A COMPARISON OF MYOGLOBIN AND

N10N~HEME IRON AS PROOXIDANTS . V
IN COOKED MEAT AND DiSPERSIONS

0F PHOSPHOLIPID.

Thesis for the Degree of Ph. D.
, MECHEGAN STATE UNWERSETY
9 7 JANE DAVIS LOVE
1972

LIBRARY

Michigan State

University

 

 

 

ABSTRACT

A COMPARISON OF MYOGLOBIN AND NON-HEME IRON AS PROOXIDANTS
IN COOKED MEAT AND DISPERSIONS OF PHOSPHOLIPID

BY

Jane Davis Love

Oxidation of tissue phospholipids occurs readily in
cooked meat, resulting in the development of off-flavors
and odors. Hemoprotein muscle pigments and non-heme iron
have been implicated as the major prooxidants in cooked
meat. While both hemoproteins and non-heme iron have been
shown to be effective in accelerating the oxidation of
unsaturated fatty acids in model systems, their role in
promoting lipid oxidation in meat is not clearly defined.

The present investigation was designed to study the
relative importance of myoglobin and ferrous iron as
prooxidants in cooked meat. In the first phase of the
study, beef and pork muscle were water-extracted and the
aqueous extract was subjected to extensive dialysis. The
aqueous muscle extract and the dialyzable and non'dialyzable
portions of the extract were added back to the water-
extracted muscle, which was heated to 70°C and stored at
4°C. The Z-thiobarbituric acid (TBA) test was used to

assess the extent of lipid oxidation in the stored samples.

Jane Davis Love
Results indicated that the prooxidant substances in muscle
tissue were removed by water extraction. While some of the
prooxidant activity appeared to be located in the non—
dialyzable portion of the aqueous muscle extract, a major
portion of the prooxidants could be removed from the aqueous
extract by dialysis. These results indicated that non-
heme iron plays a major role in accelerating lipid oxidation
in cooked meat.

To test the hypothesis that non-heme iron rather than
hemoproteins was the active prooxidant in the muscle extract,
model systems were devised. Purified myoglobin and ferrous
iron were added to the water—extracted beef and pork muscle,
the samples were heated to 70°C and stored at 4°C until
analyzed using the TBA test. The addition of myoglobin at
levels ranging from 1 to 10 mg/g failed to accelerate the
oxidation of the heated muscle. Ferrous iron was an
effective prooxidant at levels as low as 1 ppm. Low levels
of ascorbic acid enhanced the activity of Fe+2.

The next phase of the study was designed to assess the
ability of non-heme iron and hemoproteins to catalyze the
oxidation of a purified phospholipid, phosphatidyl ethanol-
amine (PB). Phospholipids are known to be the components
of cooked meat that oxidize most rapidly. The model systems
provided a less variable and less complex environment for
studying their oxidation than meat. A variety of methods

was used to assess the effects of ferrous iron and

Jane Davis Love
metmyoglobin on the oxidation of PE in aqueous dispersions
buffered at pH 5.5 and 7.0. Both metmyoglobin and ferrous
iron increased oxygen uptake by PB dispersions at pH 5.5.
A corresponding increase in TBA reactive material and
fluorescent products of lipid oxidation was also noted.
Due to increased oxidation in the presence of metmyoglobin
and ferrous iron at pH 5.5, a greater loss of unsaturated
fatty acid residues was apparent in the PB containing met-

myoglobin or Fe+2

than in the control samples.

At pH 7.0, neither metmyoglobin nor ferrous iron
accelerated oxygen uptake in aqueous PE dispersions.
Ferrous iron did not increase the production of fluorescent
materials under the same conditions, which confirmed the
lack of prooxidant activity. Metmyoglobin did increase the
fluorescence at pH 7.0, in spite of its apparent lack of
prooxidant activity. Addition of ferrous iron slightly
increased the loss of unsaturated fatty acids at pH 7.0,
although the effect was much greater at pH 5.5. Addition
of metmyoglobin at pH 7.0 failed to increase the oxidative
degradation of unsaturated fatty acids.

Spectral studies indicated that the heme group of
metmyoglobin was destroyed after 24 hr of incubation in a
pH 5.5 PB dispersion. At pH 7.0, metmyoglobin was not
degraded under identical conditions.

The volatiles found in the headspace above oxidizing

phospholipid dispersions were similar to the headspace

volatiles in a sample of oxidized cooked meat. Hexanal

Jane Davis Love
was a major component in the volatiles of both oxidizing
phospholipids and cooked meat. The concentration of n-
hexanal increased as a result of oxidation and apparently

is related to the development of off-odors in cooked meat.

A COMPARISON OF MYOGLOBIN AND NON-HEME IRON AS PROOXIDANTS
IN COOKED MEAT AND DISPERSIONS OF PHOSPHOLIPID

By

P

1‘ ‘-
Jane Dav1s Love

A THESIS

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

1972

6,“ ACKNOWLEDGMENTS

The author is indebted to Dr. A. M. Pearson for his
direction throughout the course of graduate study and
during the preparation of the thesis. Sincere apprecia-
tion is also extended to the guidance committee: Drs.

L. L. Bieber, L. R. Dugan, Jr., G. A. Leveille and D. E.
Ullrey for their critical review of this dissertation.

Appreciation is expressed to the General Foods Corpora-
tion, White Plains, New York, for providing a fellowship,
administered by the Institute of Food Technologists, to
the author and to Armour and Company, Chicago, Illinois,
for providing funds to support this study.

Finally, the author is especially grateful to her
husband, Mark. His encouragement, understanding and

Support are deeply appreciated.

ii

TABLE OF CONTENTS
Page
INTRODUCTION. . . . . . . . . . . . . . . . . . . . . 1
REVIEW OF LITERATURE. . . . . . . . . . . . . . .
Mechanisms of Lipid Oxidation.
Lipid Autoxidation. . . . . . . .
Metals as Prooxidants . . . .

Heme Compounds as Prooxidants
Heme Compounds as Antioxidants.

”V010! Oi Oi

Oxidation of Lipids in Meat. . . . . . . . . . 10
Composition of Muscle Lipids. . . 10
Phospholipid Oxidation in Cooked Meat . ll
Catalysts of Lipid Oxidation in Meat. . l3
Lipid Oxidation in Tissue Homogenates

and Model Systems . . . . . . . . . . 16

Lipid Oxidation and Meat Flavor. . . . . . . . 19

Measures of Lipid Oxidation. . . . . . . . . . 20
2-Thiobarbituric Acid (TBA) Test. . . . 20

EXPERIMENTAL. . . . . . . . . . . . . . . . . . . . . 23

Materials. . . . . . . . . . . . . . . . . . . 23
Samples . . . . . . . . . . . . . . . . 23
Solvents. . . . . . . 23

Thin- -Layer Chromatography (TLC) . . . . 23
Gas- -Liquid Chromatography (GLC) . . . . 23
Myoglobin . . .-. . . . . . . . . . . 24
Chemicals . . . . . . . . . . . . . . . 24

Experimental Systems . . . . . . . . . . . . . 24

Methods. . . . . . . . . . . . . . . . . . . . 25

Extraction of Muscle Tissue . . . . . 25

Fractionation of the Water Extract. . . 26
Preparation of Model Systems to Study

Lipid Oxidation in Cooked Meat. . . . 26

iii

Page
TBA Analysis of Lipid Oxidation

in Meat Systems . . . . 26
Isolation of Egg Yolk Phosphatidyl

Ethanolamine (PB) . . . . . . 27
Preparation of Aqueous Dispersions

of PE . . . . . . . . . . 30
Phosphorus Determination. . . . . . . 31
TBA Analysis of PE Dispersions. . . . . 32
Fluorescence of Oxidizing PE

Dispersions . . . . . . . 32
Oxygen Uptake by Oxidizing PE

Dispersions . . . . 33

Gas- -Chromatographic Analysis of the

Fatty Acid Composition of Phospha-

tidyl Ethanolamine. . . . . . 34
Gas- -Chromatographic Analysis of

Volatiles Produced by Oxidizing

Phospholipids . . . . . . 35
Gas- -Chromatographic Analysis of
Volatiles Produced by Cooked Meat . . 35

RESULTS AND DISCUSSION. . . . . . . . . . . . . . . . 37

Prooxidant Activity of Muscle Extract,
Dialyzate and Diffusate. . . . . . . . . . . 37

Prooxidant Activity of Myoglobin, Ferrous
Iron and Ascorbic Acid . . . . . . . . . . . 42

Oxidation of Aqueous, Buffered Dispersions
of Phosphatidyl Ethanolamine . . . . . . . . 52

Oxygen Uptake by Aqueous, Buffered PE
Dispersions. . . . . . . . . 53

TBA Values of Oxidizing Aqueous Dispersions
Of PB. 0 O O O O O O O O O O O O O O O O O O 58

Fluorescence of Oxidized Dispersions of PB . . 60
Fatty Acid Composition of Fresh and
Oxidized PE. . . . . '.° . . . . . . . . . . 66
Volatiles Produced by Oxidizing Phospho-
lipid Dispersions. . . . . . . 69
Volatiles Produced by Oxidized Meat. . . . . . 69
SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . 74
BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . . 76

iv

Table

LIST OF TABLES

The effect of adding various fractions of
aqueous muscle extracts-on TBA values of
water-extracted, heated beef muscle stored

‘for 18 hr at 4°C . . . . . . . .

The effect of adding various fractions of
aqueous muscle extracts on TBA values of
water-extracted heated beef muscle stored
for 48 hr at 4°C . . . . . . . .

The effect of myoglobin, ferrous iron and
myoglobin plus ferrous iron on TBA numbers

*of cooked, watereextracted beef stored for

48 hr at 4°C .

The effect of myoglobin, ferrous iron and
ferrous iron plus ascorbic acid on TBA

‘numbers of cooked, water-extracted pork
'samples stored-for 24 hr at 4°C.

The effect of PE concentration‘on oxygen

‘consnmption'during'lo hr oxidation at 38°C .

The effect of Fe+2, metmyoglobin and
ascorbic acid on the production of TBA
reactants in PE dispersions held at 38°C
for 24 hr. . . . . . . . . . . .

The change in fluorescence of an oxidizing
PE dispersion due to the addition of Fe 2

'or metmyoglobin. . . . .

The fatty acid composition of PE from egg

yolk and pork and beef muscle phospholipids.

'Ratio of unsaturated fatty acids to saturated

fatty acids in fresh PB and PE held for
24 hr at 38°C.. . . . . . . .

s

Page

38

39

44

SO

53

59

64

68

'Figure

10‘

11 '

LIST OF FIGURES

The effect of increasing‘levels of extract,
dialyzate and diffusate on the TBA numbers
of water-extracted beef stored for 18 hr
at 4°C after heating to 70°C .

The effect of myoglobin on the TBA numbers
of cooked, water-extracted beef stored for
24 hr at 4°C .

The effect of Fe+2 on the TBA numbers of
cooked, water-extracted'beef*stored for 24
hr at 4°C.

Changes in TBA numbers of water-extracted,
cooked pork loin muscle stored at 4°C.

Oxygen uptake of an oxidizing PE dispersion
at 38°C.

Oxygen uptake+of an oxidizing PE dispersion
containing Fe 2 at 38°C. . . .

Oxygen uptake of an oxidizing PE dispersion
containing metmyoglobin (MMb) at 38°C.

Excitation and emission curves for the
chloroform extract of an oxidizing PE

‘dispersion .

Fluorescent intensity in a PE dispersion as
a function of hours of oxidation at 38°C

'Gaseliquid'chromatogram'of'headspace‘vola-

tiles from oxidized phospholipid dispersions
Gas-liquid chromatogram of headspace vola-

tilgs from oxidized meat containing 5 ppm
Fe O O O O O 0 O O O O O O O O O O 0

vi

Page

41

43

45

51

SS

56

57

62

63

70

72

INTRODUCTION

Lipid oxidation is a major cause of deterioration in
the quality of meat and meat products. Fresh and frozen
meat undergo a gradual deterioration due to lipid oxida-
tion, while rapid acceleration in oxidation of lipids
occurs in cooked meat. With greater commercial production
and increased consumption of precooked meat items for
institutional and home use, elucidation of the mechanisms
involved in lipid oxidation has become increasingly
important.

Rapid deterioration in flavor occurring in cooked
meat has been linked to oxidation of the highly unsaturated
tissue phospholipids. Heme pigments have traditionally
been considered the catalysts of lipid oxidation in meat.
Recent evidence indicates that non-heme iron may play a
major role in accelerating the oxidation of meat lipids
(Sato and Hegarty, 1971).

Both hemoproteins and non-heme iron have been shown
to be effective in accelerating the oxidation of unsaturated
fatty acids in model systems. Ferric hemes are more active
prooxidants than ferrous hemoproteins, thus, metmyoglobin
win raw meat and the ferric denatured hemichromes in cooked

meat have been considered the catalytically active forms

2
of the muscle pigments. Non-heme iron in the ferrous
oxidation state is a more active prooxidant than ferric
iron. Ascorbic acid has been shown to enhance the pro-
oxidant activity of non-heme iron.

While model system studies indicate that either heme
or non-heme forms of iron are capable of accelerating lipid
oxidation under specified experimental conditions, in the
complex and more-variable meat system the prooxidant roles
of heme and non-heme iron are less clear. Consequently the
present investigation was designed to study the relative
importance of myoglobin and ferrous iron in catalyzing
lipid oxidation in cooked meat. The prooxidant activity
of heme and non-heme iron was evaluated in systems contain-
ing cooked, water-extracted muscle tissue or purified

phospholipids.

REVIEW OF LITERATURE

Mechanisms of Lipid Oxidation

Lipid Autoxidation

The spontaneous reaction between lipids and atmospheric
oxygen has been termed autoxidation. Lundberg (1962) has
pointed out that autoxidation of lipids involves a free
radical chain mechanism. Dugan (1961) described the early
stages of the reaction by the following simplified scheme:

Initiation

(1) RH + O2 --——-9 Rf + 'OH
Propagation

(2) R' + 02 ————4 R00'

(3) R00“ + RH ———-) ROOH + R'

The reaction is initiated when a labile hydrogen is
abstracted from a site on the lipid (RH), with the pro-
duction of lipid radicals (R’) as shown in step (1).
Reaction with oxygen (step 2), which yields peroxyl radicals
(ROO'), is followed by the abstraction of another hydrogen
(step 3). A hydroperoxide (ROOH) and another free radical
(R'), which is capable of perpetuating the chain, are formed.

Decomposition of the ROOH species forms more free
radicals, which participate further in the chain reactions

(Dugan, 1961). Lundberg (1962) stated that the

4
hydroperoxides may exist in an equilibrium as follows:
2 ROOH ““4 (ROOH)
x—-——- 2
At low total peroxide concentrations, hydroperoxide decompo-
sition is mainly monomolecular (Lundberg, 1962) and may be
illustrated by ROOH ————+~RO° + 'OH (Dugan, 1961). At
high hydroperoxide concentrations a bimolecular decomposi-
tion occurs (Lundberg, 1962). The decomposition may be

schematically illustrated by ZROOH ~—~—+ ROO' + R0” + H O

2
(Dugan, 1961).

Lea (1962) indicated that the hydroperoxides possess
very little "off-flavor." Dugan (1961) has suggested that
a variety of hydroxy and carbonyl compounds and shorter-
chain fatty acids are responsible for the odors and flavors
that are characteristic of rancidity.

Ingold (1962) presented the following mechanism to
summarize reactions resulting in termination of the lipid
oxidation chain:

ROZ’ + X —-——9 inactive products
He stated that X may be either a free radical (ROZ', RO°
etc.) or a free radical inhibitor.

While an autocatalytic mechanism may describe the
oxidative processes occurring in a highly refined fat or
oil, the situation in a food, such as meat, is more complex.
In meat, both neutral lipids and phospholipids are present,
and either class may undergo oxidation (Watts, 1962). The
reactivity of lipids is, of course, influenced by the

degree of unsaturation of the constituent fatty acids, as

5
well as by the presence of activating or inhibiting sub-
stances (Lundberg, 1962). The speed with which undesirable
rancid odors and flavors develop in animal tissues indi-
cates the presence of substances promoting the oxidation
of the unsaturated lipid components (Dugan, 1961). Hemo-
protein muscle pigments and metals, especially iron, have
been implicated as prooxidants in meat (Tappel, 1962;
Watts, 1962; Liu and Watts, 1970; Sato and Hegarty, 1971).
The proximity of lipids to muscle catalysts of lipid oxida-
tion may influence the extent of oxidation (El-Gharbawi

and Dugan, 1965).

Metals as Prooxidants

 

Ingold (1962) has summarized the activity of heavy
metals in increasing the rate of oxidation of food lipids.
He pointed out that metals such as iron, cobalt and copper,
possessing two or more valency states with a suitable
oxidation-reduction potential between them, are particue
larly important catalysts. He also stated that the effect
of metals can be reflected in an altered rate of chain
initiation, propagation or termination, as well as by an
altered rate of hydroperoxide decomposition.

The basic function of the metal catalyst is to
increase the rate of formation of free radicals (Ingold,
1962). Heaton and Uri (1961) have shown that metal ions
in their higher valency state will react directly with
lipid substrates. Ingold (1962) has proposed the following

equation for a reaction involving metal catalysis of this

type:

6

Mord)+ + RH ——-—> M(n+1)+H + R‘

Heaton and Uri (1961) have shown that metals in their
lower valency states may also initiate lipid oxidation
chain reactions directly. Uri (1956) and Heaton and Uri
(1961) suggested that the first stage in this process may

be represented as an activation of dissolved oxygen:

n+-

M +o ——‘M(n+1)+ .. '

. O2

2 ¢____.
A subsequent reaction with the organic substrate then
generates free radicals (Brown et al., 1963).

A Uri (1956) has described a commonly accepted mechanism
for metal catalysis involving the oxidation of a metal ion

with hydroperoxide decomposition resulting as follows:

M+n +(n+l) +

+ ROOH ——-> M ‘OH + RO’
Ferrous iron has been shown to have greater prooxidant
activity than iron in the ferric oxidation state (Brown et"
aZ., 1963; Wills, 1965; O'Brien, 1969). Bawn (1953)
noted that oxidation commences at the same time as metal
first appears in its higher valency state. A number of
investigators (Wills, 1965; Barber, 1966; Sato and Hegarty,
1971) have reported that low levels of ascorbic acid
increase the efficiency of iron as a catalyst for lipid
oxidation, presumably by regenerating the active ferrous
ion. }
Ingold (1962) has pointed out that metals exist as

hydrated ions in an aqueous lipid system. He proposed that

the water soluble radicals and products of lipid oxidation

7
enter the aqueous phase as lipid oxidation proceeds. He
then suggested that metals may react with the water soluble
species in the aqueous phase or with the substrate, radicals

or hydroperoxides at the oil-water interface.

Heme Compounds as Prooxidants

 

The catalytic effect of iron porphyrins on the oxi-
dative decomposition of polyunsaturated fatty acids was
first described by Robinson (1924). The heme catalyzed
oxidation of unsaturated fatty acids has been extensively
studied, and acceleration of lipid oxidation due to a
variety of heme compounds is now a generally accepted
phenomenon. Tappel (1962) has reviewed some of the
voluminous literature dealing with heme catalyzed lipid
oxidation.

According to Tappel (1962), catalysis by iron porphyrins
is characterized by rapid initiation and propagation of the
lipid oxidation chain reaction. He stated that the cata-
lytic homolytic cleavage of the O-OH bond of the hydro—
peroxide is a general property of hematin catalysts. He
suggested the following mechanism as the most probable one

for hematin-catalyzed unsaturated lipid oxidation:

OH
L.
LOOH Nyé /N
N / \
L LH
o
o 0'
I N \' N
N /N /
N>°N N,FRN
LO'

Tappel (1962) also suggested that a direct attack on the
lipid by the heme compound could result in generation of
lipid radicals according to the following mechanism:

3 2

LH + Hematin-Fe+ -————) 13 + Hematin-Fe+ + H+
Tarladgis (1961) attributed the catalytic activity of
ferric hemoproteins to the paramagnetic character of the
porphyrin bound iron. He suggested that the presence of
five unpaired electrons in metmyoglobin produces a strong
magnetic field that would favor the initiation of free
radical formation. He further reported that the decompo-
sition of hydroperoxides by a ferric porphyrin was mediated
through the donation of an electron from the v cloud of the

porphyrin ring.

Heme Compounds as Antioxidants
While the prooxidant activity of hemes has been known

for many years, it has been recognized more recently that

9
heme compounds can also act as antioxidants. Maier and
Tappel (1959), using a fixed heme concentration, observed
that when the linoleate concentration dropped below a
specified level, lengthy induction periods occurred.
Banks et at. (1961) found acceleration of fatty acid oxi-
dation with increasing cytochrome c concentrations up to
a maximum, while further increases resulted in inhibition.
Lewis and Wills (1963) have also reported that the pro-
or antioxidant activity of a heme compound is determined
by the ratio of heme to unsaturated fatty acids.

Linoleate:heme ratios for maximum catalysis of lipid
oxidation were determined by Kendrick and Watts (1969).
They reported Optimum linoleate: heme ratios of 100 for
hemin and catalase, 250 for metmyoglobin, 400 for cyto-
chrome c and 500 for methemoglobin. At heme concentrations
of two to four times the optimum catalytic amount, they
noted that oxidation did not occur. They theorized that
a stable lipid hydroperoxide-heme derivative was formed at
inhibitory heme concentrations. At‘lower heme concentra-
tions, it was postulated that the heme may be unable to
contain the lipid radicals and oxidation results, with
eventual destruction of the hemes.

Nakamura and Nishida (1971) reported that association
of fatty acids with hemoglobin was responsible for the
observed dependence of lipid oxidation on hemoglobin con-
centration. As the hemoglobin concentration in a linoleic

acid emulsion increased, they noted that an increasing

10
amount of linoleic acid was associated with the hemoglobin.
When more than 77% of the linoleate was bound to the hemo-
globin, a lengthy induction period was observed. The
authors reported that the visible spectra of the hemoglobin
indicated that it existed in a low spin ferric form during
the induction period. A carboxylate ion and a cis double
bond in the fatty acid structure were also reported to be

required for the binding of the fatty acid to the hemoglobin.

Oxidation of Lipids in Meat

 

Composition of Muscle Lipids

 

Lipids found in meat can be classified as intermuscular
or depot lipids and intramuscular or tissue lipids.« The
depot lipids are generally stored in specialized connective
tissues in relatively large deposits. The tissue lipids
are widely distributed throughout the muscle tissue, where
they are integra1“parts of cellular structures. ‘The intra-
cellular lipids exist in close association with proteins
and contain a large percentage of the total phospholipids
(Watts, 1962).

According to Hornstein et al. (1961) the phospholipids
in muscle contain a larger percentage of unsaturated fatty
acids than the neutral lipids. They found that 19% of the
fatty acids in beef phospholipids had four double bonds,
while only 0.1% of the triglyceride fatty acids showed this
degree of unsaturation. Particularly high levels of
linoleic and arachidonic acids are present in the phospho-

lipids (Giam and Dugan, 1965).

11

Phospholipids contribute less than 1% of total muscle
weight, while the triglyceride fraction is about five times
as large (Hornstein at aZ., 1961). The amount of phospho-
lipid has been shown to be relatively constant in muscles
from different animals or carcass locations, while the
amounts of total and neutral lipid are more variable
(Hornstein at aZ., 1967; O'Keefe at aZ., 1968).

Several researchers (Kuchmak and Dugan, 1965; Hornstein
et aZ., 1967; O'Keefe et aZ., 1968) have reported that the
fatty acid composition of phospholipids varies with carcass
location. 'Luddy et a2. (1970) reported that phospholipid
fatty acids from muscles classified as light colored had a
predominance of monoenes, while polyunsaturated fatty acids
were present in greater amounts in phospholipids from dark
muscles.

Giam and Dugan (1965) suggested that the tendency of
pork muscle to undergo oxidative deterioration may be due
to the high levels of linoleic and arachidoniC"acids in the
"free" and "bound" lipids. Compositional differences in
the fatty acids of the phospholipid fraction certainly
might be expected to result in varying susceptibilities
to oxidative rancidity in different species or in cuts from

different carcass locations.

Phospholipid Oxidation in Cooked Meat
Early studies of rancidity in meat were concerned
with the oxidation of adipose tissue lipids (Watts, 1954;

1961). Timms and Watts (1958) first noted that the rapid

12
deterioration in the flavor of cooked meat did not
correlate with measures of the oxidation of the neutral
lipids. Later work by Younathan and Watts (1960) showed
that more oxidation had occurred in the phospholipid
fraction from rancid, cooked pork than in the neutral
lipids.

Hornstein et a2. (1961) also observed that the phospho-
lipid fraction and the total lipids from pork and beef
became rancid quickly when exposed to air. These authors
concluded that the neutral fats developed off-flavors less
readily.

Thus, even though phospholipids are present in
relatively small amounts in muscle tissue, they are
apparently oxidized rapidly and contribute significantly
to stale flavors in meat (Watts, 1962). Their rapid oxi-
dation reflects the highly unsaturated nature of the
phospholipid fatty acids (Lea, 1957). Oxidized flavors
may be particularly intense in lean meat, since less
neutral fat is present to trap volatile decomposition
products of polar lipids (Watts, 1962).

On heating, the neutral lipids are lost from the meat
more readily than the phospholipids (Campbell and Turkki,
1967). The relative concentration of phospholipids in
meat thus increases as'a result of cooking. ‘Campbell and
Turkki (1967) have reported that cooking pork or beef by
a dry heat method fails to appreciably change the fatty
acid composition of the phospholipids. Giam and Dugan

(1965) also observed that there was little difference in

13
the fatty acid content of free or bound lipids in freeze
dried raw or cooked meat. Chang and Watts (1952) had
previously noted that the fatty acid composition of ether-
extractable lipids of meat and poultry was not changed by

cooking.

Catalysts of Lipid Oxidation in Meat

 

The rapidity with which animal tissue lipids undergo
oxidation has stimulated efforts to identify the catalysts
involved in the reaction. The accelerating effect of
iron porphyrins on the oxidation of lipids is a generally
accepted phenomenon. ‘Meat contains significant quantities
of hemoproteins, consisting of some hemoglobin from the
blood and larger amounts of the muscle pigment, myoglobin
(Craig et aZ., 1966).

Fox (1966) has reviewed the chemistry of meat pigments.
He stated that in fresh meat myoglobin exists in three
interconvertible forms. Oxymyoglobin imparts a desirable
bright red color to meat.‘ Reduced myoglobin is purplish-
red in color and metmyoglobin is responsible for the unde-
sirable brown to black discoloration in fresh meat. Watts
et al. (1966) stated that the balance between the different
pigment forms is affected by the activity of enzymatic
reducing systems in the meat and the oxygen concentration
of the surrounding atmosphere.

During cooking, the pigments are irreversibly con-
verted to denatured ferric hemichromogens (Fox, 1966).

Ledward (1971) has studied the nature of the cooked meat

- .l'lyl .IfllulI \ (I'll.

 

.lll l.l I in I'll.‘ . I'll I!“ III III

14
pigment. His study indicatesthat myoglobin coprecipi-
tates with other muscle proteins during heating. Increas-
ing temperature was presumed to result in conformational
changes in the hematin environment. Ledward (1971) also
postulated that denatured proteins may then attack the
hematin, resulting in replacement of apomyoglobin by other
proteins. He suggested that the pigments then aggregate
and precipitate with other unreacted denatured proteins
to form a range of denatured hemoproteins.

It has been'generally accepted that the heme compounds
of meat are catalysts of lipid oxidation (Watts and Peng,
1947; Tappel, 1952; Younathan and Watts, 1959; Watts,
1961; 1962; Tappel, 1962; Liu and Watts, 1970).“Brown and
co-workers (1963) reported that ferric hemes were more
active catalysts of lipid oxidation than ferrous hemes-
The rapid oxidation of lipidS'in cooked meat-has been
attributed to catalysis by the denatured ferric hemi-
chromes (YOunathan and Watts, 1959; Liu and Watts, 1970).

The ferric heme, metmyoglobin, has been implicated
as a catalyst of lipid oxidation in fresh meat (Greene et
aZ., 1971). Hutchins et a1. (1967) have reported a positive
correlation between metmyoglobin accumulation and-lipid
oxidation in raw meat. 'Ledward and MacFarlane (1971)
observed that high TBA values corresponded to high metmyo-
globin'concentrationS‘in"stored,'frozen'beef;'however,
apparently both TBA values and metmyoglobin accumulation

were dependent on pro-freezing‘treatmentz' Interestingly

15
enough, however, high initial metmyoglobin concentrations
did not affectthe rate of lipid oxidation during freezer
storage.

Even though heme pigments have traditionally been
implicated as the major prooxidants in meat, there is
evidence that non-heme iron may play an important role in
accelerating oxidation of muscle lipids (Moskovits and
Kielsmeier, 1960; MacLean and Castell, 1964; Sato and
Hegarty, 1971). Moskovits and Kielsmeier (1960) demon-
strated that contaminating iron functions is a prooxidant
in sausage. MacLean and Castell (1964) found that trace
amounts of iron added to cod'muscle produced a rancid
odor. Sato and Hegarty (1971) showed that non-heme iron
accelerated the oxidation of lipids-in water extracted
cooked meat. They also reported that myoglobin and heme-
globin failed to act as prooxidants in cooked meat.

If non~heme iron is the major prooxidant in muscle
tissue, the effectiveness of polyphosphates as inhibitors
of lipid oxidation in meat (Timms and Watts, 1958) is
easily explained. Presumably phosphates do not inhibit
hemeecatalyzed reactions, but act by sequestering trace
metals (Liu and Watts, 1970).

Kesinkel at at. (1964) observed that raw muscle was
subject to more extensive lipid oxidation at lower pH
values. Since non-heme iron is more active at acid pH‘
values (Wills, 1965; 1966), the observations of Kesinkel
et a1. (1964) support a major prooxidant role for non-heme

iron.

16

Lipid Oxidation in Tissue Homogenates
and Model Systems

Many studies have focused on the rapid non~enzymatic
lipid oxidation taking place in tissue homogenates and
particulate fractions exposed to atmospheric oxygen
(Bernheim, 1964; Robinson, 1965; Wills, 1965; 1966;
Barber, 1966). Lipid oxidation may occur in animal
tissues deficient in natural antioxidants (Desai and
Tappel, 1963), in tissues subjected to irradiation
(Kokatnur et aZ., 1966) or in tissues in which other
cell damage has occurred (Nishida and Nishida, 1965).

Wills (1966) attempted to assess the relative
importance of hemoprotein and non-heme catalysts of lipid
oxidation in various animal tissues incubated in air. He
compared the activity of tissue homogenates in“promoting
lipid oxidation of unsaturated fatty acid emulsions to
the results obtained in*mode1 systems~containing only
unsaturated fatty acids and home or non-heme iron. He
reported that both heme and non-heme iron were‘present in
most tissue fractions, and were capable'of”cata1yzing the
oxidation of added unsaturated fatty acids. ‘His results
indicated that non-heme iron was a more active prooxidant
at acid pH values, whereas hemoproteins were reported to
be less pH sensitive. He also reported that non—heme iron
was apparently more important than heme proteins in
catalyzing the oxidationof the endogenouslipids in.
tissue homogenates. Barber (1966) has also suggested that

non-heme iron and ascorbic acid constitute the normal

17
prooxidant system in animal'tissues exposed to atmospheric
oxygen.

Liu (1970) reported that a 1:1 complex of Fe (II)-EDTA
accelerated oxidation of a linoleic acid emulsion at pH
values lower than 6.4, while myoglobin catalysis increased
directly with pH from pH 5.6-7.8.. This is in contrast to
several other studies (O'Brien, 1969;‘Kendrick and Watts,
1969; Ben-Aziz at aZ., 1970) which have reported that while
myoglobin catalysis is less pH sensitive than iron catalysis,
it may be enhanced at acid pH values.

O'Brien (1969) studied the decomposition of linoleic
hydroperoxides by metal ions*and heme compounds. He
observed a‘pH maximum for‘the*catalyticactivity of iron
between pH 5.0 and pH 5.5. He also reported that the
activity of hemoproteinS'waS'enhanced at acid pH values,
presumably due toincreased exposure’of the heme“moiety of
the'hemoproteins.

'A number of studies (Wills, 1965; 1966; Barber, 1966;
O'Brien, 1969; Kendrick and Watts, 1970; Liu, 1970) of
heme and metal catalyzed lipid oxidation have been con-
ducted using purified unsaturated fatty acids a5*the lipid
reactant. 'Simila ‘studies wit 'phospholipids would be of
interest, since phospholipids are the lipid component
mostrapidly oxidized in cooked meat (Younathan and-Watts,
1960), in freeze-dried beef (EleGharbawi and Dugan; 1965)
and in cod muscle (Roubal, 1967). While an aqueous dis-
persion of phospholipids might be expected to behave like

an emulsion of unsaturated fatty acids, the phospholipids

18
offer a more complex oxidation system than the neutral
lipids. The phospholipids contain phosphorus and a
nitrogenous moiety as well as unsaturated fatty acids.

The tendency of phospholipids to oxidize very rapidly
is at least partially due to their high content of unsatur-
ated fatty acids (Lea, 1957). The phosphorylated bases
may also affect the oxidation of the unsaturated fatty
acids in the'phospholipid*molecu1e'(Corliss, 1968; Lee
Shin and Smith, 1970).

Lee Shin and Smith (1970) studied the effects of the
functional groups of the phosphoryl bases of phosphatidyl
ethanolamine‘and phosphatidyl choline on the oxidation of
methyl linoleate in aqueous emulsions. They reported that
ethanolamine and o-phosphoethanolamine increased oxygen
uptake by methyl linoleate at pH 7.9 and decreased oxygen
consumption at pH 10.2. The choline containing groups were
reported to have no effect on the rate of lipid-oxidation.

Corliss (1968) has reported that the induction period
for oxidation of phospholipids is a function of the
nitrogen containing moiety. He found that the ethanolamine
moiety of phosphatidylethanolamine exerts a greater pro-
oxidant effect than does the choline portion of phospha~
tidyl choline. ‘He then concluded that the rate of phospho-
lipid oxidation during the steady state was a function of
the unsaturation of the fatty acid components of

phospholipids.

19

Lipid.Oxidation and Meat Flavor

 

The flavor of cooked meat changes rapidly during
storage. Various terms such as "warmed over", "stale"
or "rancid" have been used to describe the characteristic
flavors developing in cooked, stored meat (Timms and
Watts, 1958). The compounds responsible for the undesirable
flavor changes are presumably degradation products of the
unsaturated phospholipid fatty acids, since these lipids
are rapidly oxidized in cooked meat (Younathan and Watts,
1960).

In recent years a number of attempts have been made
‘tO'characterize'the compounds involved in meat flavor.
Herz and Chang (1970) summarized much of this research in
a recent review. Some studies of meat flavor have dealt
with the contribution of lipids to flavor. Hornstein et
a1. (1960) and Hornstein and Crowe (1960) have shown that
depot fat plays an important role in the development of
species-specific aromas of cooked beef, pork and lamb.
However, they did not investigate the possible contribu-
tion of intramuscular fat to either desirablesor undesirable
flavors.

Q The most numerous members of any class of compounds
identified in meat flavor'concentrates are the carbonyl
compounds (Herz and Chang; 1970). Hexanal has been identi-
fied as a major component of meat flavor concentrates
(Herz and Chang, 1970). Hexanal, as well as-a variety of
other carbonyls, is a degradation product of oxidized

linoleic fatty acid (Gaddis at aZ., 1961).

20

There is some evidence that hexanal is a product of
lipid oxidation occurring-in cooked meat (Cross and
Ziegler, 1965). The volatile fractions isolated from
cured and uncured pork were compared by Cross and Ziegler
(1965). They reported that hexanal and valeraldehyde were
present in uncured, but not in cured, pork. The failure
of cured, cooked meat to undergo the extensive lipid
oxidation observed in cooked, uncured meat has been noted
by Younathan and Watts(l959).~ Thus, certain of the
carbonylcompoundsidentified in meat flavor concentrates
may result from lipid oxidation and contribute undesirable

instead of desirable flavors to cooked meat.

MeasureS‘of‘Lipid'Oxidation
Oxidative deterioration of lipids can be followed by
measuring the oxygen consumption, by determining the
concentration of products of the reaction or by assessing
the decrease in the concentration of unsaturated fatty
acids. All known measures of lipid oxidation have limi-
tations and cannot be applied with equal success in all

systems of oxidizing lipids.

Z-Thiobarbituric Acid (TBA) Test

The Z-thiobarbituric acid (TBA) test has been used
extensively in the study of lipid oxidation (Lea, 1962).
Sinnhuber at al. (1958) indicated that the TBA test measures
the pink color produced by reacting TBA with malonaldehyde.
Malonaldehyde and/or substance5°closely resembling it

occur in foods as end products of the oxidation of

21
polyunsaturated fatty acids (Kwon et aZ., 1965). Malon-
aldehyde may be'a secondary oxidation product derived from
unsaturated aldehydes resulting from the cleavage of
hydroperoxides (Evans, 1961). According to Tarladgis
at al. (1960) malonaldehyde*itself does not contribute to
typical rancid odors, although a high correlationsbetween
malonaldehyde content and rancid odors has been noted
(Zipser at aZ., 1964). 'The relationship may‘be limited
to moist foods, especially animal tissues (Kwon and Watts,
1964).

The TBA test can be applied directly to a lipid-
containing material without prior extraction of the fat
(Lea, 1962). However, substances other than lipid oxida-
tion products can react with the'TBA*reagent and give the
red colored compound measured spectrophotometrically in
the TBA test (Kwon et a1., 1965). Alternatively, the
product can be acidified and steam distilled and the TBA
reaction can be carried out on the distillate (Tarladgis
at aZ., 1960).

2 or Fe+3 has

The presence of trace amounts of Fe+
been reported to cause spurious increases in.TBA values
(Wills, 1964; Castell et aZ., 1966; Castell and Spears,
1968). Presumably this is due to iron catalysis of addi-
tional lipid oxidation during blending, heating or distilla-
tion procedures (McKnight and Hunter, 1965).' Castell et
a1. (1966) have recommended that EDTA be added during the
blending process. McKnight and Hunter (1965) suggested

that samples containing iron be heated or distilled

22
anaerobically. Ascorbic acid has also been reported to

result in high TBA values (Wills, 1966).

‘EXPERIMENTAL

Materials

Samples

Fresh, U.S. Choice grade beef round and fresh, grade
A hens' eggs were obtained from the Michigan State Uni-
versity Food Stores. Pork shoulder and loin samples were
obtained from the’Meats Laboratory at Michigan State

University.

Solvents
All solvents were freshly redistilled before use.
Methanol was refluxed with 2,4~dinitr0phenylhydrazine and

trichloroacetic acid prior to distillation.

ThineLayer Chromatography (TLC)
Silica Gel G, obtained from*Brinkmann Instruments,

Westbury, New York, was used for all TLC work.

Gas-Liquid Chromatoggaphy'(GLC)

Chromosorb W, acid washed (80-100 mesh), coated with
15% diethylene glycol succinate (DEGS), and mixtures of
fatty acid methyl esters were obtained from Applied Science
Laboratories, Inc., State College, Pennsylvania. Acid

washed, 60h80 mesh Chromosorb W“coated with 20% Carbowax

23

24
20M was obtained from Varian Aerograph, Walnut Creek,

California.

Myoglobin
Whale skeletal muscle myoglobin, A grade, was

purchased from Calbiochem, Los Angeles, California.

Chemicals

 

All other chemicals were of reagent grade, unless

otherwise specified.

Experimental Systems

The effect of heme and non-heme iron on lipid oxida-
tion was assessed in various systems. Initial studies
were designed to determine whether the prooxidants occur-
ring in muscle were low-molecular weight, dialyzable
substances (non-heme iron) or non-dialyzable proteins
(myoglobin and hemoglobin). Water-extracted, ground beef
or pork muscle was the basis of the systems used in this
phase of the study. The water extract contained prooxidant
substances. The activity of the dialyzate and diffusate
from the extract in accelerating lipid oxidation in the
extracted meat samples was evaluated.

The next series of experiments involved model systems
designed to ascertain whether purified metmyoglobin and
non-heme iron would exhibit prooxidant activities similar
to those observed for the dialyzate and diffusate. Puri-
fied heme proteins, ferrous iron (Fe+2) or Fe+2 and ascorbic
acid were added to samples of the water-extracted muscle,

and the extent of lipid oxidation was evaluated.

25

Since"phospholipid5‘are'the“component5‘of'cooked
meat undergoing the most rapid oxidative deterioration,
model'systems-containing*purified*phospholipidS'(phosphatidyl
ethanolamine) and heme and non~heme iron were devised.
While model systems do not duplicate the‘environment found
in meat, the activity of heme and non-heme iron'in these
systems may help‘to elucidate the prooxidant mechanisms

inameat.
“'Methods

Extraction'of'Muscle‘Tissue

After removing all visible fat and connective tissue,
meat samples were ground through a 1/8 in plate of a
Hobart'grinder."Weighed"sample5’of'ground.meat'were
extracted overnight at 4°C with 3 volumes of distilled,
deionized water containing a few drops of chloroform to
retard“microbial'growth.““SampleS‘were~stirred'occasionally
during extraction. The meat slurry was filtered through
cheesecloth and the residue reeextracted until it appeared
to be devoid of any myoglobin or hemoglobin pigment. Ten
gram portions of the extracted meat were placed in Kapak
bags, flushed with nitrogen gas, sealed and stored frozen
at -25°C. The samples of watereextracted'tissue were used
in the model systems devised to study the acceleration of

oxidation'in'cookedvmeat.

I I I I1 I N l Illl .II

26

Fractionation of the Water Extract

’The'water extracts were combined and freeze dried.
The concentrated extracts were dialyzed against distilled,
deionized water at 4°C, with several changes of water.
Cellulose dialysis tubing with an average pore diameter
of 48 A was obtained from the Arthur H. Thomas Company,
Philadelphia, Pennsylvania. The muscle extract was also
concentrated by ultrafiltration at 4°C, using an Amicon
Model 402 ultrafiltration cell‘CAmicon Corporation,
Lexington, Massachusetts) with a UMlO membrane; which
retains substance with a molecular weight greater than
10,000.
Preparation‘of'Model‘SystemS't0*Study
Lipid Oxidation'in'CookedEMeat

'Frozen, 10 g samples of meat were thawed, mixed with
fractions of the meat extracts, myoglobin, ferrousiron or
distilled water and heated to 70°C in‘a boiling water bath.
Samples were cooled rapidly and stored at 4°C for speci-
fied periods of time prior to analysis for TBA reactive
material.
TBA Analysis of Lipid Oxidation
‘inIMeat Systems

The distillation method of Tarladgis et a1- (1960)
was used to analyze for thiobarbituric acid reactive
material. A 10 g sample of meat was homogenized with 45
ml of distilled, deionized water for 2 min in a Waring
blender. EDTA was added at a'level of 0.2% to minimize

ironecatalyzed oxidation during the blending'process. The

I Ills] II ill I A

27

blended sample was transferred to a macro-Kjeldahl flask
by washing with 47.5 ml of distilled, deionized water.
The pH of the meat slurry was lowered to 1.5 by adding
2.5 ml of 4 N HCl. The samples were steam distilled at
the highest setting of'a macrObKjeldahl apparatus until
50 m1 of distillate was collected.

The distillate was mixed and 5"m1 were transferred
to a 50 m1 test tube. Then S‘ml of TBA reagent (0.02 M
2-thiobarbituric acid in 90% glacial acetic acid) were
added, the tubes were*stoppered'and the contents were
mixed. Samples were heated in a‘boiling water bath for
35 min. 'After cooling to room temperature, absorbance
was read at 538 nm, against a blank containing only
distilled water and TBA reagent. ‘Results are expressed

as mg malonaldehyde per 1000 g of meat.

Isolation of E 'Yolk Phosphatidyl
EthanoIamine (5%)

To study the ability of heme and non-heme iron to

 

'catalyze lipid oxidation in a less variable and complex
system than meat, model systems containing purified
phospholipids were devised.' While it is not feasible to
isolate muscle phospholipids in sufficient quantities to
use in model systems, phospholipids can be readily iso-
lated from egg yolks (Holub and Kuskis, 1969). Egg yolk
phosphatidyl ethanolamine (PE) has a relatively high pro-
portion of linoleic and arachidonic acids (Holub and

Kuskis, 1969), as do muscle phospholipids (Hornstein

28
at aZ., 1961; Giam and Dugan, 1965). Egg yolk phospha-
tidyl ethanolamine has been'observed to oxidize rapidly
in aqueous emulsions (Corliss, 1968).

Egg yolk lipids were extracted by a modification of
the method of Olivecrona and Oreland (1971). Egg yolks
and whites were separated and the membranes removed from
the yolks prior to filtration through cheesecloth. Fifty
grams of filtered yolk were homogenized with 300 m1 of
chloroform:methanol (2:1, v/v) for 3 minutes at highspeed
on a Waring blender. The homogenate was filtered'through
Whatman #1 filter paper on a Bfichner funnel. The“residue
was re-extracted with chloroform.‘ The combined extracts
were mixed with 150 ml of‘water, and the phases were allowed
to separate at 4°C. The water layer was removed by aspira-
tion, and the chloroform‘layer was dried over anhydrous
sodium sulfateprior to concentration under vacuum on a
Bfichi rotary evaporator. Final traces of the solvent were
removed under nitrogen and the residue was dissolved in
chloroform.

An aliquot was taken to determine total lipid content
before the concentrated lipids were added to cold‘(4'C)‘
acetone. 'Twenty-five grams of total lipid were made up to
50 ml volume with chloroform and added"to 700 ml of acetone.
The acetone insoluble material was redissolved in chloro-
form and reprecipitated twice. Thin—layer'chromatography
on Silica Gel G with a solvent system of chloroform:methanol:

water (65:25:4, by volume) indicated that the final

‘1 ll] Ill‘l III.\ II A
1' I'll: IIIJIIII'I'

29
precipitate contained mainly phospholipids with only a
trace of neutral lipids.

Phosphatidyl ethanolamine was isolated by thinelayer
chromatography. A 0.75 mm thick layer of Silica Gel G
was applied to methanol rinsed, 20‘x 20 cm glass plates
using a Desaga.spreader.' The-plates were air dried for
24 hr, activated for 1 hr at 105°C and cooled to room
temperature in desiccator cabinets.“ Approximately 20 mg
of lipid were streaked'on each plate, undernitrogen,
using a narrOW°tipped'pipette.' The plates were developed
in chloroform:methanol:water (65:25:4, by volume) and the
PE band was identified'by spraying one edge of the plate
with 0.25% ninhydrin in butanoi'. The" PE zone was scraped
from'each‘plate‘and'eluted'with‘SO'ml'of'chloroform:
methanol (4:1, v/v). Samples were evaporated to dryness
under nitrogen and redissolved in chloroform or“pentane.
'After centrifuging for-5 min at 1500 x g in an Inter-
national Refrigerated Centrifuge, Model PRe6, the solvent
containing PE was transferred to glass vials, which were
flushed with nitrogen gas, sealed with teflonhlined screw
caps and stored at -30°C.

Before use in model systems,the PE was checked for
purity by TLC. ‘Thin-layer'plateS'were also sprayed with
ferrous thiocyanate, which is capable of detecting low
levels of lipid peroxides (Stahl, 1969). Purified PE was
used shortly after preparation, before detectable levels

of peroxides had accumulated.

30
Preparation'of"Aqueous'DispersionS'of PB

Several authors (Corliss, 1968; Carraway and Huggins,
1972) encountered difficulty in dispersing PE in aqueous
systems.' Since the size of the dispersed lipid micelles
affects the oxidation of-lipids,‘it'is important to be
able‘tO'form'uniform,'reproducible'dispersions.

Several methods of dispersing PE in buffer*(0.05 M
borate or 0.05 M tris-maleate buffer) were used in the
model system studies. 'In one'method, an aliquot of the
solvent'containing‘a known'quantity of phosphorus was
pipetted into a small beaker and the*solvent was evaporated
under nitrogen. Buffer containing'0.25% Tween 20, a non-
ionic emulsifier, was added to the PE, and the mixture was
stirred magnetically until the PE was mixed with the buffer.
Potassium hydroxide (1.0 M) was added drapbwise, while
stirring was continued, until the PE was uniformly dis—
persed. 'Buffe ~was added to the desired volume and the
pH was adjusted with 1N HCl, if necessary. Aliquots were
then transferred to the reaction vessels.

Alternatively, an aliquot containing a known quantity
of phosphorus was pipetted directly into the flasks in which-
the study was to be conducted. The‘desired volume of
buffer containing 0.25% Tween 20 was added, and the samples
were placed in a water bath shaker at 38°C.' After about
20 min of shaking, the PE was uniformly dispersed in the‘

2

buffer. Then solutions of myoglobin or Fe+ were~added and

various'measures of lipid oxidation~were'initiated.

31

Since Tween 20 interfered with the gas chromatographic
analyses of fatty acid methyl esters, it could not be used
for dispersing PE in these studies. The PE was dispersed
in buffer by stirring, raising the pH with potassium
hydroxide then readjusting to the desired pH and volume
with buffer.' Lowering the pH to 5.5 resulted in the forma-
tion of'a‘turbid'dispersion.* However, the PE dispersion
became somewhat clearer as shaking and incubation at 38°C

proceeded.

PhosphoruS‘Determination

 

The PE content of lipid samples and model systems was
quantitated by analyzing for'phosphorus content using the
colorimetric procedure described by'Rouser et a1. (1966).

A standard curve was prepared using aliquots of monobasic
potassium phosphate solution (HartmaneLeddon'Company,
Philadelphia, Pennsylvania) containing from Zelo'ug of
phosphorus.

' Organic solvents were completely evaporated from
lipid samples prior to digestion in order to minimize the
danger of explosion (McClare, 1972). 'Samples were digested
by heating with 0.9 m1 of 72% perchloric acid for 20 min
at the lowest setting of an Aminco micro-Kjeldahl rack.
A glass fume cover connected to a water aspirator removed
fumes from the perchloric acid digest: ”After digestion was
complete, the sides of the flask were rinsed with 5 m1 of
distilled water and 1 ml of 2.5% ammonium molybdate was

added.' After swirling the flask, 1 ml of freshly prepared

32
10% ascorbic acid was added. Two milliliters of distilled
water were added, and the contents of the*digestion flask
were transferred tO‘a test tube and heated in a boiling
water bath for 5 min. After cooling the samples to room
temperature, absorbance*wa5'read at 820 nm with a Beckman
DU-Z spectrophotometer. Water was used as a blank and a
control reading was subtracted from the absorbance of the
samples. The absorbance at 820 nm was multiplied by a
factor of 11.0, which was calculated from a standard curve,

to convert the readings to pg of phosphorus.

TBA Analysis of'PE'Dispprsions

Aqueous, buffered dispersions of PE were held in a
water bath shaker maintained at 38°C.‘ At intervals, 0.5
ml aliquots were withdrawn and mixed with 1 m1 of TBA
reagent (0.02 M thiobarbituric acid in 90% glacial acetic
acid) in a l5'm1 test tube. The stoppered tubes were
heated at 100°C for 20 min.‘ After cooling‘to room
temperature, 1.5 m1 of acetone was added, the samples
were mixed thoroughly and the absorbance at 538 nm was
read against a blank containing only TBA reagent and
buffer. 'Results are expressed as absorbance'units (A)

at 538 nm.

Fluorescence of'OxidizingpPE'Dispersions

Fluorescent material was extracted from oxidizing PE
by shaking 4 m1 of the PE dispersion with 7 ml of
chloroform:methanol (2:1, v/v). Samples were then centri-

fuged for 5 min at 1500 x g in‘an International Refrigerated

33

Centrifuge, Model PR-6. The aqueous (upper) phase was
removed by aspiration‘and the lower phase was dried over
3 g of chloroform rinsed, anhydrous sodium sulfate. The
final volume of all samples was adjusted to 6.5 ml.

Fluorescence measurements were made on a 1‘ml aliquot
of the extract using an Aminco-Bowman spectrophotofluorometer
and an Aminco X-Y recorder (American Instrument Company,

Silver Springs, Maryland).

Oxygen Uptake by Oxidizing;PE'Dispersions

 

Phosphatidyl ethanolamine and buffer or the buffered

PE dispersion were"pipetted into a 15 ml respirometer flask
and a few potassium hydroxide pellets were added to the
center well. Buffer or buffer containing myoglobin or
Fe+2 were added to the side arm of the flask.‘ After 20
min of equilibration at 38°C, the contentS‘of the sidearm
was added to the reaction'mixture'and'measurement of
oxygen consumption began. The final volume of reaction
mixture in each flask was 5 m1. All experiments were
conducted at 38°C, in an air atmosphere with constant
shaking.

""Blank5'consisting‘of'buffer'alone,'buffer‘containing

2‘were run with each

. . . +
myoglob1n or buffer containing Fe
'experiment."Micrometer‘readings were*converted to moles
'dry gas at STP, and results are expressed as moles of oxygen

absorbed per mole of PE.

34

Gas Chromatggraphic Analysis of the Fatty
Acid Compos1tionofPhosphatidyl Ethanolamine

 

PE was extracted from aqueous, buffered dispersions
with chloroform:methanol in the same manner as the
fluorescent materials. The solvent was evaporated from
the PE under nitrogen: Then liml of 14% boron trifluoride
in methanol was added-under nitrogen to the lipid residue
in a 10’x 125 mm centrifuge tube.' The tubes were sealed
with teflon-lined screw caps and heated for-10 min at
100°C, as suggested by Morrison:and'Smith'(1964). After
cooling the samples to room temperature, 2 m1 of pentane
and 1 m1 of water were added to each tube. The contents
were mixed, then the samples were centrifuged at 1500 x g
for 5 min. The upper pentane layer was removed and con-
centrated under nitrogen.

Gas chromatographic analyses were performed using a
Varian Aerograph 200 chromatograph, equipped with a flame
ionization detector. A 1/8 in o.d., 7 ft stainless steel
column was packed with 15% high efficiency DEGS on acid-
washed, 80~100 mesh Chromosorb W. The column oven tempera-
ture was 180°C, the injection port was maintained at 210°C
and the detector at 195°C.- The nitrogen flow rate through
the column was 40 ml/min. Flow rates of hydrogen and air
were 20 m1/min and 250 ml/min, respectively. "Samples of
Z‘ul were injected, and the emerging peaks-were identified
by comparing retention times to those of standard mixtures
'of known fatty acid methyl'esters.' Peak areas were calcu-
lated by multiplying peak height times peak width at
half-height.

35

Gas ChromatoEraphic AnalysiS"of Volatiles
Produced by xidizing:§hospholipids

A 20 ml volume of an aqueous dispersion of phospho-
lipids, buffered at pH 7.0, or 20 ml of buffer were placed
in 50 ml Erlenmeyer' flasks, which were sealed with rubber
serum stoppers. The*samples were"incubated in a 38°C
water bath, with constant'shaking."At‘intervals, a 2.5
m1 sample of the headspace vapors above the oxidizing
lipid dispersion or the buffer control was withdrawn with
a gas-tight syringe for gas chromatographic analysis. A
standard technique for‘sampling (including flushing of the
syringe with headspace vapors) was necessary'to obtain
reproducible'results.

Separation of the headspace volatiles was accomplished
on a 1/8 in by 10 ft stainless steel column packed with
20% Carbowax 20 M on 60/80 mesh, acid washed, Chromosorb
W. A Varian Aerograph Model 200 gas chromatograph equipped
with a flame ionization detector was used in the analyses.
Column oven temperature was 70°C, the injection port was
maintained at 150°C and the detector at 140°C. Nitrogen
flow rate through the column was 15 ml/min. Flow rates
for hydrogen and air were 20 ml/min and 200 ml/min,
respectively.

Gas Chromato raphic Analysis of Volatiles
Producedmby ooked Meat

Water-extracted muscle tissue was heated to 70°C.

Tripolyphosphate was added to a 20 g portion-of the cooked

meat at a level of 0.5%. Ferrous iron (5 ppm) was added

36
to a second 20 g sample,'while the control sample contained
no added prooxidants or'inhibitors.* The samples of cooked
meat were stored in Erlenmeyer flasks sealed with rubber
serum stoppers. They were held at 4°C for 72 hours prior
to headspace sampling. 'Before sampling, the samples were
heated at 100°C for 5 minutes. 'The serum stoppers were
vented to the atmosphere during the heating period. A
2.5 ml volume of the headspace vapor was withdrawn'with
a gas-tight syringe and was analyzed in the same manner as

the headspace samples for oxidizing phospholipids.

RESULTS AND DISCUSSION

Prooxidant Activity of Muscle Extract,
Dialyzate and Diffusate

The first series of experiments was designed to
determine whether the prooxidants in muscle tissue could be
extracted with water, and if so, whether they were located
in the dialyzable or nonbdialyzable fraction of the extract.
The aqueous extract and the dialyzable and nonedialyzable
fractions of the extract were added back to the water-
extracted muscle tissue at the approximate levels present
in the unextracted meat.‘ A control sample was also pre-
pared using water~extracted’muscle plus deionized distilled
water. Samples were then heated to 70°C and~stored~at 4°C
for specified periods of time prior to TBA analysis.

The TBA values presented in Table‘l show the effect
of the muscle extract, the dialyzate and the diffusate on
the production of TBA reactive'materia1*in'thewwater-
extracted meat samples. ‘Addition of the aqueous extract
resulted in a 2.8 fold increase in the TBA values as compared
to control samples.' Apparently substances capable of
accelerating lipid oxidation in cooked meat were removed
from the muscle tissue during water extraction. ~Lipids

capable of undergoing oxidation apparently remained in the

37

38
extracted muscle tissue as indicated by an increase in the

TBA values for the control sample during storage.

Table 1. The effect of adding various fractions of aqueous
muscle extracts on TBA values of water-extracted,
heated beef muscle stored for 18 hr at 4°C

 

 

TBA Number

Sample (mg malonaldehyde/1000 g meat)
Control--extracted muscle 0.34

+ water

Extracted muscle + aqueous 0.96

extract

Extracted muscle + dialyzate 0.49

Extracted muscle + diffusate 0.54

 

Both the dialyzate (containing hemoproteins) and the
colorless diffusate were capable“of accelerating lipid
oxidation in cooked meat‘(Tab1e l).‘ Apparently, a pro-
oxidant substance in the muscle extract was diffusing
through the dialysis tubing. The initial experiments
failed to isolate the prooxidant activity in either the
dialyzate or diffusate. The prooxidant activity of the
muscle extract appeared to be approximately equally divided
between the dialyzate and the diffusate.‘ This suggests
that either a non-diffusing catalyst was present in the
dialyzate, a lowemolecular weight catalyst was partially
bound to a nonediffusable protein or else dialysis was

incomplete.

39

In subsequent experiments, dialysis was carried out
for longer periods of time. 'The diffusate was collected
and concentrated in two fractions. Each fraction consisted
of 5 volumes of distilled, deionized water containing the
material dialyzed from 1 volume of muscle extract. The
two fractions of the diffusate, as well as the aqueous
extract and the dialyzate,were added back to the water-
extracted muscle. The samples were then heated to 70°C
and stored at 4°C for 48 hr prior“to TBA analysis. Table
2 shows that both fractions of the diffusate were capable
of accelerating the oxidation of lipids in waterbextracted

cooked meat.‘ In spite'of the longer dialysis period, some

Table 2. The effect of adding various fractions of aqueous
*muscle extracts on TBA values of water-extracted
heated beef muscle stored for 48 hr at 4°C

 

 

TBA‘Number
Sample (mg malonaldehyde/1000 g meat)
Control-sextracted muscle 1.23
+ water
"Extracted muscle + aqueous 5.21
extract
Extracted muscle + dialyzate 1.66
Extracted muscle + first 1.48
diffusate fraction
Extracted muscle + second 1.40

diffusate fraction

 

40
prooxidant activity still remained in the dialyzate. The
longer dialysis period decreased the prooxidant activity
of the dialyzate relative to the combined activities of the
diffusate fractions. The combined diffusate fractions
caused a 2.4 fold increase over control TBA values° Addi-
tion of the dialyzate resulted in TBA numbers 1.4 times
controls.

TBA values have been reported to correlate with the
development of off-flavors in cooked meat products (Zipser
et aZ., 1964). A TBA value of 1-2 has been reported as the
threshold level for the detection of objectionable odors
in cooked meat products (Watts, 1962). In the present
study, the’control samples developed offeflavors very
slowly. ‘A comparison of the data in Tables 1 and 2 shows
an increase of 0.89 TBA units for the control samples when
the'length"of‘storage“wa5’increased from 18-48 hours. In
contrast, addition of the aqueous extract caused rapid
increases in TBA numbers and the development of rancid
odors. Increasing the storage time from 18 to 48 hr (Tables
1 and 2) resulted in an increase of 4.25 in the TBA numbers
of the water-extracted sample plus the aqueous extract.

The data in Figure 1 show that TBA values increased as
the amount of muscle extract, dialyzate or diffusate added
'to the sample was increased. The concentrated extract or
fractions added in volumes ranging from 1-5 ml produced TBA

values that were greater than control values in all cases.

41

0.4 '

0.2 r

TBA Number (mg malonaldehyde/1000 g meat)

 

1 l l

4.0 5.0

L

1.0 21.0 3.0
Amount of Additive (m1)

Figure 1. The effect of increasing levels of extract,
dizrlyzate and diffusate on the TBA numbers of water-extracted
beef stored for 18 hr at 4°C after heating to 70°C. A =
water-extracted muscle plus aqueous extract; B = water-
eXIrracted muscle plus diffusate; C - water-extracted muscle
Pltls dialyzate.

42

Lewis and Wills (1963) reported that the hemes in
tissue homogenates were effective catalysts of the oxida-
tion of added unsaturated fatty acids in dilute (1.0%, w/v)
suspensions. At higher concentrations (5%, w/v), they
reported that the tissue homogenates had an inhibitory
effect. If hemoproteins served as active prooxidants in
the muscle extract or dialyzate in the current study, no
inhibitory effect was apparent. Either an inhibiting con-
centration of hemOproteins was not reached or the heme
compounds did not influence the extent of lipid oxidation.

Prooxidant Activity of Myo lobin,
_Ferrous Iron and Ascorbic Acid

 

In the initial experiments, it was established that
prooxidant substances in muscle tissue could be extracted
with water, and that a significant amount of the prooxidant
activity could be removed from the aqueous muscle extract
by dialysis. This would indicate, contrary to traditional
views, that non-heme iron may be a more important catalyst
of lipid oxidation in cooked meat than the denatured muscle
pigments. To test the hypothesis that non-heme iron rather
than hemoproteins was the prooxidant substance in the muscle
extract, a series of experiments with model systems was
designed.

Myoglobin was added to water-extracted muscle at levels
ranging from 1.0-10.0 mg/g. The samples were heated to
70°C and stored at 4°C for 18 hr prior to TBA analysis.

The results shown in Figure 2 demonstrated that addition of

myoglobin to beef muscle did not increase TBA values.

43

5.0 F

4.0 ~

3.0 r

2.0 -

 

TBA Number (mg malonaldehyde/1000 g meat)

 

 

Myoglobin (mg/g)

Figure 2. The effect of myoglobin on the TBA numbers
of cooked, water-extracted beef stored for 24 hr at 4°C.

r- : £ a
1.0 e
1 I I L I 1 1 1 'L in
2.0 4.0 6.0 8.0 10.0

44
Ferrous iron was added to samples of water-extracted
muscle at levels ranging from 1-5 ppm. The samples were
treated in the same manner as those containing purified
myoglobin. 'As shown in Figure 3, the TBA numbers of the
water-extracted muscle“tissue showed a linear increase as

the concentration of Fe+2

2

increased from 1-5 ppm. When

both 1 ppm Fe+ and 5 mg/g of myoglobin were added to a

sample of water-extracted muscle tissue, the TBA values were

not significantly greater than those for'samples containing

2

only 1 ppm Fe+ (Table 3) suggesting that addition of

myoglobin did not catalyze lipid oxidation.

Table 3. The effect of myoglobin, ferrous iron and myo-
globin plus ferrous iron'on TBA numbers of
cooked, watereextracted beef stored for 48 hr

 

 

at 4°C
TBA Number

Sample (mg malonaldehyde/1000 g meat)
Control--extracted meat 2.10

Extracted meat + 1 ppm Fe+2 2.50

Extracted meat + S mg/g 1.97

myoglobin

Extracted meat + 1 ppm Fe+2 2.62

+ 5 mg/g myoglobin

 

The levels of myoglobin and ferrous iron were chosen
to approximate the concentrations of heme proteins and non-
heme iron reported to occur in beef muscle. ’Myoglobin, the

oxygenbbinding'muscle pigment, and hemoglobin, from blood

45

5.0

4.0

3.0

2.0

 

1.0

TBA Number (mg malonaldehyde/1000 g meat)

I l A l l L l l J I

1.0 2.0 3.0 4.0 5.0
Fe+2 (ppm)

 

Figure 3. The effect of Fe+2 on the TBA numbers of
cooked, water-extracted beef stored for 24 hr at 4°C.

 

46
trapped in muscle tissues, are the major sources of iron
in muscle. Craig et a2. (1966) reported that the total
heme pigment content in the Zongiaaimua dorai muscle of
beef animals is about 3.79 mg/g. In beef foreshank, these
authors reported the‘myoglobin and hemoglobin content to
be 4.25 mg/g.

In undenatured muscle, myoglobin'iS*in solution in the
cytoplasm.' The highly unsaturated phospholipids are located
in the membranes of cellular structures. 'Thus, the-contact
between myoglobin and the unsaturated fatty acids of the
phospholipids may be limited by cellular structure.

It is known that when the concentration‘of heme to
polyunsaturated fatty acids at'a*particular site exceeds a
fixed ratio, the heme acts as an antioxidant“rather than
a prooxidant (Lewis and Wills, 1963; Kendrick and Watts,
1969). Thus, it is possible that the soluble hemoproteins
may be out of contact with unsaturated lipids in some
portions of the cell and be present in inhibiting concen—
'trations in other areas.

Sato and Hegarty (1971) reported that“myoglobin did
not accelerate lipid oxidation when it was added to water-
‘extracted muscle tissue at'a level of 5 mg/g. This level
'of myoglobin approximates the total amount of hemoprotein
present in muscle tissue. ‘However, grinding and extracting
muscle tissue may alter the structure so that added myo-
globin does not make the same degree of contact with the
lipids as in unextracted tissue. Therefore, it was decided

to examine the effect of both high and low levels of

”' 47
myoglobin on production'of TBA reactive'materials in muscle
tissue. ‘The data-presented in‘Figure’Z and Table 3 support
the hypothesis of Sato and Hegarty (1971) that*myoglobin
does not cause the oxidation of lipids in cooked meat, while
non-heme iron is an"effective catalyst of lipid oxidation,
even at very low concentrations.

"Ledward"(1971) stated that‘on“heating, hemoproteins a:
in'muscle'tissue'are'denatured'and'form'aggregateS'with
other'denatured“muscle'proteins.“Denaturation*of~hemo—
proteins has been‘reported to'increase their prooxidant

activity (Banks, 1961; Eriksson et a2., 1971). Banks (1961)

‘0.

 

proposed that undenatured'cytochrome-c*may'actually prolong
the induction period during oxidation-of‘unsaturated fat.
He suggested that undenatured'hemoproteins may break down
hydroperoxides so that they are not capable of initiating
oxidation. 'He advanced the hypothesis that a denatured
form of the heme compound accelerates oxidation following
'the"induction'period.

'Eriksson et a2. (1971) reported that heating the
'hemoproteins, catalase and peroxidase, increased their
efficiency as heme catalysts of linoleic acid oxidation.
Spectral*analysis indicated that there was no change in
the oxidation state~of‘the hemoprotein iron. The authors
attributed the observed increase in prooxidant’activity to
a possible unfolding of the protein to cause a greater
exposure of heme groups to the substrate. Koch (1962)
cited Bishov and Henick (unpublished data)*a5'showing that

heat denaturation of heme pigments reduced their prooxidant

48
activity. 'They observed that cooked dehydrated meat was
more stable than raw dehydrated‘meat. Lipid oxidation rates
in model systems containing heated*muscle myoglobin extracts
were also reduced.

While denaturation of myoglobin makes the heme group
more available, apparently the exposed heme group has an
affinity for other proteins. 'If other proteins are present,
as would be the case in meat or meat extracts, other
denatured proteinS‘apparently'coeprecipitate with myoglobin
to form denatured hemOprotein aggregates.' This could
decrease the exposure of the heme groups and thus decrease
their prooxidant‘activity. ‘Another possible explanation'
for thereducedprooxidant activity of denatured hemo-
proteins'iS'the'observationfby’Ledward~(197l) that the
ferric hematin in cooked meat possesses low—spin character-
istics.' Generally, ironeporphyrins which have low spin
stateS'are poor catalysts of lipid oxidation.

"'Whil 'most of the iron in animal tissue occurs in
hemoproteins, a number of substanceS'in'the"anima1 body
contain non-heme iron. Iron-protein'complexes (ferritin,
'hemosiderin'and'transferrin)'function’in'the storage and
transport of noneheme iron. Small‘amountS"of non-heme
irbn proteins appearto perform key functions*in electron
transport (Mahler and Cordes, 1966). Enzymes such as
'succinic dehydrogenase, DPNH-cytochrome reductase and
xanthine"oxidase'contain'noneheme iron. ‘Wills (1966) has
demonstrated that when non-heme iron'is released from

ferritin, it becomes an active catalyst of lipid oxidation.

49
Ascorbic acid can function in the'release of'noneheme iron
from iron-containing proteins (Wills, 1966). This may
partially explain why low levels of ascorbic acid enhance
tissue lipid'oxidation.

'Sato and Hegarty (1971)-reported that'the‘non~heme
iron content of beef muscle'waS'about 1 ppm. The data in
Figure 3 indicate’that'very low levels of'ferrous iron
can-result'in'significant“acceleration“of'lipid‘oxidation.
The non-heme iron containing proteins, which function in
electron transport, would be located in close proximity to
the phospholipids in the mitochondrial membranes.’ Heat
denaturation of muscle tissue may cause“cellular"disorgani-
zation and thereby release nonhheme iron from proteins.
Degradation'of hemoproteins during cooking and storage
could also make non-heme iron available. “In the present
study TBA values in cooked, watereextracted meat samples
were about twice the values observed for unheated, water-
extracted samples of muscle stored for 24 hours. 'Several
authors have noted that a more rapid increase in TBA values
takes place in heatbtreated'muscle'than in'unheated muscle
tissue (Timms and Watts, 1958; Sato and Hegarty, 1971).
This suggests that heating decreases the stability of the
muscle phospholipids, perhaps by increasing the activity
of muscle catalysts of lipid oxidation.

The studies reported previously were carried out with
eye of round (aemitendinosus) muscle from beef. Similar
results were obtained with water-extracted pork muscle.~

Ferrous iron at a level of 1 ppm was observed to increase

50
the production of'TBA reactive'material'relative't6'con~
trol values (Table 4). ‘Ascorbic acid (5 ppm) increased
the prooxidant activity of added Fe+2. The addition of
5 mg/g of myoglobin had no effect upon the’production of

TBA reactive material.

Table 4. “The effect of*myoglobin, ferrous iron and
‘ ‘ferrous iron plus ascorbic acid*on TBA numbers
'of cooked,water~extracted pork samples stored
for 24 hr at 4°C

 

 

, TBA Number
Sample (mg malonaldehyde/1000 g meat)
Controle‘watereextracted 0.80
muscle
Water-extracted muscle + 1.16
1 ppm Fe*2

Watereextracted"musc1e +
1 ppm Fe*2 + 5 ppm 1.61
ascorbic acid

‘Watereextracted muscle + 0.77
5 mg/g myoglobin

 

'A typical pattern of change in TBA values over a 3-day
periodWis shown in Figure 4. A combination of ferrous iron
and ascorbic acid resulted in*the'most"rapid rate of
‘oxidation, particularly during the'first'24'hr period.
Ferrous iron alone accelerated oxidation to a lesser extent

2

than a combination of Fe‘+ and ascorbic acid. ‘Myoglobin.

failed to accelerate oxidation of the lipids in"cooked pork.

51

 

205 P
f'"
g
4)
E P
no 2.0
O
O
O
H
\
Q)
'3.
I: 1.5
Q)
“U
H
(d
I:
O
H
(U
a 1.0 .
OD
E
U
In
0)
'3
5 005 f
2
a?
H

.I .l |

24 48 72
Time (Hours)

Figure 4. Changes in TBA numbers of water-extracted,
cooked pork loin m scle stored at 4°C. A = water-extracted
muscle + 1 ppm Fe+ + 5 ppm ascorbic acid; B = water
extracted muscle + 1 ppm Fe+2; C - water-extracted muscle
+ distilled water; D 8 water-extracted muscle + 5 mg/g
myoglobin.

52

MuscleS'classified as darkecolored have been reported
to contain more highly unsaturated‘phospholipids than light-
colored muscles (Luddy et aZ., 1970). Thus, dark-colored
muscles would be expected to undergo more extensive lipid
oxidation than light muscles. ‘During the course of the
present study TBA values of light and dark beef and pork
muscles, which“had been water-extracted, cooked and stored
at 4°C, were'compared.’ Dark muscles had about"l.7 times
higher TBA values than light'muscles.‘ Although only a
limited number'of samples were used,‘results indicate
that darkhcolored muscles undergo more lipid oxidation
than light muscles. 'This*could be of practical signifi-
cance, since the proportions of light and dark muscle
incorporated into precooked, processed products may
'influence the extent of lipid oxidation.

Oxidation of Aqueous, Buffered Dispersions
'of_Phosphatidyl Ethanolamine

The studies discussed in the previous section showed
that myoglobin failed to accelerate lipid oxidation in
cooked beef or pork muscle, even though catalysis of
lipid oxidation by home compounds iS'a generally accepted
phenomenon. To explorethe effect of myoglobin on lipid
oxidation in a less complex and variable environment than
meat, model systemS'containing'purified phospholipids
were devised. Younathan and Watts (1960) have shown that
phospholipids are the class of lipids most rapidly oxidized
in cooked meat. Thus, phospholipid (phosphatidyl ethanolamine)

was used in the model systems. The ability of non-heme iron

53
and-myoglobin:to*accelerate‘the'oxidation"of‘phospholipids
in aqueous, buffered dispersions was measured by a variety

of methods.

 

'Oxygen Uptake by Aqueous,
u ere E Dispersions

As shown in Table 5, increasing the PE concentration
in the buffered dispersions resulted in an increase in
oxygen consumption. “At*all except the“lowest PE concentra-
tion studied, the moleS'of"oxygen consumed per mole of PE
were relatively constant. In subsequent experiments, a PE

4

concentration of'either 2.5 or 3.0 x 10’ M was used, since

 

the rate of oxygen uptake was readily measured and repro—

ducihle at these levels.

Table 5. The effect of PE concentration on'oxygen con-
sumption during 10 hr oxidation~at 38 C

 

 

PE Concentrationa Moles O2 Consumedb Moles OZ/Mole PE
1.23 x 10’4 M 0.7 1.2
2.46 x 10" M 2.0 1.6
4.00 x 10‘4 M 2.9 1.5
5.0 x 10" M 3.7 1.7

 

aPE was dispersed in 0.05 M borate buffer, pH 7.0,
containing 0.25% Tween 20.

bOxygen uptake was measured in a Differential
Respirometer (Model GR14, Gilson Medical Electronics,
Middleton, Wisconsin).

54

The PE dispersions were buffered at pH 5.5 and 7.0
with 0.05 M tris-maleate'or borate buffer. A pH value of
5.5 is in the pH range where noneheme iron exerts maximum.
prooxidant activity (Wills, 1965)- Undenatured meat would
normally have-a pH value close to 5.5; however, heating
causes a slight increase in the pH of muscle (Bendall and
Wismir-Pedersen, 1962).‘ While the pH of meat would not
normally be as high as 7.0, PE has been-reported to oxidize
most rapidly at a pH between 7.0 and 8.0 (Corliss, 1968).
Myoglobin has been reported to be an active catalyst of
lipid oxidation at both acid and slightly alkaline pH
values (O‘Brien, 1969; Liu, 1970).

The rate of oxygen consumption by PE dispersions over
a 24 hr period is shown in Figure 5. *PE underwent immediate'
and rapid oxidation at pH 7.0. At pH 5.5 the initial rate
of oxygen consumption was much slower.

The effect of ferrous iron on the oxidation of an
aqueous, buffered PE'dispersion*is shown‘in Figure 6.

4 M Fe+2

At pH 5.5, 1.5 x 10' increased the rate of oxygen
uptake. ‘When'the same concentration of Fe+2 was added to
a PE dispersion buffered at pH 7.0, a slightly inhibitory
effect'on“oxygen'consumption'waS‘observed.

Figure 7 shows that metmyoglobin at a concentration of
5.0 x 10'6 M also increased the rate of oxidation of a
PE dispersion atpH 5.5. However, at pH 7.0 the same
concentration of metmyoglobin was observed to have a
slightly depressing effect on the oxygen consumption of

PE dispersions. Lower metmyoglobin concentrations

55

.newmcoumaz .ceuoaevwz .muacouuooam

Heuwuoz cowawo .vamo Hovozv nouoaoufinmom Hewucouommfin e cw renameoa

ozone: nomxxo .om noose amm.o ucfiaweuuou .m.m mm .hommsn oueoaea-mwpu

z mo.o ed vomuommwu mm 2 oaoH x o.m u m “cm noose wm~.o mcfiufleucoo

.o.a ma .uoemsa ooooHaE-meo z mo.o eh eomuoamhe ma 2 e-oa x o.n u <
.u.mm we :ofimuommfiv mm muwuwvwxo an we oxeuns comxxo .m ousmfim

Amuaomv mega
aw Nu om ma. 0H ea NH ca a o e N

.1 A A 1 q a

d
q
-

 

(3d atom/lo setom) exasdn ueBKxg

56

3.0 P
E? A
De
,3
g B
\\ 2.0 ’
N
O
3 c
H
0
£3
0
M
S l 0
g e
c:
8.
>~
g _
/D

 

 

f 4‘ I I I I L l L

2 4 e e l 10 12
Time (Hours)

Figure 6. Oxygen uptake of an oxidizing PE dispersion
containing Pe+2 at 38°C. A - 2.5 x 10-4 M.PE dispersed in
0.05 M tris-maleate buffer containing 0.25% Tween 20 at pH
7.0; B - conditions same as A, 1.5 x 10'4 M Fe+2 added;

0 - 2.5 x 10'4 M PB and 1.5 x 10-4 M Fe+2 dispersed in 0.05
M tris-maleate buffer, pH 5.5, containing 0.25% Tween 20;
D - conditions same as C, no Fe+2.

57

 

 

300 -
L.
E?
a.
o A
v; B
43 2.0 E
N
O
In
0
H
0
£3
0
0
.2
I!
3. 1.0 e
:3
t:
0
en
.>~
a + D
2 4 6 8 10 12

Time (Hours)

Figure 7. Oxygen uptake of an oxidizing PE dispersion
containing metmyoglobin (MMb) at 38°C. A a 2.5 x 10‘4'M
PE in 0.05 M tris-maleate buffer, pH 7.0, containing 0.25%
Tween 20; B a same conditions as A + 5.0 x 10“5 M MMb; C 8
2.5 x 10-4 M pa and 5.0 x 10-6 M MMb in 0.05 M tris-maleate
buffer, pH 5.5, containing 0.25% Tween 20; D a same condi-
tions as C, except no MMb.

S8

6 M or 4.0 x 10'7 M) did not

M, 1.0 x 10'°

(2.5 x 10‘
accelerate oxygen uptake at“pH'7.0.

Apparently both metmyoglobin and ferrous iron are
capable of accelerating oxygen uptake by a dispersion of
PE buffered at pH 5.5. AtpH 7.0, neither‘metmyoglobin
nor ferrous iron increased the rate of oxygen uptake.
Ferrous iron has been reported to have maximum catalytic
activity at pH 5.0 to 5.5'(Wills, 1965) and to cease
functioning as a prooxidant above pH 6.4 (Liu, 1970).
Myoglobin catalysis has been reported to be less pH
sensitive than'noneheme iron catalysis (Wills, 1965).

The metmyoglobin catalysis at‘pH 5.5 noted in the present
study may be due to some unfolding of the myoglobin struc-
ture. This would result in greater exposure of the heme
group, thus accelerating oxidation. At pH 7.0, the heme

group may be less available for interaction with the

phospholipid'fatty'acids.

TBA'ValueS'of'OxidizingpAqueous Dispersions of PE
Table 6 shows the absorbance due to TBA reactive
material measured after 24 hr of oxidation at 38°C.. Higher
absorbance values were observed for the PE dispersions
buffered at pH 7.0 than for those at pH 5.5. Ferrous iron
accelerated production of TBA reactants at pH 5.5. The

2

TBA values for the samples-with Fe+ were 1.4 times greater

than control values. At pH 7.0, the incorporation of Fe+2
in PE dispersions resulted in‘a 1.2 fold increase in TBA

values. Ascorbic acid plus Fe+2 increased TBA values to

59
Table 6. The effect of Fe+2, metmyoglobin and ascorbic

acid on the productionof TBA reactants in PE
dispersionsheld at 38°C for 24 hr

J—

 

 

 

A (538-nm)

Sample pH 5.5 pH 7.0
Control--PE alonea 0.37 0.46

+213
PE + Fe 0.53 0.54
PE + Fe+2 and ascorbic acidC 0.74 0.72
PE + metmyoglobind 0.45 0.48

85.0 x 10"4 M PB dispersed in 0.05 M borate buffer.

bFe+2--l.5 x 10'4 M.

cAscorbic acid--l.0 x 10'4 M.

dMetmyoglobin--1.0x 10'6 M.

twice the control level at pH 5.5. At pH 7.0, the addition

of Fe+2

plus ascorbic acid resulted in‘a 1.6 fold increase
in the production of TBA reactants- Myoglobin increased
TBA values slightly at pH 5.5; however, at pH 7.0, no
increase in TBA absorbance was noted due to the addition
of myoglobin.
With one exception, the TBA values reported in Table
6 are in agreement with oxygen uptake data presented in
the previous section. Ferrous iron did not accelerate oxygen
uptake at pH 7.0 (Figure 6); however, an increase-in TBA

2

absorbance was noted for the PE samples containing Fe+ and

+2

Fe plus ascorbic acid (Table 6). Non-heme iron and

ascorbic acid have been reported to cause high TBA-values,

60
which do not reflect the true extent of oxidation occurring
in a sample (Wills, 1964; Castell at aZ., 1966; Castell and
Spears, 1968). Thus, the TBA values measured in the samples
containing iron and ascorbic acid may reflect interference

in the TBA test.

Fluorescence of Oxidized Dispersions of PE

Aqueous dispersions of PE developed a yellow-brown
discoloration after several hours of oxygen uptake. The
extent of browning-appeared to be related to the amount of
oxygen consumed by the dispersed PE. Lea (1957) and Corliss
(1968) have noted the rapid development of brown discolora-
tions in oxidizing PE.

The concentration-of free amino groups has also been
shown to decrease during PE oxidation (Lea, 1957; Corliss,
1968). Mono- and dicarbonyls, which are secondary degra-
dation products of peroxides produced by lipid oxidation,
are known to condense with primary amines to yield conjugated
chromophoric Schiff-base systems (Fletcher-and Tappel,
1970). The fluorescence spectra of Schiffbbase compounds
formed by condensation of carbonyls and amino groups have
been reported by Fletcher and Tappel (1970).

Measurement of fluorescent products could provide
another method for evaluating the extent of oxidation in
PE dispersions.' Consequently, the fluorescence of oxidiz-

ing PE dispersions was measured.

61

In'the‘present'investigation, the oxidizing aqueous
dispersions of PE contained a fluorescent substance, which
could be extracted with chloroform. The excitation and
emission maxima*for'the chloroform"extract‘of oxidized PE
are shown in Figure 8. The wavelength for maximum excita—
tion was 360 nm.* Maximum fluorescent emission was at 430
nm. The fluorescing compounds were not further characterized;
however, their fluorescent excitation and emission maxima
correspond to those reported for Schiffebase compounds formed
by carbonyl-amine condensation (Fletcher and Tappel, 1970).

Figure 9 illustrates the development of fluorescent
'material in oxidizing PE dispersions. 'At‘pH 7.0, fluorescent
intensity increased slowly during the first 6 hr of oxida-
tion. *As shown in Figure 5, a very rapid rate of oxygen
uptake was noted during this time period. ‘From‘6-12 hours,
the fluorescent intensity of oxidizing PE dispersions
increased rapidly (Figure 9). “Apparently "the maximum
‘fluorescent intensity was reached after 12 hours, and by
24 hours the relative intensity had decreased. At pH 5.5,
where oxygen uptake was less rapid, less fluorescent
material was formed. The production of fluorescent material
at pH 5.5 increased rapidly between 6 and 12 hr. However,
the relative intensity after 24 hr of oxidation at pH 5.5
was slightly greater than the value at 12 hr.

4M

Table 7 shows that incorporation of 1.5 x 10'
Fe+2 into the PE dispersions increased the production of

fluorescent material at pH 5.5. At pH 7.0, however, a

62

 

 

Z: 100
p
0H
(n
I:
330.
.5 B
d.)
:6!-
“0
c6
3 A
b
407
0
U
c:
0
§20-
H
O
3.
no I J_
200 300 400 500 600

Wavelength (nm)

Figure 8. Excitation and emission curves for the
chloroform extract of an oxidizing PE dispersion. A =
excitation spectrum, B = emission spectrum.

63

100

60

40—

20 '

Fluorescence (relative intensity)

 

A 6 12 18 ' 24
Time (Hours)

Figure 9. Fluorescent intensity in a PE dispersion as
a function of hours of oxidation at 38°C. A = PB dispersed
in 0.05 M tris-maleate buffer, pH 7.0, containing 0.25%
Tween 20; B = Same conditions as A, except pH of 5.5.

64

decrease in the fluorescent intensity was noted when Fe+2

was added to the oxidizing PE dispersions.

Table 7. The change in fluorescence of an oxidizing PE
dispersion due to the addition of‘Fe+2 or

 

 

 

 

‘metmyoglobin
Increase over PE alone
Sample pH 5.5 pH 7.0
a
PE + Fe+2 2.7x 0.73x
PE + metmyoglobinb 2.8x 1.9x
33.0 x 10.4 M PB and 1.5 x 10.4 M Fe+2 dispersed in

0.05 M tris-maleate buffer containing 0.25% Tween 20.
b3.0 x 10’4 M'PE and 2.5 x 10‘° M metmyoglobin

dispersed in 0.05 M tris-maleate buffer containing 0.25%
Tween 20.

These results reflect the effect of Fe+2 on oxygen
consumption in the PE dispersions. At pH 5.5, Fe+2 increased
oxygen uptake (Figure 6), apparently resulting in‘a greater
concentration of carbonyl reactants, as shown by the increase
in fluorescence (Table 7). Ferrous iron decreased oxygen
consumption at pH 7.0 (Figure 6) and a corresponding decrease
in'fluorescent7materia15'was noted (Table 7).

Metmyoglobin increased the production of fluorescent
material at pH 5.5 (Table 7). Again, this presumably
reflects the greater avilability of carbonyl reactants due
to the prooxidant activity of myoglobin. At pH 7.0, addi-
tion of metmyoglobin to PE dispersions also resulted in

an increase in the production of fluorescent materials,

65
even though metmyoglobin failed to increase oxygen uptake
at pH 7.0 (Figure 7).

No fluorescence was present in blanks containing only
buffer and metmyoglobin or ferrous iron. The aqueous phase
remaining after the PE dispersion had been extracted with
chloroform also failed to fluoresce. ’It was also noted
that the myoglobin which precipitated during the chloroform
extraction and centrifugation was bright red in color at
pH 7.0 and grayish~brown at pH 5.5. The visible spectra of
PE dispersions containing metmyoglobin after 24 hr of oxi-
dation at pH 7.0 were very similar to the spectra of freshly
prepared samples or of buffer and metmyoglobin which had
been incubated for 24 hr at 38°C.' After 24 hr of oxidation
at 38°C, the PE dispersions containing myoglobin at pH
5.5 failed to show the peaks which are characteristic of
metmyoglobin.' The spectra of control samples containing pH
5.5 buffer and metmyoglobin were not changed by 24 hr of
incubation at 38°C.

Apparently the heme moiety of metmyoglobin is capable
of catalyzing phospholipid oxidation at pH 5.5 and is
degraded as a result of lipid oxidation. At pH 7.0, lipid
oxidation occurs rapidly in PE dispersions, even though
metmyoglobin7is not capable of accelerating lipid oxidation
at this pH. The configuration of the apomyoglobin at pH
7.0 apparently shields the heme group from destruction by

lipid oxidation.

66

Kendrick and Watts (1969) and Nakamura and Nishida
(1971) have reported that the spectra of hemoproteins are
altered when inhibiting concentrations of hemoproteins are
in contact with unsaturated fatty acids.‘ They have reported
that peaks at 535 and 560-563 nm are characteristic of the
inhibitory complexes. ‘This spectral shift was not noted
in the current study in any of the PE dispersions containing
metmyoglobin. However, the spectral shift was noted when
metmyoglobin was dissolved in‘a buffer containing 0.002 M
sodium lauryl sulfate. A corresponding change from brown
to red was apparent in the metmyoglobin solution. The
ability of these solutions to accelerate or inhibit lipid

oxidation was not evaluated.

Fatty Acid Composition'of Fresh and Oxidized PB

The fatty acid composition offreshly prepared egg
yolk phosphatidyl ethanolamine is shown in Table 8. Values
reported by Hornstein et a2. (1961) for beef and pork
phospholipids are also given for comparison. The egg yolk
PE contained 18.4% arachidonic acid, which is similar to
the level reported for muscle phospholipids. While egg
yolk PE contained less linoleic acid (18:2) than muscle
phospholipids, it also contained 11.3% of long chain fatty
acids, tentatively identified as 22:3 and 22:6 fatty acids.
It should be noted that these*fatty acids were extremely
labile and care in extracting and isolating PB was neces-

sary in order to obtain the reported levels.

67

Table 8. 'The fatty acid composition of PE from egg yolk
and pork and beef muscle phospholipids

 

% Composition

 

 

. a Beef . . b Pork . . b
Fatty ac1d Egg yolk PE phospholipids phospholipids

16:0 15.2 13.2 20.0

16:1 Trace 2.2 2.3

18:0 25.4 15.6 11.0

18:1 19.4 21.2 16.2

18:2 10.1 20.2 27.9

18:3 Trace 1.8 1.0

20:4 18.4 19.2 16.3

22:3 + 11.3 - -

22:6

 

aPeak area percent.

bHornstein at al. (1961).

While the fatty acid composition of the purified egg
yolk PE and muscle PE are not identical, the values reported
in Table 8 indicate that both contain large quantities of
fatty acids which would be susceptible to oxidation. Thus,
purified muscle PE might be expected to behave similarly
to isolated egg yolk PE in model systems.

Table 9 shows the loss in unsaturated fatty acids in
PE dispersions after 24 hr of oxidation at 38°C. Since
loss of unsaturated fatty acids results in increased per-
centages<xfthe remaining fatty acids, results are expressed

as the ratio of unsaturated fatty acids to saturated fatty

68
acids. Thus, the lower the ratio, the greater the degree

of oxidation'occurring in the sample.

Table 9. Ratio of unsaturated fatty acids to*saturated
fatty acids in fresh PE and PE held for 24 hr

 

 

 

 

at 38°C ‘
Ratio of unsaturated fatty acids
to saturated fatty acids
Sample pH 5.5 pH 7.0
Freshly prepared PEa-- 1.60 1.60
no storage
PE held 24 hr at’38°ca 0.80 0.65
pa +bFe+2 held 24 hr at 0.32 0.50
38°C
PE +ch held 24 hr at 0.72 0.69
38°C
83.0 x 10'4 M PE dispersed in 0.05 M tris-maleate
buffer.
b3.0 x 10"4 M'PE and 1.5 x 10‘4 M Fe+2 dispersed in
0.05 M tris-maleate buffer.
C3.0x 10'4 M PE and 2.5 x 10"6 M metmyoglobin

dispersed in 0.05 M trisemaleate buffer.

The destruction-of unsaturated fatty acids was more
rapid at pH 7.0 than at pH 5.5 (Table 9). The most highly
unsaturated fatty acids in PE were oxidized most rapidly.
The 022:6’ C22:3 and C20:4 fatty acids were completely
degraded-during 24 hr of oxidation in the samples containing
Fe+2. Although extensive degradation of C18:2 occurred,
residual levels of this fatty acid were observed in all

samples.

69

Z

The addition of Fe+ "resulted in extensive fatty acid

'degradation at pH 5.5 (Table 9). At pH 7.0, however, the

prooxidant effect of Fe+2

was less pronounced. When met-
myoglobin was added to the PE dispersion at pH 5.5, the

loss of unsaturated fatty acids was also accelerated.
Metmyoglobin did not increase the'degradation of unsaturated
fatty acids at pH 7.0.

'VolatileS'Produced’by'Oxidizing
Phospholipid'fiispersions

 

During the previous experiments, a characteristic odor
was noted in the oxidized PE dispersions. In an attempt to
isolate and identify some'of the odorous components, the
volatiles from'the headspace of an’aqueous dispersion of
phospholipids were'analyzed by gas~1iquid chromatography.

A typical chromatogram of the headspace volatiles from
oxidized phospholipidsis shown in Figure 10. One of the
major peaks (peak 5) had a retention time identical to
n-hexanal. The odor of the column effluent from peak 5

was the same as that of standard nhhexanal. ‘GaddiS'et a2.
(1961) have reported that n-hexanal is a product of the
autoxidation of linoleate, and Evans (1961) has proposed

a mechanism to explain its formation.” Peak 7 had the same
retention time as heptaldehyde. The retention time of peak

9 was identical to a nonanal standard.

Volatiles Produced by Oxidized Meat
The odors of the volatiles in the headspace above the

cooked meat samples stored for 72 hr at 4°C was noted.

70

 

 

 

 

 

 

 

 

 

 

1,2 5
i" '
O
m
:3
O
a.
V8
0
a:
H
«‘3 3
F0
0
U
0
a 1 6
4
8
\J
7 9 10
\Or ,/\ n

 

 

Time (Minutes)

Figure 10. Gas-liquid chromatogram of headspace
volatiles from oxidized phospholipid dispersions.

71

The control sample and the sample containing tripoly-
phosphate had a pleasant, "boiled-meat" aroma, whereas

those containing 5 ppm Fe+2

were distinctly rancid. Samples
of the headspace vapors above these samples were analyzed
by gas-liquid chromatography.'"A'typical“chromatogram of
the volatiles produced in meat'samples containing Fe+2 is
shown in Figure 11. Peak 4 had a retention time identical
to that of n-hexanal.* The odor~of the column~effluent
from this peak wasthe same-as the hexanal standard. Peak
5 and peak 7 had retention timeS‘identical‘to those of
heptaldehyde and nonanal. The identity ofthese peaks was
not confirmed by other~means.‘ However, all of these com-
pounds have been reported to occur in heated beef (Herz
and Chang, 1970). 'Severa 'of the‘unidentified peaks had
characteristicrodors;for'example,*peak5'l'and 2 had an
odor like sulfurbcontaining compounds.

"ChromatogramS‘of the headspace vapors above control
meat samples and the'samp1e57containing tripolyphosphate
were similar to the chromatogram shown in Figure 11. Dif-
ferences were noted in the relative areas of many of the
peaks. This was especially noticeable in the case of the
n-hexanal peak, which was apparently related to the degree
of.oxidation occurring in the samples. The hexanal peak
was smallest in the samples containing tripolyphosphate.
The control sample, which did not contain~either anti- or
prooxidants, had a hexanal peakstwice the size of that in
the tripolyphosphate sample. The addition of 5 ppm Fe+2

resulted in a hexanal peak that was 4.5 times that of the

72

 

 

 

Recorder Response

 

 

 

 

 

 

 

 

 

N MRI.” 6 E s

Time (Minutes)

 

Figure 11. Gas-liquid chromatogram of headspace
volatiles from oxidized meat containing 5 ppm Fe+ .

73

sample with tripolyphosphate.‘ The area of several other
peaks, including the peaks with an odor like sulfur-
containing compounds, increased somewhat—in the sample
with Fe+2. A variety of these compounds undoubtedly
contribute to the characteristic stale flavors of cooked,
stored meat.

A comparison of the chromatograms in Figures 10 and
11 indicated that many of the same volatile compounds were
present in“the headspace above oxidizing PE dispersions
and above oxidized meat. ‘Hexana 'appeared"t0'increase as
a result of lipid oxidation in both PE and cooked meat.

El-Gharbawi and Dugan (1965) reported that n-hexanal

increased greatly during the storage of freezeedried beef.

Cross and Ziegler'(l965) also~found that”hexanal was present

in greater quantities in uncured than'cured meat, presumably

due to greater lipid oxidation in the uncured meat.
Hexanal is obviously only one of the components
‘associated with the characteristic odor of stored, cooked

meat products.' The quantities of hexanal present indicate

that it may make a significant contribution to stale flavors

in cooked meat. 'Further isolation and identification of the

volatiles produced by oxidized meat is needed for a better
understanding of the formation of undesirable odors in

‘cooked, stored meat products.

SUMMARY AND CONCLUSIONS

Studies with heated, watereextracted beef and pork
muscle demonstrated that‘prooxidant'substances can be
extracted from muscle tissue with water. Much of the
prooxidant activity was removed from"the aqueous muscle
extract by dialysis.' These observations indicated that
low-molecular weight substances play a major role in
acceleration of lipid oxidation in cooked meat.

It was hypothesized that'nonsheme iron is a major
prooxidant in the"muscle extract. This hypothesis was
tested by measuring the ability'of ferrous iron to accel-
erate lipid'oxidation"in:watereextracted,"cooked beef and

2

pork muscle. Levels of Fe+ as low as 1 ppm were effective

in accelerating lipid oxidation. Ascorbic acid enhanced
the prooxidant activity of Fe+2. However, myoglobin added
to the waterhextracted muscle at levels ranging from 1-10
mg/g failed to exhibit prooxidant activity. It was con-
cluded that:noneheme'iron is a major prooxidant in cooked
meat, while myoglobin does not accelerate lipid oxidation
in denatured muscle.

The effect of metmyoglobin and noneheme iron on the
oxidation of phosphatidyl ethanolamine dispersions was
evaluated by'a variety’of methods. ”Both metmyoglobin and

ferrous iron were effective in accelerating the oxygen

74

75
uptake in PE dispersions bufferedsat pH 5.5. The accelerat-

ing effect of metmyoglobin and Fe+2

"on oxygen uptake in

PE dispersions at pH 5.5 was confirmed by higher TBA values,
greater production'of fluorescent material and increased
losses of unsaturated fatty acids. The data clearly
indicate that both metmyoglobin and Fe+2are'capable of
catalyzing the oxidation of phospholipids at‘pH 5.5.

At pH 7.0, neither metmyoglobin nor Fe+2 accelerated
oxygen uptake in PE'dispersions.' Metmyoglobin did increase
the production of fluorescent‘materials at pH 7.0, but
failed to increase TBA values or enhance fatty acid
destruction.“ Ferrous iron increased TBA values at pH 7.0,
but did not accelerate the production‘of fluorescent
materials. The results suggested that metmyoglobin and

Fe+2

failed to accelerate phospholipid‘oxidation at pH
7.0. ‘Spectral studies indicated that metmyoglobin catalysis
at pH 5.5 was accompanied by degradation‘of the heme groups.
At pH 7.0, however, the heme groups were not destroyed.
Gas-liquid chromatography was used to separate the
volatile compounds produced by oxidizing PE dispersions
and cooked meat.r The volatiles found in the headspace
'above oxidizing phospholipid dispersions were similar to
the headspace volatiles in a samplemof oxidized, cooked
meat. Hexanal was a major component of the volatiles from
both oxidized phospholipids and rancid cooked meat. The

concentration of hexanal was apparently related to the

degree of off—odors in the meat.

BIBLIOGRAPHY

‘BIBLIOGRAPHY

Banks, A. 1961.' The role of cytochromeec in the autoxi-
'dation of unsaturated fatty‘acids. 'Chem. and Ind.
No. 2:40.

Banks, A., Eddie, E. and Smith, J.G. M. 1961. Reactions
of cytochromebc with methyl linoleate*hydroperoxide.
Nature 190:908.

Barber, A. A. 1966.. Lipid peroxidation in rat tissue
homogenates: Interaction of iron and ascorbic acid
as the normal catalytic mechanism.” Lipids 1:146.

Bawn, C. E. H. 1953. Free radical reactions in solution
‘initiated'by heavy metal ions. 'Discussion5"Faraday
Soc. 14:181.

Ben-Aziz, A., Grossman, S., Ascarelli, I. and Budow, P.
1970. Linoleate oxidation inducedby lipoxygenase
and heme proteins. Anal. Biochem. 34:88.

Bendall, J. R. and Wismer-Pedersen, J. ‘1962. Some proper-
ties of the fibrillar proteinS‘of'normal and watery
pork muscle. J. Food Sci. 27:144.

Bernheim, F. '1964.‘ Effect of Z-mercaptoethylamine on the
peroxidation in microsomes of unsaturated lipid
catalyzed by cysteine or thioacetic acid and UV
light. Radiation Res. 23:454.

Brown, W- 0., Harris, L. S. and Olcott, H. S. 1963.
'Catalysis of unsaturated lipid oxidation by iron
protoporphyrin derivatives. Arch. Biochem. Biophys.
101:14.

Campbell, A. M. and Turkki, P. R. 1967.. Lipids of raw
and cooked ground beef and pork. J. Food Sci.
32:143.

Carraway, K. L. and Huggins, J. W.‘ 1972. Reactivity of
amino groups in phospholipid dispersions. 'Chem.
Phys. Lipids 8:56.

Castell, C. H., Moore, B. and Neal, W. ~1966."Control and
cause-of spurious TBA values from fish muscle in the
presence of inorganic iron salts. J. Can. Fisheries
Res. Board 23:737.

76

77

Castell, C. H. and Spears, D. M. 1968. Heavy metal ions
and the development of rancidity in blended fish
muscle. J. Can. Fisheries Res. Board 25:639.

Chang, I. and Watts, B. M. 1952. The fatty acid content
of meat and poultry before and after cooking. J.
Amer. Oil Chem. Soc. 29:334.

Corliss, G. A. 1968.' Phospholipid oxidation in emulsions.
Ph.D. Thesis, Michigan State University, East Lansing,
Mich.

Craig, H. B., Blumer, T. N., Smart, W. W. 6., Jr., and
Wise, M. B. 1966.“ Evaluation of hemoglobin, myo-
globin and blood oxygen content and organoleptic
qualities of beef from steers fed grain on pasture
or cut forage and grain in drylot. J. Anim. Sci.
25:1128.

Cross, C. K. and Ziegler, P. 1965. A comparison of the
volatile fractions from cured and uncured meats.
J. Food Sci. 30:610.

Desai, I. D. and Tappel, A. L. 1963. Damage to proteins
by peroxidized lipids. J. Lipid Res. 4:204.

Dugan, L. R., Jr. 1961. Development and inhibition of
oxidative rancidity in foods. Food Technol. 15:10.

El-Gharbawi, M. I. and Dugan, L. R., Jr. 1965. Steam-
volatile components of freeze~dried beef stored at
room temperature. J. Food Sci. 30:814.

El-Gharbawi, M. I. and Dugan, L. R., Jr.' 1965. Stability
of nitrogenous compounds and lipids during storage
of freeze-dried raw beef. J. Food Sci. 30:817.

Eriksson, C. E., Olsson, P. A. and Svensson, S. G. 1970.
Oxidation of fatty acids by heat treated hemopro-
teins. Lipids 5:365.

Evans, C. 0.. 1961. Chemical changes accompanying flavor
deterioration of vegetable oils. Proc. Flavor
Chem. Symposium, p. 123. 'Campbell Soup Co., Camden,
New Jersey.

Fletcher, B. L. and Tappel, A. L. 1971. Fluorescent modi-
fication of serum albumin by lipid peroxidation.
Lipids 6:172.

Fox, J. B., Jr. 1966. The chemistry of meat pigments. J.
Agr. and Food Chem. 14:207.

78

Gaddis, A. M., Ellis, R. and Currie, G. T. 1961. Carbonyls
in oxidizing fat. V. The composition of neutral
volatile monocarbonyl compounds from autoxidized
oleate, linoleate, linoeate esters and fats. J.
Amer. Oil Chem. Soc. 38:371-

Giam, I. and Dugan, L. R., Jr. 1965. The fatty acid compo-
sition of free and bound lipids in freeze—dried
meats. J. Food Sci. 30:262.

Greene, B. E., Hsin, I. and Zipser, M. W. 1971. Retarda-
tion of color changes in raw ground beef. J. Food
Sci. 36:940.

Heaton, F. W. and Uri, N. 1961. The aerobic*oxidation of
unsaturated fatty acids and their esters: Cobalt
stearate-catalyzed oxidation of linoleic acid. J.
Lipid Res. 2:152.

Herz, K. O. and Chang, S. S. 1970. Meat flavor. In
"Advances in Food Res.,"Vol. 18, p. 1. Academic
Press, New York.

Holub, B. J. and Kuskis, A. 1969. Molecular species of
phosphatidyl ethanolamine from egg yolk. Lipids
4:466.

Hornstein, I. and'Crowe, P. F. 1960. Flavor studies on
beef and pork. J. Agr. Food Chem. 8:494.

Hornstein, I., Crowe, P. F. and Heimberg, M. J. 1961.
Fatty acid composition of meat tissue lipids. J.
Food Sci. 26:581.

Hornstein, I., Crowe, P. F. and Hiner, R.’ 1967.' Composi-
tion of lipids in some beef muscles.' J. Food Sci.
32:650.

Hornstein, I., Crowe, P. F. and Sulzbacher, W. L. 1960.
Constituents of meat flavor: Beef. J. Agr. Food
Chem. 8:65.

Hutchins, B. K., Liu, T. H. P. and Watts, B. M. 1967.
Effect of additives and refrigeration on reducing
activity, metmyoglobin and malonaldehyde of raw
ground beef. 'J. Food Sci. 32:214.

Ingold, K. V. 1962. Metal catalysis. In "Symposium on
Foods: Lipids and Their Oxidation," ed. Schultz,
H. W., Day, E. A. and Sinnhuber, R. 0., p. 93. AVI
Publishing Co., Westport, Conn.

Kendrick, J., and Watts, B. M. 1969. Acceleration and
inhibition of lipid oxidation by heme compounds.
Lipids 4:454.

79

Kesinkel, A., Ayres, J. C. and Snyder, H. E. 1964.
Determination of oxidative changes in raw meats
by the 2-thiobarbituric acid method. Food Technol.
18:223.

Koch, R. B.‘ 1962. Dehydrated foods and model systems.
In "Symposium on Foods: Lipids and Their Oxidation,"
ed. Schultz, H. W., Day, E. A. and Sinnhuber, R. 0.,
p. 230. AVI Publishing Co., Westport, Conn.

Kokatnur, M. G., Bergan, J. G. and Draper, H. H. 1966.
Observations on the decomposition of hemin by fatty
acid hydroperoxides. Proc. Soc. Expt. Biol. Med.
123:314.

Kuchmak, M. and Dugan, L. R., Jr. 1965. Composition and
positional distribution of fatty acids in phospho-
lipids isolated from pork muscle tissue. J. Amer.
Oil Chem. Soc. 42:45.

Kwon, T. W., Menzel, D. B. and Olcott, H. S. 1965.
Reactivity of malonaldehyde with food constituents.
'J. Food Sci. 30:808.

Kwon, T. W. and Watts, B. M. 1964. Malonaldehyde in
aqueous solution and its role as a measure of
lipid oxidation in foods. ‘J. Food Sci. 29:294.

Luddy, F. E., Herb, S. F., Madigman, P., Spinelli, A. M.
‘ and Wasserman, A. E. 1970. Color and the lipid
composition of pork muscles. J. Amer. Oil Chem.

Soc. 47:65.

Lundberg, W. O. 1962. Mechanisms. In "Symposium on
Foods: Lipids and Their Oxidation," ed. Schultz,
H. W., Day, E. A., and Sinnhuber, R. 0., p. 31.
AVI Publishing Co., Westport, Conn.

MacLean, J. and Castell, C. H. 1964. Rancidity in lean
fish muscle. I. A proposed accelerated copper-
catalyzed method for evaluating the tendency of
fish muscle to become rancid. J. Can. Fisheries
Res. Board 21:1345.

Mahler, H. R. and Cordes, E. H. 1966. "Biological
‘Chemistry", p. 581. Harper and Row, New York.

Maier, V. P. and Tappel, A. L. 1959. Rate studies of
unsaturated fatty acid oxidation catalyzed by
hematin compounds. J. Amer. Oil Chem. Soc. 36:8.

Lea, C. H. 1957. Deteriorative reactions involving phospho-
lipids and lipoproteins. J. Sci. Food Agr. 8:1.

80

Lea, C. H.‘ 1962. The oxidative deterioration of food
lipids.' In'"Symposium'on”Foods: 'Lipids and Their
Oxidation," ed. Schultz, H. W., Day, E. A. and
Sinnhuber, R. 0., p. 3. AVI Publishing Co., West-
port, Conn.

Ledward, D. A. 1971. On the nature of cooked meat hemo-
protein. J. Food Sci. 36:883.

Ledward, D. A. and MacFarlane, J. J. 1971.‘ Some observa-
'tions on myoglobin and lipid oxidation in‘frozen
beef. J. Food Sci. 36:987.

Lee Shin, T. and Smith,-L. M. 1971. Role of the bases
and phosphoryl bases of phospholipids in the
autoxidation of methyl linoleate emulsions.
Lipids 6:196.

Lewis, s. E. and Wills, E. 0. 1963. Inhibition of the,
autoxidation of unsaturated fatty acids byhematin
proteins. Biochim. Biophys. Acta 70:336.

Liu, H. 1970. Catalysts of lipid peroxidation in meats
l. Linoleate peroxidation catalyzed by Meth or
Fe (II)-EDTA. ‘J. Food Sci. 35:590.

Liu, H. and Watts, B. M. 1970. Catalysts of lipid
peroxidation in meats. 3. Catalysts of oxidative
rancidity in meats. J. Food Sci. 35:596.

McClare, C. W. F. 1972. Addendum: A warning. An
accurate and convenient organic phosphorus assay.
Anal. Biochem. 45:658.

McKnight, R. C. and Hunter, F. E. 1965. "EffectS'of inor-
'ganic iron on the thiobarbituric acid method for
determination of lipid peroxides. Biochim. Biophys.
‘Acta 98:643.

Morrison, W. R. and Smith, L. M.‘ 1964. Preparation of
fatty acid methyl esters and dimethylacetals from
lipids with boron fluoride-methanol. J. Lipid
Res. 5:600.

Moskovits, V. G. and Kielsmeier, E. W. 1960. An effect
of iron and sodium chloride on flavor-in'sausage.
Proc. 12th Res. Conf. Amer. Meat Institute Founda-
tion, University of Chicago, Chicago, Illinois.

Nakamura, Y. and Nishida, T. 1971. Effect of hemoglobin
*concentration‘on the oxidation of linoleic acid.
J. Lipid Res. 12:149.

Nishida, T. and Nishida, H.’ 1965. Interaction of ferri-
‘hemoglobin with7peroxidized low density lipoproteins.
'J. Biol. Chem. 240:225.

81

O'Brien, P. J. 1969. Intracellular mechanisms for the
decomposition of a lipid peroxide. 1. Decomposi-
tion of a lipid peroxide by metal ions, heme compounds
andnucleophiles. Can. J. Biochem. 47:485.

O'Keefe, P. W., Wellington, G. H., Mattick, L. R. and
Stouffer, J. R. 1968. Composition of bovine muscle
lipids at various carcass locations. J. Food Sci.
33:188.

Olivecrona, T. and Oreland, L. 1971. Reassociation of
soluble monamine oxidase with lipidedepleted mito-
chondria in the presence of phospholipids. Bio~
chemistry 10:332.

Robinson, J. D. 1965. Structural changes in microsomal
suspensions. III. Formation of lipid peroxides.
Arch. Biochem. Biophys. 112:170.

Robinson, M. E. 1924. Haemoglobin and methaemoglobin as
oxidative catalysts. Biochem. J. 18:255.

Roubal, W. T. 1967. Oxidative deterioration of flesh
lipids of pacific cod (Gadua macrocephalue). J.
Amer. Oil Chem- Soc. 44:325.

Reuser, G., Siakotos, A. N. and Fleischer, S. 1966.
Quantitative analysis of phospholipids by thin-
layer chromatography andphosphorus analysis of
spots. Lipids 1:85.

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

Sinnhuber, R. 0, Yu, T. C. and Yu, Te Chang. 1958.
Characterization of the red pigment formed in the
2-thiobarbituric acid determination of oxidative
rancidity. Food Res. 23:626.

Stahl, E., ed. 1969. "Thin-Layer Chromatography," 2nd ed.,
p. 877. Springer-Verlag, New York.

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

Tappel, A. L. 1962. Home compounds and lipoxidase as
‘ biocatalysts. In "Symposium on Foods: Lipids and
Their Oxidation," ed. Schultz, H. W., Day, E. A.,
and Sinnhuber, R.O., p. 122. AVI Publishing Co.,
Westport, Conn.

Tarladgis, B. G. 1961. An hypothesis for the mechanism
of the heme-catalyzed lipid oxidation in animal
tissues. J. Amer. Oil Chem. Soc. 38:479.

82

Tarladgis, B. 6., Watts, B. M., Younathan, M. T. and Dugan,

Timms,

Uri, N.

L. R. 1960. A distillation method for the quanti-
tative determination of malonaldehyde in rancid
foods.‘ J. Amer. Oil Chem. Soc. 37:44.

M. J. and Watts, B. M.‘ 1958.‘ Pretection*of cooked

“meats with phosphates. Food-Technol.'12:240.

1956. Metalion'catalysis and polarity of

‘environment’in‘the"aerobic"oxidation‘of'unsaturated

Watts,

Watts,

Watts,

Watts,

fatty'acids."Nature"177:1177.

B. M. 1954. Oxidative rancidity and discoloration
in meat. In "Advances in Food Research," Vol. 5,,
p. 1. Academic Press, Inc., New York.

B. M. 1961. ‘The role of lipid oxidation in lean
tissues in flavor deterioration of meat and fish.
In "Proceedings Flavor Chemistry Symposium," p.
83. 'Campbell Soup Co., Camden, New Jersey.

B. M. 1962. Meat products. In "Symposium on

'Foods: 'Lipids and Their Oxidation," ed. Schultz,

H. W., Day, E. A. and Sinnhuber, R. O., p. 202-
AVI Publishing Co., Westport, Conn.

B. M., Kendrick, J., Zipser, M., Hutchins, B., and
Saleb, B. 1966. Enzymatic reducing pathways in
meat. J. Food Sci. 31:855.

B. M. and Peng, D. 1947. Lipoxidase activity of

‘hog hemoglobin and muscle extract. J. Biol. Chem.

Wills,'

‘Wills,

Wills,

170:441.

E. D. 1964. The effect“of“inorganiC'iron‘on the
thiobarbituric acid method for the determination of
lipid peroxides. Biochim. Biophys. Acta 84:475.

E. D. 1965. Mechanism oflipid peroxide formation
in tissues.' Role of metals and haematin proteins
in the catalysis of the oxidation of unsaturated
fatty acids. Biochim. Biophys. Acta 98:238.

E. D. 1966.' Mechanism of lipid peroxide formation
in animal tissues. Biochem. J. 99:667.

Younathan, M. T. and Watts, B. M. 1959. Relationship of

meat'pigment5“to lipid oxidation. 'Food Res. 24:728.

'Younathan, M. T. and Watts, B. M. 1960. 'Oxidation of

tissue lipids in cooked pork. J. Food Sci. 25:538.

'Zipser, M. W., Kwon, T. W. and Watts, B. M. 1964. Oxidative

changes in cured and uncured frozen cooked pork. J.
Agr. Food Chem.‘12:105.

HICHIGQN STRTE UNIV. LIBRRRIES
\IHI Hill WIIIWI {WI M I“ ll“! I11! IIHH \llll NH |H!|
31293102003690