days). The addition of various salts to the
cooked model system containing 3 parts per million added iron produced greater TBARS values
compared to the model systems containing added iron alone. They also did not signiﬁcantly
increase the non-heme iron concentration. Because sodium chloride promoted lipid oxidation
without increasing the non-heme iron fraction in the model system. These data suggest that salt

does not increase lipid oxidation by releasing iron from heme proteins.

88

89
Introduction

Sodium chloride has been reported to promote lipid oxidation in diﬂ‘erent species
ofmeat (King and Bosch, 1990; Arnold et al., 1991; Kanner et al., 1991; Ahn et al. 1993a,
b). For example, lipid oxidation in raw pork patties during frozen storage was rapidly
accelerated by 1% salt concentration (Andersen and Skibsted, 1991). Raw turkey meat
patties with 2% NaCl had greater TBARS values than turkey meat patties without NaCl
(Ahn et al., 1993b). Recently, it has been suggested that the prooxidant mechanism of
NaCl is to increase the free iron concentration by either breaking iron-protein
macromolecular interactions (Kanner et al.,l991) or by releasing iron from high molecular
weight (>10 kilodalton)proteins (Osinchak et al., 1992; Ahn et al., 1993c).

Kanner et al. (1991) suggested that when free iron ions were added to minced
turkey dark muscle, a large part of the added iron interacted with protein macromolecules.
This interaction prevented iron ions from contacting with membranous lipids and acting as
a catalyst of lipid oxidation. Addition of NaCl interrupted the interaction between the iron
and macromolecules. Therefore, more free iron ions were available to interact with the
lipid fraction, thereby enhancing lipid oxidation. Osinchak et al. (1992) used a different
model system (phosphatidylcholine liposomes) but a similar approach to study the
prooxidant effect of NaCl. They conﬁrmed that iron ions were involved in NaCl-mediated
lipid oxidation. They also observed that using NaCl and a high molecular weight (>10
kilodalton) fraction of mackerel press juice as a prooxidant in the model system increased
the amount of lipid oxidation 7-8 times compared to using the high molecular weight (>10
kilodalton) ﬁ'action of mackerel press juice alone. They hypothesized that iron was

released from iron-containing protein in the high molecular weight fraction (>10

90

kilodalton) of mackerel muscle press juice when NaCl was present. The free iron
subsequently catalyzed lipid oxidation in this model system.

Iron in the high molecular weight iron—containing proteins such as hemoglobin,
myoglobin and ferritin account for 51 to 74% of the total iron in muscle depending on the
type of muscle (Hazell, 1982). It has been reported that these proteins will release iron
during cooking (Love and Pearson, 1974; Oellingrath, 1988; Buchowski et al., 1988) or
during storage (Decker and Hultin, 1990). Cooking denatures the heme proteins,
degrades the heme, and subsequently releases iron (Oellingrath, 1988). Salt may have a
similar effect as heat in denaturing the heme-containing proteins (Kijowski and Mast,
1988; Ahn and Maurer, 1989). Salt can alter the stability of meat proteins, causing a
reduction in the denaturation temperature (Quinn et al. 1980; Kijowski and Mast, 1988).
Ahn and Maurer (1989) reported that NaCl will decrease the heat stability of hemoglobin
and myoglobin.

Although Kanner et al. (1991) and Osinchak et al.(1992) suggested that the
prooxidant effect of NaCl is to increase the availability of iron ions by either breaking the
iron-protein macromolecule or releasing iron ﬁom the high molecular iron-containing
proteins, there are no data available which veriﬁes the source of ﬁee iron in a meat system
when NaCl is present. The objective of this study is to verify the free iron source by
examining the effects of various salts on iron release from hemoglobin and myoglobin as
well as to study the recovery of added iron in a meat system. Speciﬁc objectives are: 1.
To determine the effect of chloride and bromide salts on iron release from heme proteins
in a water-washed muscle ﬁber model system. 2. To determine the effect of these salts on

iron recovery in a water-washed muscle ﬁber model system. 3. To compare the effect of

91

chloride and bromide salts on non-heme iron release ﬁ'om heme proteins as well as the

recovery of the added iron with lipid oxidation.

92
Materials and Methods

Pork legs were obtained ﬁ'om a local packer within 48 hr after slaughter. Bone
and skin were removed ﬁom the legs the following day after arriving at Michigan State
University. The boneless pork was stored at -10 °C before use. Bathophenanthroline
disulfonic acid (sodium salt), hemoglobin, myoglobin, thioglycolic acid (96-99%), NaCl,
KCl, NaBr and KBr were purchased ﬁ'om Sigma Chemical Co. (St. Louis, MO).
Trichloroacetic acid, ammonium acetate, nitric acid and perchloric acid were obtained
ﬁ'om J.T. Baker Inc. (Phillipsburg, NJ). Sodium nitrite and hydrochloric acid were from
purchased Mallinckrodt Inc. (Paris, KY). All chemicals were reagent grade. All the
glassware used in this study was immersed overnight in a 4 N hydrochloric acid solution to
remove potential trace iron and rinsed with distilled water to remove hydrochloric acid
residues. Disposable plastic test tubes (Baxter Diagnostics Inc., McGraw Park, IL) were

used instead of glass test tubes to prevent possible iron contamination.

Sample preparation
Experiment I

Water-washed muscle ﬁbers (WF) were prepared according to the method of
Hazell (1982)(Figure 3-1). One hundred twenty g WF, 12 ml 1.65M salt solution (NaCl,
KCl, NaBr or KBr) and/or heme proteins (5 mg hemoglobin or myoglobin/g wet ﬁber)
were mixed in a 250 ml beaker for 1.5 min then held refrigerated for 30 min. In order to
more easily study the effect of NaCl on non-heme iron release from heme proteins, higher
heme proteins than naturally exists in pork were used in this study. The target iron

concentration in treatments containing heme proteins was 16.5 ug Fe/g meat (Hazell,

93
1982). The target salt concentration in each treatment that contained salt was 8.8 mg

NaCl/g meat, 11.2 mg KCl/g meat, 15.5 mg NaBr/g meat and 17.9 mg KBr/g meat (equal
to 0.15 M salt in meat). For the control, distilled water instead of a salt solution was
added to the WP and handled in a similar fashion. Iron release and lipid oxidation were
studied in both raw and cooked systems. The samples were cooked by placing the
samples in a 250 ml beaker, covering with aluminum foil and heating to an internal
temperature of 70 °C in a 83 °C water bath. Cooking time was approximately 50 min. All
samples were stored at 4°C during the storage study.

Lipid oxidation was monitored by the 2-thiobarbituric acid procedure (Tarladgis et
al., 1960) as modiﬁed by Crackel et al. (1988). Total iron in each sample was determined
using the extraction method of Igene et al. (1979), which utilizes nitric acid and perchloric
acid digestion followed by quantitation by atomic absorption spectroscopy. The non-heme
iron fraction of each sample was prepared using the modiﬁed Schricker method (Rhee and
Ziprin,1987) and quantitated using atomic absorption spectroscopy.

Experiment 11

Hemoglobin or myoglobin solutions were prepared to target 16.5 ug F e/g meat
(approximately 5 mg hemoglobin or myoglobin per ml water). Different salts (NaCl, KCI,
NaBr and KBr) were added to the heme protein solutions to reach a 17.4 mg salt/g meat.
Samples (15 ml) were cooked in a 88-90 °C water bath to an internal temperature of 70
°C. For the hemoglobin stability test, 15 ml of the hemoglobin solution were cooked in a
88-90 °C water bath to internal temperatures of 75, 85 and 87.5 °C (Chen et al., 1984).
The non-heme iron concentrations in the cooked heme proteins solution were quantitated

by the method of Igene et al. (1979).

94

Experiment 111

Three ug Fe/g meat was incorporated into the WP by mixing 120 g of WF with a
12 ml 33 ug Felg meat ferrous sulfate solution in a 250 ml beaker for 1.5 min. The
samples were reﬁigerated for 30 min and cooked as described previously. Lipid oxidation
was monitored by the TBARS method (Tarladgis et al., 1960) as modiﬁed by Crackel et
al. (1988). Non-heme iron was extracted according to Igene et al. (1979) and quantitated

using atomic absorption spectroscopy.

Data Analysis

The experiment was performed using a split-plot design with repeated measures
(Gill, 197 8). Means, standard errors, sum of squares and mean square errors were
calculated using the MSTAT-C microcomputer statistical program (Michigan State

University, 1989). The Bonferroni t test was used to compare selected contrasts.

95

ground pork: water = 1:4

blending (blender) a

 

centrifuge at 3,000xG + water

 

 

l l

supernatant precipitate —

l

water-washed muscle fibers

 

Figure 3-1 Water-washed muscle fibers preparation procedure

96

Results and Discussion

The effect of salt on non-heme iron concentrations in the WF system containing
heme proteins

Water-washed muscle ﬁbers systems were used to study the eﬁ‘ect of salt on non-
heme iron release ﬁom heme protein. This model system most closest resembles the meat
system when comparing with other model systems such as liposome systems, emulsions of
methyl linoleate and linoleic acid systems. Total and non-heme iron concentrations for the
raw and cooked WF model systems in the presence of various salts and heme proteins are
presented in Tables 3-1 and 2-2, respectively. Hemoglobin and myoglobin treatments had
an average of 16.2 ug Fe/g meat more total iron than the WP (control). This is close to
the 16.5 ug F e/ g meat which was targeted. Addition of hemoglobin and myoglobin
increased (p<0.05) non-heme iron in raw samples at day 0 (Table 3-1). This observation
is in agreement with Ahn et al. ( 1993c) who reported that the addition of hemoglobin to
turkey breast meat increased the ﬁee iron in that system. They attributed the iron increase
to the damaged heme proteins and which may
have released iron during the sample preparation procedure (Chen et al., 1984; Ahn et al.,
1993 c). There were no differences (p<0.05) in non-heme iron concentrations present in
either the myoglobin or hemoglobin treatments in this study (Table 3-1). Neither salt
addition nor different salts had any effect on the non-heme iron concentration in
treatments containing either heme protein at day 0. This is consistent with data reported

by Ahn et al (1993c) which indicated that NaCl does not affect the non-heme iron

97

Table3-1. Total and non-heme iron in a raw water-washed muscle ﬁber (WF) model
system in the presence of selected heme proteins1 or salts2

 

 

 

Non-heme iron (ug Fe/g meat) Total iron
Treatment Day 03’” Day 3 Day 6
1. WF (control) 4.5‘I 4.9' 4.0‘ 5.9‘
2. WF + Mb (myoglobin) 6.5" 7.2" 5.1" 21.4"
3. WF + Mb + NaCl 6.2" 7.7" 8.7" 21.0"
4. WF + Mb + KCl 7.1" 7.7" 6.7" 20.9"
5. WF + Mb + NaBr 5.9" 6.9" 5.9" 20.7"
8. WF + Mb + KBr 8.4" 7.8" 8.4" 21.5"
7. WF + Hb (hemoglobin) 8.4" 7.9" 5.7" 22.4"
8. WF + Hb + NaCl 8.4" 7.3" 8.8" 22.5"
9. WF + Hb + RC] 7.0" 7.9" 6.1" 21.4"
10. WF + Hb + NaBr 5.9" 7.8" 8.5" 22.5"
11. WF+Hb+KBr 8.6" 7.2" 5.8" 21.9"

 

H

Hemoglobin and myoglobin solutions were prepared to target 16.5 ug Fe/g meat or 5
mg myoglobin or hemoglobin/g meat.

NaCl, KCI, NaBr and KBr concentrations were 8.8 mg salt/ g meat.

Samples were held at 4 °C.

Mean square error = 0.98.

Mean values within the same column with different superscripts are different

(p < 0.05).

.V‘PP’!"

98
concentrations in raw turkey meat. After 6 days of reﬁigerated storage at 4 °C, salt did

heme iron release from hemoglobin treatments was greater (p<0.05) than from myoglobin
treatments in the cooked samples. This may be because hemoglobin is more heat sensitive
than myoglobin (Lewis, 1926; Oellingrath, 1988). Oellingrath (1988) reported that
hemoglobin-heme was more heat sensitive than myoglobin-heme between 78 °C and 100
0C. Therefore, more non-heme iron was released ﬁom hemoglobin than myoglobin in
cooking. Neither salt addition nor diﬂ‘erent salts signiﬁcantly increased the non-heme iron
concentrations in the cooked treatments containing either hemoglobin or myoglobin
during 2 days of reﬁigerated storage at 4 °C (Table 3-2). Ahn et al. (1993c) reported that
with 2% NaCl, the amounts of non-heme iron in cooked turkey leg meat and mechanically
deboned turkey meat were 3.5-4.0 ug/g greater than those without NaCl. They attributed
the increase in non-heme iron to NaCl addition which destabilized the heme pigment and
therefore released iron. The differences between the current study and the one performed
by Ahn et al. (1993c) may be because a different meat system or cooking temperature
were used. In the Ahn et al. (1993c) study, the experiment was conducted with turkey
meat and samples cooked in a 350 °C oven to an internal temperature of 80 °C. In this
study, WF from pork was used and samples were cooked in a 83 0C water bath to an
internal temperature of only 70 °C. It is also possible that the increase in non-heme iron in
Ahn’s study came from sources other than the heme proteins such as the water-insoluble
fraction of meat. In the Ahn et al. (1993c) study, the experiment was conducted in a meat
system. No speciﬁc evidence was presented indicating that the non-heme iron increase
was from heme pigments when NaCl was present. On the other hand, Decker et al.

(1993) demonstrated that non-heme iron in the water-insoluble fraction of beef diaphragm

99

Table 3-2. Total and non-heme iron in a cooked1 water-washed muscle ﬁber (WF) model
system in the presence of selected heme proteins2 or salts3

 

 

 

Non-heme iron (ug Felg meat) Total iron
Treatment Day 0"” Day 1 Day 2
1. WF (control) 5.0‘ 5.4' 6.1‘I 7.0'
2. WF + Mb (myoglobin) 8.7" 8.7" 7.0" 23.8"
3. WF + Mb + NaCl 8.4" 7.5" 7.2" 22.7"
4. WF + Mb + KCl 8.2" 7.8" 7.3" 22.7"
5. WF + Mb + NaBr 5.9" 7.3" 7.4" 21.7"
8. WF + Mb + KBr 8.7" 7.3" 7.8" 22.8"
7. WF + Hb (hemoglobin) 70° 82° 78° 23.0"
8. WF + Hb + NaCl 73° 83° 77° 23.4"
9. WF + Hb + KCl 71° 77° 85° 23.4"
10. WP + Hb + NaBr 77° 77° 84° 23.7"
11. WF+Hb+KBr 71° 81° 85° 22.7"

 

ﬂ

. The cooked samples were heated to an internal temperature of 70 °C in a 83 °C water
bath.

Hemoglobin and myoglobin solutions were prepared to target 16.5 ug F e/g meat or
approximately 5 mg myoglobin or hemoglobin/g meat.

NaCl, KCI, NaBr and KBr concentrations were 8.8 mg salt/g meat .

Samples were held at 4 °C.

Mean square error = 0.86.

Mean values within the same column with different superscripts are different

(p < 0.05).

.N

9‘95“.“

100
muscle can be released and will catalyze lipid oxidation. Hazell (1982) indicated that

water-insoluble iron accounted for 58% of the total iron in chicken. Carpenter et al.
(1995) reported that about 57 % (7.4 ug F e/g meat iron) of iron in turkey is non-heme
iron. However, Kanner et al. (1988) reported that only a small portion of the non-heme
iron in dark turkey meat is water-soluble (2.5 ug F e/ g meat iron). This study indicated
that the majority of the non-heme iron in turkey is in the water-insoluble ﬁ'action and
could be the source of the non-heme iron.

To conﬁrm that NaCl did not accelerate the release of iron from heme-containing
proteins in the WP model system, hemoglobin and myoglobin solutions (targeting 16.5 ug
Fe/g water) were cooked with greater salt concentrations (17.4 mg salt/g meat) or
temperatures (75, 85 and 87.5 °C). In previous studies ground pork signiﬁcant differences
were always found at this salt concentration. Both hemoglobin and myoglobin were stable
in 17.4 mg NaCl/g meat during heating (Table 3-3). Salt and heating to 70 °C did not
increase (p<0.05) non-heme iron release from these proteins (Table 3-3). When
hemoglobin solutions targeting the same iron concentration were cooked to internal
temperatures of 75, 85 and 87.5 °C in a 88-90 °C water bath, the corresponding non-heme
iron concentrations were 0.52, 0.52 and 0.54 ug Fe/g meat, respectively. Non-heme iron
analysis indicated that cooking at these temperatures and NaCl concentrations did not
signiﬁcantly increase non-heme iron concentrations. This study demonstrated that heme-
containing proteins were stable at cooking temperatures and NaCl concentrations
consistent with Oellingrath (1988). Oellingrath (1988) reported that the heme degradation
in metmyoglobin was less than 25% during 2 hr of heating at 78 °C in a test tube. When

heated to 100 °C for 2 hr, the reduction of heme in metmyoglobin was similar to the

101

Table 3-3. Effect of cooking1 and diﬁ‘erent salts2 on the iron release ﬁ'om heme3

proteins in a model system stored at 4 °C for 2 days

 

Non-heme iron (ug Fe/ g meat)4

 

 

 

Treatment Raws Cooked
1. Mb (myoglobin) 0.3' 0.3'
2. Mb + NaCl 0.3‘ 0.3‘
3. Mb + KCl 0.3‘ 0.4‘
4. Mb + NaBr 0.3' 0.2'
5. Mb + KBr 0.3‘ 0.3'
6. Hb (hemoglobin) 1.2" 1.2b
7. Hb + NaCl 1.2" 1.2"
8. Hb + KCl 1.2" 1.4"
9. Hb+NaBr 1.1" 1.4"

10. Hb + KBr 1.2" 1.2"

1. The samples were placed into test tubes and cooked in a 88 °C water bath to an internal

wlo

we

temperature of 70 0C.

NaCl, KCl, NaBr and KBr concentrations were 0.3mg salt/g meat .
. Hemoglobin and myoglobin solutions were prepared to target 16.5 ug Felg meat or
approximately 5 ml myoglobin or hemoglobin/ml water.

Mean square error = 0.02.

(p < 0.05).

. Mean values within the same column with different superscripts are different

102
reduction found at 78 °C, while methemoglobin solutions decreased about 35% after 2 hr

of cooking at 100 °C. Janky and Froning (1973) reported that only 13.7 percent of the
metmyoglobin in solution was denatured at 70 °C. This study indicated that the majority
of heme-containing proteins were stable even when held at 100 °C for 2 hr.

One factor which may have an important impact on non-heme iron concentration
in meat during cooking is the cooking rate (Chen et al., 1984; Buchowski et al., 1988).
Chen et al. (1984) demonstrated that slow heating increased the amount of non-heme iron
more than rapid heating in muscle pigment extracts. They suspected that heating rate may
have changed coagulation of the myoglobin molecule in such a way that the heme iron
could not be cleaved from the globin moiety. In this study, when the temperature in the
hemoglobin solutions were raised to internal temperatures of 75, 85 and 87.5 °C in a 88-
90 °C water bath in approximately 50 see, there were no signiﬁcant differences in the non-
heme iron concentrations. These data were consistent with Chen et al (1984) who
reported that rapid heating did not increase the concentration of non-heme iron.

However, recently published research concerning the effect of the rate of cooking on the
non-heme iron concentration has produced opposite results. Conforti and Giuffrida
(1995) reported that the rate of cooking (slow vs fast cooking) did not have an effect
(p<0.05) on the non-heme iron concentration of drumsticks or breasts.

Based on the cooking time, temperature and heating rate in our study, it is possible
that heme-containing proteins were stable at these temperatures and cooking rate.
However, these studies were conducted in a model system. There is always a limitation
whether the data gathered from a simple model system can be representative of what

really happens in the complex meat system. Many components in a meat system are not

103
present in the model system. For example, Schricker et al. (1982) reported that heating 3

g of beef for 20 min in a boiling water bath increased the non-heme iron concentration of
the ﬂesh beef muscle ﬁom 9.9 to 20.9 jig/g and that of aged beef muscle from 11.5 to
16.0 ug/g. The cooking conditions (100 °C for 20 min) in Schricker’s study was less
severe than the cooking conditions (100 °C for 2 hr) in Oellingrath’s study. However, the
non-heme iron concentration increase in a meat system (Schricker et al., 1982) is far
greater than the increase in a model system (Oellingrath, 1988). Oellingrath (1988)
reported that the denaturation temperatures for myoglobin in meat systems (Slinde, 1987)
were lower than those determined in the highly diluted myoglobin solution model systems.
He attributed the difference in the myoglobin denaturation temperature to the presence of
other less thermally stable proteins in meat which may destabilize the hemeproteins due to
interaction between denatured protein and undenatured myoglobin (Ledward, 1978). Han
et al. (1995) reported that ascorbic acid, which was not present in the heme protein model
system, caused iron release from hemoglobin. They reported that after 5 hr incubation at
37 °C, about 4.7% of total heme iron had been liberated from 3.2 mM hemoglobin by 5.0
mM ascorbic acid.

The non-heme iron concentrations in cooked samples were greater (p<0.05) than
raw samples (Tables 3-1 vs. 3-2). It has been suggested by other researchers that cooking
will denature the heme proteins, degrade the heme and subsequently release iron (Igene et
al., 1979; Chen et al., 1984; Oellingrath, 1988). Buchowski et al. (1988) studied the effect
of heating (60, 77 and 97 °C) on the iron distribution between meat and broth. They
reported that total, heme and non-heme iron concentrations in cooked beef samples all

increase after cooking. For the beef samples cooked to 97 °C, heme iron concentrations

104
were greater than raw samples. Since heme structure will be degraded and subsequently

release free iron during the cooking process, it is not likely that heme iron concentration in
the cooked sample will be higher than the raw sample unless food components were lost.
The amount of total iron in meat is ﬁxed and the total iron concentration is unlikely to
increase during cooking unless some components are lost during cooking. Therefore, the
increase in total, heme and non-heme iron in cooked samples in the Buchowski et al.
(1988) study probably was a concentration effect due to water and/or lipid losses during
the cooking process. Because cooking losses are a problem in muscle samples, it is hard
to determine which portion of the non-heme iron increase is due to heme degradation and
which is attributed to a concentration effect. In this study because of the relative stability
of heme iron in hemoglobin and myoglobin during cooking and because water loss
occurred during the cooking process, the effect of cooking on the heme degradation and

non-heme iron release is small.

The effect of salt on lipid oxidation in the presence of heme iron

The effect of chloride and bromide salts on lipid oxidation in raw and cooked
treatments containing heme proteins is presented in Tables 3-4 and 3-5. For raw samples,
the WP (control) had similar TBARS values during the 6 day reﬁigerated storage period.
There were no differences between heme protein and salt treatments on day 0. After 3
days storage, all treatments containing myoglobin and hemoglobin had greater (p<0.05)
TBARS values than the control. These data are consistent with Johns et al. (1989) who
reported that heme proteins are prooxidants in meat. Treatments containing myoglobin

and various salts had greater (p<0.05) TBARS values than treatments containing

105

myoglobin alone (Table 3-4). However, there were no signiﬁcant differences in non-heme
iron concentration between myoglobin treatments and treatments containing myoglobin
and various salts (Table 3-1). This indicates that NaCl did not promote lipid oxidation in
the model system by releasing ﬁ'ee iron from heme pigments. Other mechanisms are
involved.

For the cooked samples (Table 3-5), there were no differences between these
treatments at day 0. At day 2 of reﬁigerated storage, hemoglobin/salt or myoglobin/salt
treatments had greater (p<0.05) TBARS values than hemoglobin or myoglobin treatments.
This is consistent with Ahn et al. (1993b) who reported that treatments containing
hemoglobin and NaCl had greater TBARS values than the hemoglobin only treatments. As
was consistent with raw samples, the cooked samples had no signiﬁcant differences in the
non-heme iron concentrations between treatments containing heme proteins and

treatments containing heme proteins and various salts (Table 3-2).

Effect of salt on non-heme iron recovery and lipid oxidation in a model system

Total iron in the raw samples containing added iron was approximately 3.5 ug Felg
meat greater than the control (Table 3-6). It was close to the target added iron level of 3
ug Fe/g meat. At day 0, the non-heme iron recovery was 2.6 ug F e/ g meat (74%). The
addition of NaCl or other salts did not signiﬁcantly (P < 0.05) affect the recovery of added
iron in the model system. This observation is consistent with Ahn et al.(1993b) who
reported that added NaCl had no effect on the non-heme iron concentrations in treatments
with different levels of added iron. A similar non-signiﬁcant recovery was found after 6

days of reﬁigerated storage would be expected. The effect of chloride and bromide salts

106

Table 3-4. The eﬁ‘ect of heme proteins1 and diﬂ‘erent salts2 on TBARS values in a
raw water-washed muscle ﬁber (WF) model system

 

TBARS (mg malonaldehyde/Kg meat)

 

 

Treatment Day 03“" Day 3 Day 8
1. WF (control) 069' 0.78' 0.67'
2. WP + Mb (myoglobin) 0.84‘ 1.59" 1.58"
3. WF + Mb + NaCl 0.87' 289° 402°
4. WF + Mb + KC] 093° 310° 343°
5. WF + Mb + NaBr 0.85' 257° 219°
8. WF + Mb + KBr 0.86‘ 222° 244°
7. WP + Hb (hemoglobin) 0.80' 1.00" 1.03‘l
8. WP + Hb + NaCl 0.79' 1.08‘l 1.23d
9. WP + Hb + KCl 0.78a 0.94" 1.11‘I

10. WP + Hb + NaBr 0.78° 0.94°| 0.90“l

11. WP + Hb + KBr 0.83' 1.09‘1 0.90d

 

pus

approximately 5 ml myoglobin or hemoglobin/ml water.

Samples were held at 4 °C.
Mean square error = 0.37.

weww

(p < 0.05).

NaCl, KCI, NaBr and KBr concentrations were 8.8 mg salt/g meat .

. Hemoglobin and myoglobin solutions were prepared to target 16.5 ug Fe/g meat or

Mean values within the same column with different superscripts are different

107

Table 3-5. The effect of heme proteins1 and diﬁ‘erent saltszon TBARS values in a cooked3
water-washed muscle ﬁber (WF) model system

 

TBARS (mg malonaldehyde/Kg meat)

 

 

Treatment Day 0“"6 Day 1 Day 2
1. WF (control) 046' 0.55' 0.66‘
2. WP + Mb (myoglobin) 0.38‘ 059' 0.81"
3. WP + Mb + NaCl 033' 0.79" 159°
4. WP + Mb + KCI 044' 0.82" 149°
5. WP + Mb + NaBr 042' 0.90" 168“
8. WP + Mb + KBr 043' 0.90" 189°
7. WP + Hb (hemoglobin) 0.35" 0.62‘ 0.80"
8. WP + Hb + NaCl 0.36‘ 0.80" 130°
9. WP + Hb + KC] 0.31° 0.72" 1. 13°
10. WF+Hb+NaBr 035' 0.89" 141°
11. WP + Hb + KBr 0.33' 0.95" 159°

 

1. Hemoglobin and myoglobin solution were prepared to target 16.5 ug Felg meat or
approximately 5 ml myoglobin or hemoglobin /ml water.

2. NaCl, KCl, NaBr and KBr concentrations were 8.8 mg salt/g meat .

. The cooked samples were heated to an internal temperature of ‘70 °C in a 83 °C water

bath.

4. Samples were held at 4 °C.

. Mean square error = 0.24.

6. Mean values within the same column with different superscripts are different
(p < 0.05).

U)

U:

108

Table 3-6. The effect of chloride and bromide salts1 on the recovery of added iron2
in a raw water-washed muscle ﬁber (WF) model system

 

 

 

Non-heme iron(ug F e/g meat) Total iron

Treatment Day 03"" Day 3 Day 8

1. WP 2.2‘ 2.0' 2.1' 5.3'
2. WP + Fe 4.8" 4.0" 5.1" 9.1"
3. WP + Fe + NaCl 4.2" 4.1" 4.5" 8.7"
4. WP + Fe + KCl 4.4" 3.9" 4.4" 8.8"
5. WP + Fe + NaBr 4.1" 3.8" 4.8" 8.8"
8. WP + Fe + KBr 3.8" 4.0" 4.2" 8.5"

 

NaCl, KCI, NaBr and KBr concentrations were 8.8 mg/g meat .

FeSOr was added to target 3ug Fe/g meat.

Samples were held at 4 °C.

Mean square error = 0.13.

Mean values within the same column with different superscripts are diﬂ‘erent
(p < 0.05).

M+w-

109
on the recovery of the added iron in cooked samples (Table 3-7) was similar to that for

raw samples (Table 3-6). Kanner et al. (1991) suspected that when free iron ions are
added to the WF, the iron may be either bound to protein macromolecules or may be
somehow trapped in the WP system. They proposed that the prooxidant effect of NaCl is
to release free ionic iron from iron-binding macromolecules. Because the addition of NaCl
or other salts do not increase the recoveries of added iron in the model system, the ability
of NaCl to release ﬁee iron from iron-protein macromolecular interaction was not
demonstrated in this model system. These data provide indirect evidence which indicates
that the Kanner et al. (1991) theory is not likely. This study was conducted using a model
system which is limited to predict what really happens in a meat system as discussed
previously. On the other hand, all evidence which supports Kanner’s hypothesis was
indirect. Additional direct evidence is needed to prove Kanner’s hypothesis. Kanner et
al. (1991) reported that in preparation of the WF, the amount of chelatable iron in the
eluent extracted from muscle by 0.3 mg NaCl/g meat was double that extracted by
distilled water. They suggested that because 0.3 M NaCl solution washed out more
chelatable iron than distilled water, it was possible that NaCl can more efﬁciently extract
the added iron from minced muscles. However, they did not verify whether the greater
amounts of chelatable iron in the eluent was because NaCl broke the iron-protein
macromolecular interaction in the meat system ﬁrst and subsequently released iron into the
eluent or because the iron-protein macromolecular complexes were washed out by NaCl
ﬁrst and the iron in the complexes was released by the chelating agent later.

There were no signiﬁcant differences in TBARS values for raw samples (Table 3-

8) at day 0. After 3 days of refrigerated storage treatments containing added iron had

110

Table 3-7. The effect of chloride and bromide salts1 on the recovery of added iron2
in a cooked3 water-washed muscle ﬁber (WF) model system

 

Non-heme iron (ug Fe/ g meat)

 

 

Total iron

Treatment Day 0"”6 Day 1 Day 2

1. WP 2.1' 2.1' 2.2' 6.9‘
2. WP + Fe 4.8" 5.3" 5.9" 11.0"
3. WP + Fe + NaCl 4.8" 5.4" 5.0" 11.4"
4. WP + Pe + KCl 4.1" 4.1" 4.5" 11.2"
5. WP + Fe + NaBr 5.0" 4.8" 5.3" 12.8"
8. WP + Fe + KBr 4.2" 4.0" 3.8" 10.9"

 

1. NaCl, KCI, NaBr and KBr concentrations were 8.8 mg/g meat .

2. FeSO4 was added to target 3ug Fe/g meat in samples.

3. The cooked samples were heated to an internal temperature of 70 °C in a 83 °C water
bath.

4. Samples were held at 4 °C.

5. Mean square error = 0.11.

6. Mean values within the same column with different superscripts are different
(p < 0.05).

111
greater (p<0.05) TBARS values than the control. This is consistent with Sato and

Hegarty (1971) who suggested that non-heme iron promotes lipid oxidation. After 6 days
storage, iron/salt treatments were similar to iron-only treatments. In the cooked model
system (Table 3-9), similar trends were found except that iron treatments with salt had
signiﬁcantly greater (p<0.05) TBARS values than the treatment with only added iron after
2 days of refrigerated storage. However, the non-heme iron concentrations in the iron/salt
treatments were not signiﬁcantly greater than the iron only treatments (Tables 3-7). These
results again conﬁrm that NaCl promotes lipid oxidation by mechanisms other than by
increasing the non-heme iron concentration in meat.

Kanner et al. (1991) reported that increasing the concentration of NaCl enhanced
lipid oxidation in raw minced muscle, especially after a freeze-thaw cycle. They suggested
that NaCl and freeze-thaw cycles may cause fusion of the intracellular compounds and the
destruction of the cell structure which further enhances lipid oxidation (Shomer et al.,
1987). However, the major focus of their paper is that the prooxidant effect of NaCl is to
increase the availability of iron ions during lipid oxidation. The important impact of NaCl
on structure integrity related lipid oxidation had not been emphasized. It is generally
believed that lipid oxidation in muscle foods is initiated in the highly unsaturated
phospholipid fraction in subcellular membranes (Gray and Pearson, 1987). Any disruption
of muscle membrane integrity by mechanical deboning, mincing, restructuring, or cooking
alters cellular compartmentalization. These processes will increase the chance of
interaction between prooxidant and unsaturated fatty acids and cause the increase of lipid
oxidation. Chang and Watts (1950) suggested that salt may affect the physical state of

meat in such a way that hemoglobin would be brought into closer contact with the fat and

112

Table 3-8. The effect of iron1 and chloride and bromide salts2 on TBARS values in
a raw water-washed muscle ﬁber (WF) model system

 

TBARS (mg malonaldehyde/Kg meat)

 

 

 

Treatment Day 03’4’5 Day 3 Day 6
1. WF 060‘ 0.75’I 0.78‘
2. WP + Fe 0.93' 1.81" 1.87"
3. WP + Fe + NaCl 0.88' 1.81" 1.94"
4. WP + Fe + KCl 0.88‘ 1.83" 1.85"
5. WP + Fe + NaBr 0.86' 1.53" 1.72"
6. WP + Fe + KBr 088“ 1.48" 1.82"
1. FeSO4 was added to target 3ug Fe/g meat in samples.

2. NaCl, KCl, NaBr and KBr concentrations were 8.8 mg/g meat .

3. Samples were held at 4 °C.

4. Mean square error = 0.03.

5. Mean values within the same column with different superscripts are different (p <

0.05).

113

Table 3-9. The effect of iron' and chloride and bromide salts” on TBARS values in
cookedc water-washed muscle ﬁber (WF) model system

 

TBARS (mg malonaldehyde/Kg meat)

 

 

Treatment Day 0“" Day 1 Day 2
1. WP 119' 157' 1.73'
2. WP + Fe 1.89" 2.79" 3.07"
3. WP + Fe + NaCl 2.17" 3.24" 384°
4. WP + Pe + KCl 2.09" 3.21" 353°
5. WP +Fe+NaBr 2.17" 3.41" 337°
6. WP + Fe + KBr 2.07" 3.23" 388°

 

p-u

bath.
4. Samples were held at 4 °C.
. Mean square error = 0.03.

kl!

. F eSO4 was added to target 3ug Felg meat in samples.
2. NaCl, KCI, NaBr and KBr concentrations were 8.8 mg/g meat .
3. The cooked samples were heated to an internal temperature of 70 °C in a 83 °C water

1 l4
subsequently promote lipid oxidation. This study demonstrated that salt can promote lipid

oxidation in the model system containing added iron (Tables 3-9) or heme proteins (Tables
3-4 and 3-5) without increasing the non-heme iron concentrations in the system.

However, salt alone, without another prooxidant such as iron or heme proteins, will not
promote lipid oxidation in the water-washed model system (Kanner et al., 1991). This
demonstrates the importance of a catalyst such as free iron or heme proteins for the
prooxidant effect of NaCl to occur. Since NaCl can promote lipid oxidation without
increasing the non-heme iron concentration in the system, we believe the major ﬁrnction of
NaCl in lipid oxidation is to reduce the integrity of muscle tissue instead of making more
iron available to catalyze lipid oxidation (Kanner et al., 1991). Therefore, lipids are more
accessible for the attack of the prooxidants (free iron). However, more research is needed

to verify this theory.

115

Summary and Conclusions

The effects of chloride and bromide salts on non-heme iron release from
hemoglobin and myoglobin and on the recovery of the added iron were studied in a raw
and cooked WF model system. The addition of selected chloride and bromide salts did
not increase the release of the non-heme iron ﬁ'om either hemoglobin or myoglobin in this
model. Various salts also did not increase the recovery of the added iron. However, the
addition of the same salts to the treatments with heme proteins or added iron signiﬁcantly
increased lipid oxidation. This indicates that NaCl does not promote the lipid oxidation by
increasing the non-heme iron concentration in meat. Other mechanisms must be proposed

to explain why NaCl is a prooxidant in this model system.

116

References

Ahn, D.U. and Maurer, A.J. 1989. Effects of sodium chloride, phosphate, and dextrose
on the heat stability of puriﬁed myoglobin, hemoglobin, and cytochrome c. Poultry
Sci. 6831218.

Ahn, D.U, Wolfe, F .H. and Sim, J .S. 19938. Prevention of lipid oxidation in pro-cooked
turkey meat patties with hot packaging and antioxidant combinations. J. Food Sci.
58:283.

Ahn, D.U., Ajuyah, A., Wolfe, F.H. and Sim, J .S. 1993b. Oxygen availability affects
prooxidant catalyzed lipid oxidation of cooked turkey patties. J. Food Sci. 58:278.

Ahn, D.U., Wolfe, F.H. and Sim, J .S. 1993c. Three methods for determining nonheme
iron in turkey meat. J. Food Sci. 58:288.

Andersen, H.J. and Skibsted, L.H. 1991. Oxidative stability of frozen pork patties. Effect
of light and added salt. J. Food Sci. 56: 1 182.

Arnold, A.R., Bascetta, E. and Gunstone, FD. 1991. Effect of sodium chloride on
prooxidant activity of copper (II) in peroxidation of phospholipid liposomes. J.
Food Sci. 56:571.

Buchowski, M.S., Mahoney, A.W., Carpenter, CE. and Comforth, DP. 1988. Heating
and the distribution of total and heme iron between meat and broth. J. Food Sci.
53:43.

Carpenter, CE. and Clark, E. 1995. Evaluation of methods used in meat iron analysis
and iron content of raw and cooked meats. J. Agric. Food Chem. 4321824.

Chang, I. and Watts, BM. 1950. Some effects of salt and moisture on rancidity of fat.
Food Res. 15:313.

Chen, C.C., Pearson, A.M., Gray, J .I., Fooladi, M.H. and Ku, PK. 1984. Some factors
inﬂuencing the nonheme iron content of meat and its implications in oxidation. J.
Food Sci. 49:581.

Conforti, PD. and Giuffrida, ML. 1995. Determination of oxidative changes in precooked
chicken parts by nonheme iron content and thiobarbituric acid value. Poultry Sci.
74: 1224.

Crackel, R.L., Gray, J.I., Pearson, A.M., Booren, A.M. and Buckley, DJ. 1988. Some
further observations on the TBARS test as an index of lipid oxidation in meats.
Food Chem. 28: 138.

117

Decker, E.A., Cnim, AD, Shantha, NC. and Morrissey, PA. 1993. Catalysis of lipid
oxidation by iron ﬁom an insoluble fraction of beef diaphragm muscle. J. Food Sci.
58:233.

Decker, EA. and Hultin, H0. 1990. Factors inﬂuencing catalysis of lipid oxidation by
the soluble fraction of mackerel muscle. J. Food Sci. 55:947.

Gill, J.L. 1978. “Design and Analysis of Experiments in the Animal and Medical
Sciences.” Iowa State Univ. Press, Ames, IA.

Gray, J .I. and Pearson, A.M. 1987. Rancidity and warmed-over ﬂavor. Adv. Meat Res.
3:221.

Han, D., McMillin, K.W., Godber, J .S., Bidner, T.D., Younathan, MT. and Hart, LT.
1995. Lipid stability of beef model systems with heating and iron fractions. J. Food
Sci. 60:599.

Hazell, T. 1982. Iron and zinc compounds in the muscle of meats of beef, lamb, pork and
chicken. J. Sci. Food Agric. 33:1049.

Igene, J .0, King, J .A., Pearson, A.M. and Gray, J .I. 1979. Inﬂuence of Heme pigments,
nitrite, and non-heme iron on development of warmed-over ﬂavor (W OF ) in
cooked meat. J. Agric. Food Chem. 27 :838.

J anky, D.M. and Froning, G.W. 1973. The effect of pH and certain additives on heat
denaturation of turkey meat myoglobin. Poultry Sci. 52:152.

Johns, A.M., Birkinshaw, L.H. and Ledward, DA. 1989. Catalysts of lipid oxidation in
meat products. Meat Sci. 25:209.

Kanner, J ., Hazan, B. and Doll, L. 1988. Catalytic “free” iron ions in muscle foods. J.
Agric. Food Chem. 36:412.

Kanner, J ., Harel, S. and Jaffe. R. 1991. Lipid peroxidation of Muscle food as affected by
NaCl. J. Agric. Food Chem. 39:101.

Kijowski, J .M. and Mast, MG. 1988. Effect of sodium chloride and phosphates on the
thermal properties of chicken meat proteins. J. Food Sci. 53:367.

King, A.J. and Bosch, N. 1990. Effect of NaCl and KC] on rancidity of dark turkey meat
heated by microwave. J. Food Sci. 5521549.

Ledward, DA. 1978. Scanning calorimetric studies of some protein-protein interactions
involving myoglobin. Meat Sci. 2:241.

118

Lewis, PS. 1926. The kinetics of protein denaturation. Part 1. The effect of variation in
the hydrogen ion concentration on the velocity of the heat denaturation of
oxyhaemoglobin. Biochem. J. 20:965.

Love, J .D. and Pearson, A.M. 1974. Metmyoglobin and non-heme iron as prooxidants in
cooked meat. J. Agric. Food Chem. 22:1032.

Oellingrath, I.M. 1988. Heat degradation of heme in methemoglobin and metmyoglobin
model systems measured by reversed-phase ion-pair high performance liquid
chromatography. J. Food Sci. 53:40.

Osinchak, J .E., Hultin, H.O., Zajicek, O.T., Kelleher, SD. and Huang, CH. 1992. Eﬂ‘ect
of NaCl on catalysis of lipid oxidation by the soluble ﬁ'action of ﬁsh muscle. Free
Radical Biology & Medicine. 12:35.

Quinn, J .R., Raymond, DP. and Harwalkar, V.R. 1980. Differential scanning calorimetry
of meat proteins as affected by processing treatment. J. Food Sci. 45: l 145.

Rhee, KS. and Ziprin, Y.A. 1987. Modiﬁcation of the Schricker nonheme iron method to
minimize pigment effects for red meats. J. Food Sci. 52:1174

Sato, K. and Hegarty, GR. 1971. Warmed-over ﬂavor in cooked meats. J. Food Sci.
36: 1098.

Schricker, B.R., Miller, DD. and Stouffer, J .R. 1982. Measurement and content of
nonheme and total iron in muscle. J. Food Sci. 47 :740.

Slinde, E. 1987. Color of blank salami sausage. Dissociation of heme from myoglobin and
hemoglobin. J. Food Sci. 5221152.

Shomer, 1., Weinberg, 2G. and Vasilever, R. 1987 . Structural binding properties of silver
carp muscle affected by NaCl and CaC12 treatments. Food Microstruct. 6:199.

Tarladgis, B.G., Watts, HM. and Younathan, MT. 1960. A distillation method for the
quantitative determination of malonaldehyde in rancid foods. J. Am. Oil Chem.
Soc. 37 :44.

SUMMARY AND CONCLUSIONS

A series of studies was conducted to investigate the prooxidant mechanism of
NaCl, with focus on the effect of NaCl on non-heme iron concentrations in ground pork 1
and in 8 WF model system as well as its relationship to lipid oxidation. Our hypothesis
was that the prooxidant mechanism of NaCl in meat is primarily to make more iron ions
available to promote lipid oxidation. This is done by: 1. Facilitating the release of iron
ions from iron-containing proteins (hemoglobin and myoglobin); and/or 2. Breaking
protein-iron interactions in a meat system.

Three methods for quantitating non-heme iron were evaluated to better understand
why these methods of analysis produce different values for non-heme iron when analyzing
the same sample. The extraction procedures for separating non-heme and heme iron were
responsible for the differences in non-heme iron values. The quantitation procedures for
these methods produced non-heme iron values which were not different from each other.
The data presented in this study indicated that the ferrozine method may not completely
recover added iron from ground pork. It was also recommended that the Igene method

should not be used because of its poor iron recovery. The modiﬁed Schricker’s method

119

120
was found to be the most effective procedure to completely extract non-heme iron ﬁom

meat.

The effects of NaCl on the non-heme iron concentrations and lipid oxidation were
studied in ground pork to verify that the prooxidant effect of NaCl in meat is to make
more iron ions available to catalyze lipid oxidation. Increasing NaCl concentrations
increased lipid oxidation and non-heme iron concentrations in both raw and cooked
samples. Non-heme iron analyses indicated signiﬁcant (p<0.05) increases in non-heme
iron content at day 6 for raw treatments and day 0 for the cooked treatments. NaCl, KCl,
NaBr and KBr addition resulted in similar effects with respect to lipid oxidation and non-
heme iron release. However, the exact cause for the non-heme iron increase was not
determined. It is possible that the increase in non-heme iron is caused either by a
concentration effect during cooking or by the addition of salt.

If NaCl causes non-heme iron increases in the meat, the next step is to examine
other possible iron sources such as ferritin and water-insoluble iron. A new non-heme iron
measurement method must be developed to resolve this question. The ferrozine and
modiﬁed Schricker methods are not capable of identifying whether non-heme iron release
is due to the effect of salt because most of the iron in the ferritin will be extracted because
of the extraction procedure used in the methods studied in our ﬁrst study.

The effects of NaCl on the non-heme iron released ﬁ'om the possible non-heme
iron protein sources (hemoglobin and myoglobin) and on the recovery of the added iron in
8 WF model system were studied. The addition of chloride and bromide salts had no
effect on the release of non-heme iron from hemoglobin and myoglobin in either the raw

and cooked samples in this model system. These salts had no effect on the recovery of the

121
added iron. However, the addition of the salts to the treatments with heme proteins or

added iron signiﬁcantly increased lipid oxidation compared to treatments with heme
proteins or added iron alone. This indicates that NaCl can promote the lipid oxidation
without increasing the non-heme iron concentration in meat.

It was our hypothesis that the prooxidant effect of NaCl in meat is primarily to
make more iron ions available to catalyze lipid oxidation. In this study, the addition of
NaCl in ground pork caused small increases in the non-heme iron concentration in ground
pork. However, in an uncooked system this increase in the non-heme iron concentration
was only signiﬁcant after 6 days of reﬁigeration. The exact cause for the non-heme iron
increases in ground pork (both cooked and uncooked studies) is still not determined. It is
possible that the non-heme iron increase is caused by the addition of salt. However, it is
also possible that it is the concentration effect during cooking which causes the non-heme
iron increase. Data from the WP model system indicated that various salts had no effect
on the non-heme iron release from heme proteins. Salt promotes lipid oxidation without
increasing non-heme iron concentrations in the WP model system. Therefore, increases in
non-heme iron availability in the meat system may not be a major pathway and can not be
solely responsible for the dramatic increase in lipid oxidation when NaCl was present. Our
data indicate it is likely that another prooxidant mechanism for NaCl exists. However, this
study only provides these observations in a model system and does not provide similar
observations in meat. It is unlikely that the effect of salt in meat will produce different
results. However, ﬁrrther research is needed to conﬁrm this.

Based on this study, we suggest that the major function of NaCl in lipid oxidation

is to make muscle tissue more vulnerable for the attack of the prooxidants (free iron or

122

heme proteins). It is likely that ﬁ'ee iron ion, heme proteins or other prooxidants already
exist in the meat system. When NaCl is added to the meat, the system is more susceptible
lipid oxidation due to microstructure changes. Therefore, even a small amount of free iron
will catalyze lipid oxidation because the lipids are more susceptible to oxidative change.
This salt-mediated lipid oxidation is the result of synergism between meat microstructure

damage and the presence of catalysts such as iron.

FUTURE RESEARCH

1. To firrther study the effect of NaCl on the non-heme iron concentration in the meat
system.

0 To determine whether the non-heme iron concentration increase in ground pork
is caused by concentration effects due to moisture loss during cooking or by the
addition of salt.

0 To develop a new procedure for measuring non-heme iron which detects the
release of iron from all iron sources including ferritin and water-insoluble
fractions of meat.

0 To determine the effect of NaCl on the release of non-heme iron from fenitin or
the water-insolube ﬁ'action of meat.

2. To verify the hypothesis that the prooxidant effect of NaCl is due to muscle

microstructure damage and the presence of catalysts such as free iron.

123

124

3. To re-evaluate the mechanism of cooking on the lipid oxidation.

0 To emphasize non-heme iron release in the meat system when cooking losses
are constant.

0 To determine the effect of cooking on the release of free iron from the ferritin
and water-insoluble fractions of meat.

0 To study the effect of heat-activated heme protein as a catalyst for lipid
oxidation in both model systems and meat systems.

0 To study the effect of loss of membrane integrity losses due to heating on lipid

oxidation.

APPENDICES

126

APPENDIX A

Modiﬁed Schricker non-heme iron procedure (Rhee and Ziprin, 1987)

1.

Five grams of very ﬁne ground meat (in triplicate) were weighed into test tubes with
screw caps.

The meat was mixed thoroughly in each tube with 0.4 ml NaNOz reagent (156 ug
NaNOz/g meat based on the meat weight) and 5 ml distilled water.

. Allow to equilibrate for 30 min.

. Fifteen ml of the acid mixture [HCL (6N):trichloroacetic acid (40%) = 1:1] were added

to each tube and the tube was tightly stoppered.

. The tubes were incubated in a water bath-shaker at 65 °C for 20 hr and then cooled to

room temperature in cool water.

. Allow the tubes to settle until the supernatant become clear.

One ml of the acidic liquid supernatant was transferred to a small test tube and 5 ml
bathophenanthroline disulfonate reagent solution added.

. The absorbance of the supernatant was read at 540 nm against the reagent blank (1 ml

acid mixture + 5 ml color reagent).

127

APPENDIX B

F errozine non-heme iron procedure (Carter, 1971)

1.

9.

Three grams of ground meat (in duplicate) and 9 ml 0.1 M citrate-phosphate buffer
were put into 50 ml centrifuge tube.

. Homogenize using a polytron for 2 seconds at top speed.
. Added 4 ml 2% ascorbic acid solution.
. Incubate for 15 min. at room temperature.

. Add 8 ml trichloroacetic acid solution and mix thoroughly.

Centrifuge at 3000 G for 10 min.

Transfer 5 ml supernatant into test tube.

. Add 2 ml buffer, 0.5 ml ferrozine reagent solution (0.3% ferrozine + 0.3%

neocuproine) and mix thoroughly.

Allow to stand for 5 min.

10.The absorbance of the supernatant was read at 563 nm.

128
APPENDIX C

Igene non-heme iron procedure (Igene et al., 1979)

1.

Eight grams of ground meat and distilled water (in duplicate) were weighed into test
tubes.

Add 4 ml 2% EDTA solution.

. Homogenize using a polytron for 2 seconds at top speed.
. Added 4 ml 12.5% trichloroacetic acid.

. Centrifuged at 3000 G for 10 min.

. Transfer the supernatant to another test tube.

. Measure free iron using atomic absorption spectroscopy

129
APPENDIX D

Total iron analysis procedure

1. 3 gram of ground pork were placed in a ﬂask.

2. 3 ml perchloric acid and 15 ml nitric acid were added to the ﬂask.

3. The sample was digested on a heat plate.

4. Dilute the digested solution to an approximate 1 to 10 ug F e/g solution range.

5. Measure the iron concentration in atomic absorption spectroscopy.

130
APPENDIX E

The recovery of added iron ﬁ'om raw ground pork using three procedures for determining

non-heme iron.

 

 

 

 

Non-heme iron (ug Fe/ g meat)‘ Recovery (%)”
modiﬁed modiﬁed
Igene Ferrozine Schricker Igene F errozine Schricker
pork (control) 1.1:0.2 26:03 39:03
pork (3 ug Felg meat) 29:05 53:02 74:13 61:16 89:8 117:35
pork (6 ug Fe/g meat) 38:03 82:03 10.0:0.4 46: 3 92:5 102:2
pork (9 ug Fe/g meat) 42:03 104:0.4 13.4: 1.1 35: 2 86:3 106:9

 

a.Mean : standard deviation (n =3)

b.Recovery (%) = (measured non-heme iron in treatment - control) x 100/added non-

heme iron

13 1
APPENDIX F

The recovery of added iron ﬁ'om cooked' ground pork using three procedures for
determining non-heme iron

 

Non-heme iron (ug F e/ g meat)b Recovery (%)c

 

 

modiﬁed modiﬁed
Igene F errozine Schricker Igene Ferrozine Schricker

 

pork (control) 22:06 40:03 5. 1 :04

pork (3 ug Fe/g meat) 34:09 71:05 87:02 41:11 104:15 119:16
pork (6 ug Felg meat) 4.2:1.3 99:0.9 12.4:0.6 33:13 98:18 122:13
pork (9 ug Fe/g meat) 4.8:1.6 12.4:09 15.2:1.0 29:12 94:10 111:15

 

a.The cooked samples were heated to internal temperature to 70C in a water bath
(temperature = 83 C).

b.Mean : standard deviation (n =3)

c.Recovery (%) = (measured non-heme iron in treatment - control) x 100/added non-
heme iron