Meat Science

Volume 77, Issue 1, Pages 1-147 (September 2007)

1. Editorial Board
Page IFC

2. 53rd International Congress of Meat Science and Technology, Beijing, China, 5–

10th August 2007
Page 1
Guanghong Zhou

3. Stratification of toughness in beef roasts

Pages 2-6
H.J. Swatland

4. Mechanisms controlling pork quality development: The biochemistry controlling

postmortem energy metabolism
Pages 7-16
T.L. Scheffler and D.E. Gerrard

5. Predictive microbiology: Quantitative science delivering quantifiable benefits to the

meat industry and other food industries
Pages 17-27
T.A. McMeekin

6. Managing safety and quality through the red meat chain

Pages 28-35
P. Desmarchelier, N. Fegan, N. Smale and A. Small

7. Application of genomic technologies to the improvement of meat quality of farm

animals
Pages 36-45
Yu Gao, Ran Zhang, Xiaoxiang Hu and Ning Li
8. Identification of pork quality parameters by proteomics
Pages 46-54
Dick F.M. van de Wiel and Wei Li Zhang

9. Microbial ecosystems of traditional fermented meat products: The importance of

indigenous starters
Pages 55-62
R. Talon, S. Leroy and I. Lebert

10. A fresh look at meat flavor

Pages 63-80
C.R. Calkins and J.M. Hodgen

11. Technologies to shorten the drying period of dry-cured meat products

Pages 81-89
J. Arnau, X. Serra, J. Comaposada, P. Gou and M. Garriga

12. Interventions to reduce/eliminate Escherichia coli O157:H7 in ground beef

Pages 90-96
M. Koohmaraie, T.M. Arthur, J.M. Bosilevac, D.M. Brichta-Harhay, N.
Kalchayanand, S.D. Shackelford and T.L. Wheeler

13. Application of proteomics to understand the molecular mechanisms behind meat

quality
Pages 97-104
Kristin Hollung, Eva Veiseth, Xiaohong Jia, Ellen Mosleth Færgestad and Kjell
Ivar Hildrum

14. Tenderness and oxidative stability of post-mortem muscles from mature cows of
various ages
Pages 105-113
Youling L. Xiong, Oliver E. Mullins, John F. Stika, Jie Chen, Sue P. Blanchard
and William G. Moody

15. Biochemical changes during processing of traditional Jinhua ham

Pages 114-120
G.H. Zhou and G.M. Zhao

16. Effects of metabolic modifiers on carcass traits and meat quality

Pages 121-135
M.E. Dikeman
17. Cured meat products without direct addition of nitrate or nitrite: what are the
issues?
Pages 136-147
Joseph G. Sebranek and James N. Bacus

18. AMSA Mission & Membership Application

Pages I-II
The qualities of meat — its composition, nutritional value, wholesomeness and consumer acceptability — are largely
determined by the events and conditions encountered by the embryo, the live animal and the post-mortem
musculature. The control of these qualities, and their further enhancement, are thus dependent on a fuller
understanding of the commodity at all stages of its existence — from the initial conception, growth and development of
the organism to the time of slaughter and to the ultimate processing, distribution, preparation, cooking and
consumption of its meat.
It is the purpose of Meat Science to provide an appropriate medium for the dissemination of interdisciplinary and
international knowledge on all the factors which influence the properties of meat. The journal is predominantly
concerned with the flesh of mammals; however, contributions on poultry meat may be published, especially if these
have relevance to our overall understanding of the relationship between the nature of muscle and the quality of the
meat which muscles become post mortem.

Editor
Professor David A. Ledward,
School of Biosciences, Division of Food Sciences,
University of Nottingham, Sutton Bonington Campus,
Loughborough, LE12 5RD, UK
e-mail: david.ledward@nottingham.ac.uk
Associate Editors Professor Howard J. Swatland
Dr. Michael Enser Department of Animal and Poultry Science,
Division of Food Animal Science, University of Bristol, University of Guelph, Guelph,
Langford, Bristol BS40 5DU, UK Ontario, Canada, N1G 2WI
e-mail: m.enser@bristol.ac.uk e-mail: swatland@uoguelph.ca
fax: +44 (117) 9289324 fax: +1 (519) 767 0573

Editorial Board
P. Barton-Gade K. Incze A. Suzuki
Danish Meat Research Institute, Hungarian Meat Research Institute, University of Niigata, Japan
Maglegardsvej 2, 4000 Roskilde, Budapest, Hungary
F. Toldrá
Denmark
J. Kerry Instituto de Agroquimica y
E.A. Decker University College Cork, Ireland Tecnologia de Alimentos (CSIC),
Chenoweth Lab, Department of Burjassot, Spain
Food Science, University of S.D. Morgan Jones
Massachusetts, USA Agriculture and Agri-Food Canada, E. Tornberg
Lethbridge, Canada Food Engineering Department,
P. Gatellier Lund University, Sweden
INRA Theix, Saint Genes Champanelle, R.T. Naudé
France W. Uchman
Department of Biochemistry,
University of Agriculture,
C.O. Gill University of Port Elizabeth, South Africa
Institute of Meat Technology,
Agriculture and Agri-Food Canada, Poznan, Poland
E. Puolanne
Lacombe, Alberta, Canada
University of Helsinki, Helsinki, Finland P.D. Warriss
M.L. Greaser School of Veterinary Science,
University of Wisconsin, Madison, USA J.W. Savell
Texas A&M University, College Station, University of Bristol,
K.O. Honikel Texas, USA Langford, UK
Bundesanstalt für Fleischforschung, J.D. Wood
Kulmbach, Germany R. Shorthose School of Veterinary Science,
D. Hopkins CSIRO, Queensland, University of Bristol,
DPI (NSW), Center for Sheep Meat Australia Langford, UK
Development, Cowra, NSW, Australia E.T.F. Silveira G.H. Zhou
M.C. Hunt Instituto De Tecnologia Del Nanjing Agricultural University,
Kansas State University, USA Alimentos, Campinas, Brazil Nanjing, P.R. China
 MEAT
SCIENCE
Meat Science 77 (2007) 1
www.elsevier.com/locate/meatsci

Preface

53rd International Congress of Meat Science and Technology,

Beijing, China, 5–10th August 2007

On behalf of the scientific and organizing committee of Sincere thanks are also extended to the following indi-
the 53rd International Congress of Meat Science and Tech- viduals for their efficient reviewing work:
nology, it is my great honor to present this Special Issue of
Meat Science containing the manuscripts of invited speak- J. Arnau S.D. Morgan Jones
ers for the Congress. C.R. Calkins L. Sammel
The topics addressed by the invited speakers focus on L. Cocolin J.W. Savell
meat safety, meat production, muscle biochemistry, meat M. Du J.G. Sebranek
quality and nutrition, meat processing and meat products D.E. Gerrard A. Suzuki
respectively. The latest research findings and the current C.O. Gill H. Swatland
research issues in the above aspects are generally covered. K. Hollung R. Talon
We trust that this special issue will promote the interna- M. Koohmaraie R. Taylor
tional communication of research in meat science and D. Ledward Y.L. Xiong
technology. T. McMeekin
I would like to give my sincere thanks to all the authors
who have contributed their masterpiece manuscripts. My Guest Editor
utmost appreciation goes to Dr. Howard Swatland – Guanghong Zhou
without his kind help, effective editing and hard work, we Chairman of 53rd ICoMST,
cannot image how we could fulfill this task. I also like to Professor/Vice President,
thank Prof. Weili Zhang for his coordination and Ms. Nanjing Agricultural University,
Wendy Hurp of Elsevier publishers for her kind co- Nanjing, 210095, PR China
operation. E-mail address: ghzhou@njau.edu.cn

0309-1740/$ - see front matter Ó 2007 Published by Elsevier Ltd.

doi:10.1016/j.meatsci.2007.06.001
 MEAT
SCIENCE
Meat Science 77 (2007) 2–6
www.elsevier.com/locate/meatsci

Stratification of toughness in beef roasts

H.J. Swatland *

Department of Animal and Poultry Science, University of Guelph, Guelph, Ont., Canada N1G 2W1

Received 23 February 2007; received in revised form 3 April 2007; accepted 6 April 2007

Abstract

Most meat scientists adopt a reductionist approach to the study of meat toughness, taking a few intramuscular cores from one or
more muscles to simplify the enormous complexity of toughness in all the retail cuts derived from a whole carcass. This is a valid
approach to a complex problem, but we should also start to consider how consumers respond to bulk meat such as steaks and roasts.
Probing whole roasts reveals a complex internal structure, detectable by both connective tissue fluorescence and resistance to penetration.
The dorsal aponeurosis of the Longissimus thoracis is a major connective tissue stratum in beef rib roasts and its properties are correlated
with those of adjacent intramuscular connective tissues. When the aponeurosis is cooked, its reflectance first increases with protein dena-
turation, then decreases with gelatinisation. Heat-induced contraction is concurrent with the increase in reflectance. Gelatinisation is
reduced if the aponeurosis is mechanically restrained to resist contraction. Thus, mechanical restraint interacts with heat penetration
in explaining stratification of toughness in bulk meat.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Fluorescence; Penetrometry; Connective tissue; Meat toughness; Beef

1. Introduction
Meat scientists routinely analyse meat toughness using
intramuscular core samples for methods such as the War-
ner–Bratzler shear test (Voisey, 1976). This reveals varia-
tion in toughness between and within major muscles of
the carcass (Belew, Brooks, McKenna, & Savell, 2003;
Gruber et al., 2006; Janz, Aalhus, Dugan, & Price, 2006).
But when meat is consumed as a whole steak or roast,
the stratification of connective tissues is superimposed on
the toughness of individual muscles. The agreeable sensa-
tion of an internally tender core of meat may be spoiled
by tough seams of connective tissue on the surfaces of, or
between individual muscles. Overall tenderness in bulk
meat is complicated by the amount of connective tissue gel-
atinisation, which depends on cooking method and depth
in the meat. Building on our knowledge of intramuscular
connective tissue toughness (Bailey & Sims, 1977; Dutson,
*
Tel.: +1 519 824 4120x53670; fax: +1 519 767 0573.
E-mail address: swatland@uoguelph.ca
Hostetler, & Carpenter, 1976; Loyd & Hiner, 1959; Smith
& Judge, 1991), in this presentation we will review pub-
lished experiments and other data. The objective is to look
at the problems involved in understanding toughness in
bulk meat containing different muscles and types of con-
nective tissue, as when a consumer evaluates a whole steak
or roast, or when intact sides of beef are probed in attempts
to predict meat toughness (Wulf, Emnett, Leheska, &
Moeller, 2002).
Lanius is the Latin word for a butcher (Lewis, 1891). We
may use the adjective, laniary, to denote structures in cuts
of meat prepared by a butcher, hence laniary strata for the
layers of connective tissues in bulk meat. The thickness of
the perimysium (connective tissue around bundles of mus-
cle fibres) and the diameters of its collagen fibres both affect
shear force (Brooks & Savell, 2004; Janz, Aalhus, & Price,
2006; Light, Champion, Voyle, & Bailey, 1985). But here
we are more concerned with the epimysium (connective tis-
sue on the muscle surface). Epimysium creates detectable
laniary strata when we probe bulk meat. Thus, we detect
major signals from strata (epimysium and fasciae) superim-

0309-1740/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.013
 H.J. Swatland / Meat Science 77 (2007) 2–6
posed on the background signals between the strata (endo-
mysium and perimysium within muscles, and reticular
fibres within adipose tissue). Intramuscular background
signals correspond to the vast amount of information col-
lected by shearing intramuscular cores. We will be ques-
tioning the relationships between major signals from
laniary strata and the background signals between strata,
and will then go on to look at complexities caused by
cooking.
2. Comparing a muscle surface and its interior
When a probe is pushed into the eye muscle (Longissi-
mus thoracis) of a beef rib roast, it passes through several
grades of fibrous connective tissue (Swatland, 2006a).
There is epimysium on the muscle surface, perimysium
around bundles of muscle fibres and endomysium around
individual muscle fibres. The aponeurosis is a very thick
part of the muscle epimysium. Site 1 in Fig. 1 is where
the maximum penetrometer force is encountered as the
probe penetrates through the aponeurosis. Resistance to
penetration is completely from collagen fibres. Site 2 is
the minimum force immediately under the aponeurosis,
with the probe tip passing through only muscle fibres and
endomysium. Here resistance is limited to endomysial col-
lagen fibres plus myofibrils, with the resistance from myo-
fibrils being an interaction of sarcomere length and the
extent of post-mortem autolysis. Site 3 is the maximum
force anywhere intramuscularly when the probe tip
encounters the thickest septum of perimysial connective tis-
sue. The tip of our probe can detect both resistance to pen-
etration and the optical properties of tissues at this point
(Swatland, 2005).
The fluorescence data in Fig. 2 show where the probe tip
passes through the Longissimus thoracis aponeurosis and
underlying muscle. Fluorescence increases to a maximum
where the probe penetrates the aponeurosis 2(1). Then
there is a sharp decrease in fluorescence to the background
level intramuscularly 2(2). A major connective tissue sep-
tum is encountered intramuscularly 2(3).
Fig. 1. Anatomy of a probe transect (from Swatland, 2006a). The
aponeurosis is at point (1), the muscle immediately below the aponeurosis
at point (2) contains only muscle fibres and endomysium, and a major
seam of perimysium is at point (3).
3
Fig. 2. Fluorescence signal (from Swatland, 2006a). After the probe enters
the meat at distance 0 mm, it detects high fluorescence of the aponeurosis
at point (1), low fluorescence of muscle immediately below the aponeurosis
at point (2), and the intermediate fluorescence of a major seam of
perimysium at point (3).
Looking at the corresponding penetrometer data,
(Fig. 3), we see resistance to penetration increasing until
the aponeurosis is penetrated. 3(1). Then there is a sharp
decrease in resistance until the background level associated
with endomysium and muscle fibres is reached 3(2). There
is a resistance peak 3(3) corresponding to the perimysial
septum identified by its fluorescence 2(3).
We all know meat is a highly variable commodity, and it
is no surprise to find situations where the patterns of force
and fluorescence do not match as clearly as in the example
shown here. But these mismatches can be explained. Some-
times there may be an intramuscular fluorescence peak
without any corresponding resistance. This may be where
the probe window has passed tangentially by a septum of
connective tissue, but has not actually penetrated through
it. Conversely, there may be a peak of resistance without
any increase in fluorescence. This may be where the probe
tip has penetrated muscle fibres with strong myofibrils,
either because of short sarcomeres or a lack of post-mor-
tem autolysis. Ignoring these exceptions, we will collect
our data using a split screen showing two matching data
sets (with fluorescence peaks matching the pattern of pen-
etrometer peaks).
In a typical set of Canadian prime rib roasts (well aged,
well marbled Canada Grade AAA; n = 13), and taking
the means of five measurements from each roast, fluores-
cence of the aponeurosis at site 1 is correlated with both
Fig. 3. Penetrometer resistance matching Fig. 2 (from Swatland, 2006a).
After the probe enters the meat at distance 0 mm, it detects high resistance
of the aponeurosis at point (1), low resistance of muscle immediately
below the aponeurosis at point (2), and the resistance of a major seam of
perimysium at point (3).
4 H.J. Swatland / Meat Science 77 (2007) 2–6
intramuscular fluorescence at site 2 (r = 0.81, P < 0.0005,
n = 13) and with fluorescence of perimysial septum at site
3 (r = 0.78, P < 0.005, n = 13). Fluorescence at sites 2
and 3 is correlated (r = 0.87, P < 0.005, n = 13). Mean rel-
ative fluorescence intensities at sites 1, 2 and 3 are, respec-
tively, 0.96 ± 0.40, 0.12 ± 0.08 and 0.28 ± 0.14 (relative to
white paper standard = 1). With a paired t-test, differences
are significant, P < 0.001.
For the matching penetrometer data, the force required
to penetrate the aponeurosis at site 1 is correlated with
both intramuscular resistance at site 2 (r = 0.51, P < 0.05,
n = 13) and with perimysial septum at site 3 (r = 0.63,
P < 0.025, n = 13). No relationships are detected in this
set of samples between penetrometer forces at sites 2 and
3 (r = 0.37, NS, P > 0.05, n = 13). Mean values for pene-
trometer forces at sites 1, 2 and 3 are, respectively,
27.7 ± 4.8 N, 8.5 ± 2.1 N and 18.0 ± 3.9 N. With a paired
t-test, differences are significant, P < 0.001. Pooling all
the data (five measurements from 13 roasts), force at site
3 is correlated with both force at site 1 (r = 0.39,
P < 0.0005, n = 65) and at site 2 (r = 0.29, P < 0.01,
n = 65).
Despite these obvious relationships within the data sets
for fluorescence and penetrometer force, the relationships
of fluorescence and penetrometer force are difficult to
detect – if they exist. No significant relationships of fluores-
cence with penetrometer force are detectable at sites 1
(r = 0.03, NS, P > 0.05, n = 13) or 3 (r = 0.14, NS,
P > 0.5, n = 13). Although, fluorescence is correlated with
force at site 2 (r = 0.52, P < 0.05, n = 13).
What do these data show us? Firstly, both fluorescence
and penetrometer data show there are obvious relation-
ships between connective tissues on the muscle surface
and those within the muscle. This is to be expected because
both are part of the same system developed by the animal
to transmit the forces of muscle contraction to the skeleton.
Secondly, the perimysial effect on meat toughness detected
by Brooks and Savell (2004) may now be quantified. The
strength of the strongest perimysium in aged beef roasts
is more than twice the sum of endomysial plus myofibrillar
strength (18.0 ± 3.9 N for perimysium and 8.5 ± 2.1 N for
endomysium + myofibrils). As expected, the aponeurosis is
stronger than anything else in the Longissimus thoracis
(27.7 ± 4.8 N). But, remember, penetrometer data are diffi-
cult to standardize. A sharp probe gives lower values than
a blunt probe. We can only make relative deductions: the
aponeurosis in aged roasts is stronger than the strongest
perimysium, and the strongest perimysium is stronger than
endomysium plus myofibrillar structure (P < 0.001).
This helps explain why penetrometer data are so difficult
to use on-line for the prediction of toughness in a whole
carcass. The more senior members of this audience will
doubtless remember several probe systems failing to live
up to their initial expectations. Although we see here that
the strength and composition of epimysium, perimysium
and endomysium are all interrelated, we can also see the
risk of a large sampling error. Epimysial connective tissues
are much stronger than perimysial and endomysial connec-
tive tissues, and chance encounters with epimysial tissues
will certainly increase sampling error. Thus, it is essential
to standardize the anatomical location of probe transects
and this is extremely difficult to achieve with either human
operators or robotics.
How about using fluorescence probes to predict tough-
ness? All the positive tests for predicting toughness from
fluorescence were based on intramuscular transects (Swat-
land, 1995). Thus, we see fluorescence correlated with force
at site 2 involving only muscle fibres and endomysium
(r = 0.52, P < 0.05, n = 13). But we can now see why
researchers who tested fluorescence probes without solving
the anatomical problems of sampling were doomed to fail-
ure. Despite a network of significant interrelationships
within strength and fluorescence data among different
strata of connective tissues, a key relationship fails in the
data presented here. No relationship is detectable between
the fluorescence of the aponeurosis and its strength. The
most likely explanation is elastic deformation. If a fluores-
cent septum is wrapped around an optical window, then
linearity of fluorescence with strength is unlikely – a satu-
ration effect. Thus, fluorescence probes need to cut through
small structures, generating a fluctuating signal whose
peaks can be counted and measured. Endomysial and
minor perimysial fluorescence signals are related to meat
toughness (Swatland, 1995), but not if we add in random
large peaks from connective tissues outside the muscle of
interest.
3. Cooking the aponeurosis on the muscle surface
We know a lot about meat cooking (Tornberg, 2005),
and how collagen fibres contract to squeeze out fluid from
muscle fibres (Bendall & Restall, 1983). Collagen fibres
may be completely gelatinized if meat is thoroughly
cooked, especially with moist heat, but this may not hap-
pen when steaks or roast are lightly cooked. Thermal dena-
turation of intramuscular collagen typically occurs at 53–
63 °C (Martens, Stabursvik, & Martens, 1982) but denatur-
ation may occur at 70 °C in extramuscular collagenous
structures such as ligaments (Vangsness et al., 1997). What
happens to the aponeurosis on the surface of the Longissi-
mus thoracis when it is cooked? Collagen contracts and
melts during thermal denaturation (Chien, 1975), but are
optical and rheological changes simultaneous? To answer
these questions, strips of aponeurosis were removed from
beef roasts (n = 15) similar to those used in the last exper-
iment. Strips were heated gradually in distilled water while
recording their length and reflectance (Swatland, 2006b).
Heating first caused a gradual increase in reflectance
(mean 0.026 ± 0.021, averaging reflectances at 400, 500,
600 and 700 nm), reaching half maximum at
59.9 ± 5.6 °C and maximum at 67.1 ± 5.5 °C (Fig. 4).
The decrease in reflectance associated with gelatinisation
did not occur until later, and the initial increase in reflec-
tance was thought to be caused by protein denaturation.
 H.J. Swatland / Meat Science 77 (2007) 2–6
Fig. 4. Reflectance changes in a strip of Longissimus thoracis aponeurosis
when heated (lines from top to bottom, reflectances at 400, 500, 600 and
700 nm, respectively; from Swatland, 2006b). From 40 °C to 56 °C there is
no change in reflectance, but then reflectance increases to peak at around
68 °C.
Simultaneously, as strips of aponeurosis were heated
they also contracted (Fig. 3). Half maximum contraction
was reached at 62.8 ± 3.4 °C and the maximum was at
69.2 ± 3.2 °C. These temperatures did not differ signifi-
cantly from the corresponding temperatures to reach half,
and maximum increases in reflectance (P > 0.05). The
mean contraction was 33.8 ± 15.8% of starting length
(3 cm). The lengthening of aponeurosis strips at tempera-
tures >70° was an artefact caused by the clamps failing
to hold strips securely after they softened. Unclamped
strips in the water bath remained at their contracted length
while hot.
Collagen fibres show both Mie and Rayleigh scattering
(Saidi, Jacques, & Tittel, 1995). Mie scattering is from par-
ticles >20 lm in diameter and involves visible light at all
wavelengths, whereas Rayleigh scattering is from much
smaller structures and primarily involves violet and blue
light (white clouds in a blue sky are caused by Mie and
Rayleigh scattering, respectively). Normally we only see
the aponeurosis before and after cooking – not during
cooking. It starts as a bright, white, glistening structure
and may finish as a dull, translucent layer of gelatine. Thus,
it was not anticipated the reflectance of the aponeurosis
would initially increase before it subsequently decreased
as gelatinisation occurred. In retrospect, this makes sense.
Increases in light scattering accompany protein denatur-
ation in many food systems, as when egg albumin changes
from transparent to white on heating. An increase in light
scattering is also the predominant internal optical change
occurring when meat is cooked (Swatland, 1989a). The out-
side of the meat, of course, is darkened by drying and a
variety of chemical changes and its reflectance decreases
as it is cooked.
In beef semitendinosus during cooking, the gelatinisation
of the perimysium may contribute to tenderness below
60 °C (Christensen, Purslow, & Larsen, 2000). But optical
measurements of perimysial gelatinisation in beef indicate
67 °C as the point of gelatinisation (Swatland, 1989b), clo-
ser to the reflectance and contraction peaks seen for apo-
neurosis in Figs. 4 and 5. Perhaps we are seeing
differences caused by methodology, but there might be
5
Fig. 5. Contraction in a strip of Longissimus thoracis aponeurosis when
heated, matching Fig. 4 (from Swatland, 2006b). The contraction seen as
movement on the y-axis starts around 56 °C and peaks around 70 °C just
before gelatinisation releases the strip from its clamps.
other factors involved, such as differences caused by
mechanical restraint.
4. Importance of mechanical restraint
This just about brings me up to date (February 2007)
with published experiments, but leaves us with a myriad
of complex possible interactions, and the floor will soon
be open for discussion and the floating of new ideas. One
intriguing factor many of us have ignored is the importance
of mechanical restraint on laniary septa during cooking. I
came across this by accident. I was testing a new com-
puter-operated polarising microscope I had just built. I
found a strong signal from raw aponeurosis – as one would
expect, because it has a highly anisotropic structure causing
strong birefringence. I cooked the samples and re-measured
them, but their flatness had to be maintained, so I clamped
them between two metal plates during cooking. To my sur-
prise, far beyond the normal temperature of denaturation,
the samples maintained their degree of internal structure
and optical properties. I was expecting fully gelatinised
samples with an isotropic structure and no birefringence,
as occurs when perimysial septa are cooked (Swatland,
1989b).
A literature search showed this effect is well known.
Mechanical restraint preventing heat-induced contraction
protects collagen from thermal denaturation (Wells, Thom-
sen, Jones, Baek, & Humphrey, 2005). This was confirmed
for bovine aponeurosis by cutting parallel strips (length
10 cm, width 5 mm, depth 3 mm) and clamping them at
one end and at half way along their length. Thus, after they
had been heated in distilled water to 70 °C, it was possible
to compare shear forces between restrained and unre-
strained parts. Using a Warner–Bratzler shear test (500 N
load cell, Dillon BFG-500 N, Weigh-Tronix, Fairmont,
Minnesota; compression tester, Dillon SnapShot, capacity
1 kN, range 29 cm, velocity 4 mm s 1), restrained parts of
strips were much tougher (31.9 ± 16.6 N) than unre-
strained parts (19.3 ± 7.0 N; P < 0.005 with paired t-test,
n = 14 pairs).
This could be an important factor in explaining events
occurring as a roast is cooked. For example, an aponeuro-
6 H.J. Swatland / Meat Science 77 (2007) 2–6
sis near a cut meat surface might have minimal gelatinisa-
tion because it is exposed to dry heat, whereas deeper in the
roast formation of steam might accelerate gelatinisation –
but not if the aponeurosis is mechanically restrained. Does
this help explain the popularity of boned and rolled roasts?
Bones connected to laniary septa might retard gelatinisa-
tion and reduce overall tenderness.
5. Conclusion
The classical approach to measuring meat toughness in
cores of meat taken from within muscles and carefully
stripped of extraneous connective tissue is important, but
is only part of the story. We also need to look at the strat-
ification of tissues in whole cuts of meat. This may help us
understand consumer responses to whole steaks and roasts,
as well as the problem of sampling error when making on-
line measurements from intact carcass. Is it possible to
breed or feed our meat animals to have easily gelatinised
laniary septa? Consumers would be delighted if we could.
References
Bailey, A. J., & Sims, T. J. (1977). Meat tenderness: Distribution of
molecular species of collagen in bovine muscle. Journal of the Science
of Food and Agriculture, 28, 565–570.
Belew, J. B., Brooks, J. C., McKenna, D. R., & Savell, J. W. (2003).
Warner–Bratzler shear evaluations of 40 bovine muscles. Meat
Science, 64, 507–512.
Bendall, J. R., & Restall, D. J. (1983). The cooking of single myofibres,
small myofibre bundles, and muscle strips from beef M. psoas and M.
sternomandibularis at varying rates and temperatures. Meat Science, 8,
93–117.
Brooks, J. C., & Savell, J. W. (2004). Perimysium thickness as an indicator
of beef tenderness. Meat Science, 67, 329–334.
Chien, J. C. W. (1975). Solid-state characterization of the structure and
property of collagen. Journal of Macromolecular Science, Review of
Macromolecular Chemistry, C12(1), 1–80.
Christensen, M., Purslow, P. P., & Larsen, L. M. (2000). The effect of cooking
temperature on mechanical properties of whole meat, single muscle fibres
and perimysial connective tissue. Meat Science, 55, 301–307.
Dutson, T. R., Hostetler, R. L., & Carpenter, Z. L. (1976). Effect of
collagen levels and sarcomere shortening on muscle tenderness. Journal
of Food Science, 41, 863–866.
Gruber, S. L., Tatum, J. D., Scanga, J. A., Chapman, P. L., Smith, G. C.,
& Belk, K. E. (2006). Effects of postmortem aging and USDA quality
grade on Warner–Bratzler shear force values of seventeen individual
beef muscles. Journal of Animal Science, 84, 3387–3396.
Janz, J. A. M., Aalhus, J. L., Dugan, M. E. R., & Price, M. A. (2006). A
mapping method for the description of Warner–Bratzler shear force
gradients in beef Longissimus thoracis et lumborum and Semitendi-
nosus. Meat Science, 72, 79–90.
Janz, J. A. M., Aalhus, J. L., & Price, M. A. (2006). The effect of epimysial
connective tissue on factors related to tenderness of beef semitendi-
nosus. Journal of Muscle Foods, 17, 43–55.
Lewis, C. T. (1891). An Elementary Latin Dictionary. Oxford University
Press.
Light, N., Champion, A. E., Voyle, C., & Bailey, A. J. (1985). The role of
epimysial, perimysial and endomysial collagen in determining texture
in six bovine muscles. Meat Science, 13, 137–149.
Loyd, E. J., & Hiner, R. L. (1959). Relation between hydroxyproline of
alkali-insoluble protein and tenderness of bovine muscle. Agricultural
and Food Chemistry, 7, 860–862.
Martens, H., Stabursvik, E., & Martens, M. (1982). Texture and colour
changes in meat during cooking related to thermal denaturation of
muscle proteins. Journal of Texture Studies, 13, 291–309.
Saidi, I. S., Jacques, S. L., & Tittel, F. K. (1995). Mie and Rayleigh
modeling of visible-light scattering in neonatal skin. Applied Optics, 34,
7410–7418.
Smith, S. H., & Judge, M. D. (1991). Relationship between pyridinoline
concentration and thermal stability of bovine intramuscular collagen.
Journal of Animal Science, 69, 1989–1993.
Swatland, H. J. (1989a). Effect of temperature (0–80 °C) on the interior
reflectance of ovine sternomandibularis muscle. International Journal
of Food Science and Technology, 24, 503–510.
Swatland, H. J. (1989b). Thermal denaturation of perimysial collagen in
meat measured by polarized light microscopy. Journal of Food Science,
55, 305–311.
Swatland, H. J. (1995). On-line evaluation of meat. Lancaster, Pennsylva-
nia: Technomic Publishing.
Swatland, H. J. (2005). A method for simultaneous fluorometry and
rheology of connective tissue in bulk meat. Meat Science, 70,
605–611.
Swatland, H. J. (2006a). Stratification of connective tissue toughness in
beef roasts assessed by simultaneous fluorometry and penetrometry.
Food Research International, 39, 1106–1109.
Swatland, H. J. (2006b). Thermal denaturation of bovine longissimus
thoracis aponeurosis, correlating contraction with optical changes.
Canadian Journal of Animal Science, 86, 221–224.
Tornberg, E. (2005). Effects of heat on meat proteins – Implications on
structure and quality of meat products. Meat Science, 70, 493–508.
Vangsness, C. T., Mitchell, W., Nimni, M., Erlich, M., Saadat, V., &
Schmotzer, H. (1997). Collagen shortening. An experimental approach
with heat. Clinical and. Orthopaedic Related Research, 337, 267–271.
Voisey, P. W. (1976). Engineering assessment and critique of instruments
used for meat tenderness evaluation. Journal of Texture Studies, 7,
11–48.
Wells, P. B., Thomsen, S., Jones, M. A., Baek, S., & Humphrey, J. D.
(2005). Histological evidence for the role of mechanical stress in
modulating thermal denaturation of collagen. Biomechanics and
Modeling in Mechanobiology, 4, 201–210.
Wulf, D. M., Emnett, R. S., Leheska, J. M., & Moeller, S. J. (2002).
Relationships among glycolytic potential, dark cutting (dark, firm, and
dry) beef, and cooked beef palatability. Journal of Animal Science, 80,
1895–1903.
 MEAT
SCIENCE
Meat Science 77 (2007) 7–16
www.elsevier.com/locate/meatsci

Mechanisms controlling pork quality development: The

biochemistry controlling postmortem energy metabolism
T.L. Scheffler, D.E. Gerrard *

Department of Animal Sciences, Purdue University, 915 W. State Street, West Lafayette, IN 47907, United States

Received 23 March 2007; received in revised form 2 April 2007; accepted 2 April 2007

Abstract

Pale, soft and exudative (PSE) pork represents considerable economic losses for the industry due to its limited functionality and unde-
sirable appearance. During the past several decades, exhaustive research covering various aspects of the food chain has established geno-
typing procedures, recommended handling practices, and quality indicators in order to reduce the incidence of inferior pork quality.
Despite these efforts, there is still a relatively high occurrence of PSE pork. Development of pork quality attributes is largely governed
by the rate and extent of postmortem pH decline. The combination of high temperature at low pH or abnormally low ultimate pH causes
denaturation of sarcoplasmic and myofibrillar proteins, resulting in paler color and reduced water holding capacity. The pH decline is
closely related to muscle energy availability and demand at or around slaughter. The postmortem degradation of glycogen through gly-
cogenolysis and glycolysis provides ATP to help meet energy demand and decreases pH by generating lactate and H+. Therefore, the flux
of metabolites through glycolysis, the involvement of energy signaling pathways that modulate glycolytic activity, and the inherent
metabolism of different fiber types are critical factors influencing pH decline and pork quality. Further, recent work implicates adenosine
monophosphate-activated protein kinase (AMPK) as a major energy sensor for the cell, and thus may be involved in the control of post-
mortem metabolism. The intent of this paper is to review the biochemistry controlling postmortem energy metabolism in pig muscle and
explore new information generated using genetic mutations in order to define the fundamental mechanisms controlling the transforma-
tion of muscle to meat.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Pork; Biochemistry; Glycolysis

1. Introduction texture due to enhanced water binding capability. Con-

Increased health-consciousness among consumers has
led to significant changes in pork carcass attributes over
the past several decades. Swine producers have focused
selection pressure on improving carcass merit, resulting in
increased lean growth and higher yielding carcasses. How-
ever, extreme variation in pork quality, ranging from dark,
firm, and dry (DFD) to pale, soft and exudative (PSE)
pork, has also become more prevalent. DFD pork has a
dark, unattractive appearance and a firm, dry, and sticky
*
Corresponding author. Tel.: +1 765 494 8280; fax: +1 765 494 6816.
E-mail address: dgerrard@purdue.edu (D.E. Gerrard).
versely, PSE pork is characterized by pale color, soft tex-
ture, and low water-holding capacity, and has limited
functionality in further processing. Surveys indicated that
in pork produced in the USA in 2003, 15.5% was PSE
(Stetzer & McKeith, 2003). The inferior quality attributes
of PSE pork are estimated to cost the pork industry $100
million annually (Carr, Kauffman, Meeker, & Meisinger,
1997).
Due to the considerable economic losses, PSE pork has
been a significant source of research attention. Early stud-
ies established that the development of PSE pork was lar-
gely due to an increased rate of early postmortem
glycolysis, indicated by elevated muscle temperature and
rapid pH decline (Briskey, 1964). More recently, the extent

0309-1740/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.024
8 T.L. Scheffler, D.E. Gerrard / Meat Science 77 (2007) 7–16
of glycolysis and ultimate pH have been implicated (Sellier
& Monin, 1994). Further research has established measur-
able indicators of quality, and yielded molecular diagnostic
techniques and control methods. Yet, despite the extensive
efforts to improve pork quality, Cassens (2000) concluded
that little progress has been made. A greater understanding
of postmortem metabolism during the conversion of mus-
cle to meat and the mechanisms of enhanced and extended
glycolysis are critical to characterizing pork quality
development.
2. Conversion of muscle to meat
Transitions between rest and exercise require that mus-
cle is a dynamic tissue with the ability to adapt to dramatic
changes in energy expenditure. Muscle contraction is a
rapid, energetically demanding process that requires the
splitting of adenosine triphosphate (ATP) in order to meet
energy requirements for muscle contraction and relaxation,
sequestration of calcium and maintenance of ion gradients.
The most efficient means of generating ATP is through
mitochondrial oxidative metabolism, but the ATP concen-
tration in muscle supplies enough energy for only a few
twitches. Therefore, additional reactions must be able to
buffer energy levels when other metabolic processes are
not able to meet ATP demand.
Phosphocreatine (PCr) is present at a higher concentra-
tion (18–19 lmole/g) than ATP (6.6–6.8 lmole/g) in pig
longissimus (Bendall, 1973) and can be quickly utilized in
a chemical reaction to rephosphorylate ADP to ATP by
the enzyme creatine kinase. Additionally, myokinase cata-
lyzes the conversion of two adenosine diphosphate (ADP)
to adenosine monophosphate (AMP) and ATP. Together,
these reactions allow muscle to quickly increase energy pro-
duction in order to keep cellular ATP constant and main-
tain homeostasis. Aerobic metabolism and PCr are able
to satisfy energy demand when oxygen is adequate and
muscle is working slowly, but if contraction proceeds rap-
idly, oxygen becomes limiting and the muscle will resort to
anaerobic glycolysis to supply the energy needed for con-
traction. Similarly, the removal of blood during the slaugh-
tering process eliminates the oxygen supply and forces
metabolically active muscles to adapt to new physiological
circumstances.
The basic biochemical reactions and physical changes
underlying the conversion of muscle to meat are well recog-
nized. ATP production is necessary to keep the muscle in
the relaxed state, but postmortem muscle has a high rate
of ATP turnover (Bate-Smith & Bendall, 1949). Initially,
ATP is replenished by the creatine kinase and myokinase-
catalyzed reactions. Once 70% of the PCr pool has been
degraded, ATP levels rapidly decline (Bendall, 1951), and
muscle glycogen must be degraded (glycogenolysis) and
metabolized in anaerobic glycolysis in order to rephospho-
rylate ADP to ATP and prevent the formation of perma-
nent actomyosin crossbridges. Anaerobic glycolysis also
produces lactate, H+, and heat. Because postmortem mus-
cle does not have the means to remove waste products, lac-
tate and H+ accumulate and lower the pH. As the
breakdown of ATP exceeds its synthesis by glycolysis, less
ATP is available and the formation of actomyosin bonds
shortens sarcomeres and increases muscle tension, signal-
ing the onset of rigor mortis. Rigor mortis is complete
when the ATP supply is exhausted; thus, actomyosin cross-
bridges cannot be broken and the muscle is relatively inex-
tensible. Muscle tension will eventually decrease with
postmortem storage as a result of degradation of myofibr-
illar proteins and loss of structural integrity.
The rate and extent of pH decline during the conversion
of muscle to meat significantly impact the development of
fresh meat quality attributes (Fig. 1). Normally, pH
declines gradually from 7.4 in living muscle to roughly
5.6–5.7 within 6–8 h of postmortem and then has an ulti-
mate pH at 24 h (pHu) of about 5.3–5.7 (Briskey & Wis-
mer-Pedersen, 1961). However, muscles with a hastened
pH decline exhibit rapid glycolysis and produce large
amounts of heat, which slows carcass chilling. This results
in a rapid muscle pH decline to less than 6.0 during the first
hour after slaughter and an ultimate pH of 5.3–5.7. The
onset of rigor mortis at high temperature and low pH
causes the denaturation of approximately 20% of the sarco-
plasmic and myofibrillar proteins (Honikel & Kim, 1986),
and the reduction in myosin head length is sufficient to
draw thick and thin filaments closer together, leading to
increased expulsion of water (Offer et al., 1989). Greater
precipitation of sarcoplasmic proteins is largely responsible
for paler pork color, while denaturation of myofibrillar
proteins explains the reduced water holding capacity in
PSE muscle (Joo, Kauffman, Kim, & Park, 1999). In con-
trast, an extended pH decline proceeds at a normal rate
but continues to a low pHu of roughly 5.3–5.5, resulting
in ‘‘acid meat’’. Abnormally low pH reduces the net charge
of myofibrillar proteins, and the attraction moves filaments
closer together and forces water out of the myofilament lat-
tice (Irving, Swatland, & Millman, 1989). Moreover, sarco-
plasmic protein solubility declines with decreasing pHu and
contributes to paler pork color (Joo et al., 1999). The rate
and extent of postmortem pH decline significantly influence
Fig. 1. Various pH declines occurring postmortem and associated pork
quality characteristics. Modified from Briskey (1964).
 T.L. Scheffler, D.E. Gerrard / Meat Science 77 (2007) 7–16
protein characteristics and thus critically affect pork quality
development.
3. Major genes and aberrant postmortem metabolism
Postmortem acidification is closely related to the energy
status of the muscle at slaughter. The decline in pH
depends on the initial concentration of PCr and glycogen
(Bendall, 1951), which may vary widely under different
experimental and commercial conditions. A variety of fac-
tors, such as stress, stunning method, muscle type, and diet,
affect muscle energy level and its utilization postmortem.
However, in order to develop more comprehensive and
sophisticated strategies for reducing the variation found
in pork quality, the biochemistry underlying postmortem
metabolism must be fully defined.
3.1. The halothane gene
Early studies utilized certain breeds and lines within
breeds in order to model PSE development. This tendency
to produce PSE pork was closely associated with porcine
stress syndrome, a condition synonymous with human
malignant hyperthermia (MH). Eikelenboom and Mink-
ema (1974) demonstrated that MH susceptible swine dis-
play hypermetabolism, elevated body temperature, and
muscle rigidity upon exposure to the anesthetic halothane.
Stress and excitement are also sufficient to provoke an MH
episode, and can result in the enhanced rate of postmortem
metabolism associated with PSE pork. Exposure to halo-
thane gas was used to screen for stress susceptible swine,
and the genetic component has since been referred to as
the halothane (HAL) gene. The HAL locus has two alleles:
the normal dominant allele (N) and the mutant recessive
allele (n).
MH susceptibility associated with the HAL gene is due
to a defect in the ability of the muscle to adequately regu-
late myoplasmic Ca2+ concentration. The causative poly-
morphism is the single substitution of T for C at
nucleotide 1843 in the skeletal muscle ryanodine receptor
(RYR1), leading to an amino acid change from arginine
to cysteine (Fujii et al., 1991). The RYR1, or Ca2+-release
channel, is the primary mechanism by which Ca2+ stored in
the sarcoplasmic reticulum terminal cisternae is released
into the sarcoplasm to initiate muscle contraction. The
Ca2+ release channels in MH susceptible pigs are hypersen-
sitive to agents that stimulate opening (O’Brien, 1986), thus
allowing longer open time probability and resulting in
enhanced Ca2+ release and greater twitch tension (Mickel-
son et al., 1988; Mickelson et al., 1989). The abnormal
channels flood the cell with Ca2+ and overpowers the sar-
coplasmic reticulum ATPase pumps that resequester cyto-
plasmic Ca2+, resulting in sustained contraction and
metabolism.
This defect in Ca2+ concentration has important conse-
quences for production and meat quality traits. Pigs homo-
zygous for the HAL gene have higher carcass yield and
9
lean percentage (Aalhus, Jones, Robertson, Tong, &
Sather, 1991; Herfort Pedersen et al., 2001; Leach, Ellis,
Sutton, McKeith, & Wilson, 1996; Rempel, Lu, Mickelson,
& Louis, 1995). The leaky Ca2+ release channels of HAL
mutant pigs may contribute to leanness and heavy mus-
cling by causing spontaneous muscle contraction and
greater energy utilization, leading to work induced muscle
hypertrophy and limiting fat deposition (MacLennan &
Phillips, 1992).
The positive effect of the mutant allele on performance is
negated by an increased risk for stress-induced death and
high susceptibility to acute stress prior to slaughter, which
may be manifested in an accelerated rate of pH decline and
the production of PSE pork. An MH episode triggers a
massive stimulation of aerobic and anaerobic metabolism
that depletes energy resources, generates metabolic waste
products, and disrupts cellular and extracellular ion bal-
ances. The larger muscle fiber area and lower capillary den-
sity of homozygous mutant pigs (Essen-Gustavsson,
Karlstrom, & Lundstrom, 1992) further compromises their
ability to cope with metabolic stress, and contributes to an
enhanced rate of glycogenolysis and glycolysis, evidenced
by reduced ATP, PCr, and glycogen levels, and higher lac-
tate levels and lower muscle pH at exsanguination (Essen-
Gustavsson et al., 1992; Fernandez, Neyraud, Astruc, &
Sante, 2002; Lundstrom, Essen-Gustavsson, Rundgren,
Edfors-Lilja, & Malmfors, 1989). Clearly, the mutant
HAL genotype increases the frequency of PSE meat by
accelerating the early postmortem rate of glycogenolysis
and glycolysis, leading to a more rapid pH decline.
3.2. The RN gene
Greater availability of glycogen may contribute to an
extended pH decline by providing additional substrate for
glycolysis. Sayre, Briskey, and Hoekstra (1963) showed
the Hampshire breed had more than twice the amount of
muscle glycogen compared to other breeds. Monin and Sel-
lier (1985) suggested that the higher muscle glycogen stores
were mainly responsible for an extended pH decline post-
mortem, resulting in pale meat with a low pHu, low water
holding capacity, and reduced technological yield, and
referred to this as the ‘‘Hampshire effect’’ in order to distin-
guish it from the PSE meat produced from a rapid pH
decline at elevated temperatures. Monin & Sellier (1985)
recommended that the muscle metabolites glycogen, glu-
cose, glucose 6-phosphate, and lactate should be combined
into a single measure termed glycolytic potential (GP) in
order to reflect all of the compounds in the muscle capable
of being converted to lactate, thus indicating the muscle’s
capacity for extended postmortem glycolysis. A single
point mutation, Rendement Napole (RN), is responsible
for the elevated muscle GP (>180–200 lmol/g muscle),
extended pH decline, and greatly reduced technological
yield that was first found in the Hampshire breed (Le
Roy et al., 2000). There are two alleles, the dominant allele
(RN) associated with elevated muscle glycogen and infe-
10 T.L. Scheffler, D.E. Gerrard / Meat Science 77 (2007) 7–16
rior quality, and the wild-type recessive allele (rn+)
(Le Roy, Naveau, Elsen, & Sellier, 1990). Additional alleles
in the PRKAG3 gene may play a significant role in glyco-
gen variation, including the 199I allele which is correlated
with lower glycogen and lactate, higher ham and loin pH
and improved color scores, resulting in improved meat
quality (Ciobanu et al., 2001).
The mutation responsible for the RNgene is an R200Q
substitution in the PRKAG3 gene that encodes for the
muscle specific isoform of the regulatory c-subunit of aden-
osine monophosphate activated protein kinase (AMPK)
(Milan et al., 2000). AMPK monitors the energy charge
of the muscle fiber and prevents high energy phosphate
depletion. AMPK is allosterically inhibited by PCr and
0
ATP and activated by 5 AMP. Therefore, increased rates
of ATP utilization during muscle contraction activate
AMPK, which is hypothesized to phosphorylate proteins
involved in triggering fatty acid oxidation and glucose
uptake (reviewed by Winder, 2001). Transfection experi-
ments performed with mice revealed that the PRKAG3
R225Q is a loss of function mutation that eliminates allo-
steric regulation by ATP/AMP resulting in increased basal
AMPK activity (Barnes et al., 2004). Additionally, the
adenosine monophosphate kinase gamma-3 subunit
(AMPKc3) is more highly expressed in mouse fast twitch
white glycolytic muscle compared to fast twitch red oxida-
tive muscle, but is virtually undetectable in red muscle (Yu,
Fujii, Hirshman, Pomerleau, & Goodyear, 2004).
The expression profiling of AMPKc3 supports evidence
that the RN mutation influences the compositional and his-
tochemical traits and metabolic enzyme activities in a mus-
cle-type dependent manner. Lebret et al. (1999) found the
RN mutation had no effect on red semispinalis capitis mus-
cle. However, enhanced glucose uptake due to an increase
in the translocation of GLUT4 to surface membranes
(Kurth-Kraczek, Hirshman, Goodyear, & Winder, 1999)
and faster resynthesis of glycogen after exercise (Anders-
son, 2003) may explain the up to 70% increase in muscle
glycogen localized mostly to abnormally enlarged sarco-
plasmic compartments of type IIB white fibers observed
in RN carriers (Monin, Brard, Vernin, & Naveau, 1992;
Estrade, Vignon, Rock, & Monin, 1993). Furthermore, a
twofold increase in branching enzyme activity and a ten-
dency toward increased glycogen synthase activity may
contribute to elevated glycogen levels and differences in
the molecular structure of glycogen (Estrade, Ayoub, Tal-
mant, & Monin, 1994).
The relatively high glycogen content of RN carriers
might be expected to further enhance glycolytic metabolism
in IIB fibers. Curiously, RN carriers possess enhanced
oxidative metabolism, indicated by higher citrate synthase
and b-hydroxyacyl coenzyme A dehydrogenase activities,
decreased lactate dehydrogenase activity, higher relative
area of IIA fibers and lower relative area of IIB fibers (Leb-
ret et al., 1999). Mice with the PRKAG3 R225Q mutation
also exhibited higher skeletal muscle oxidative capacity
without altered fiber type composition, suggesting that
the elevated glycogen content may be an indirect conse-
quence of altered muscle oxidation, and the role of the
AMPKc3 isoform may be to ensure that glycogen content
in muscle is restored by maintaining a high glycolytic
potential through shifting the metabolic fate of fuel toward
fat oxidation and glycogen storage (Barnes et al., 2004).
Nonetheless, high glycogen levels in RNcarriers do not
appear to alter the rate of glycogen utilization. RNcarri-
ers and wild type pigs exhibit similar glycogen degradation
during exercise (Andersson, 2003) as well as similar rates of
early postmortem pH decline (Monin & Sellier, 1985).
Estrade et al. (1994) reported no differences in either glyco-
gen phosphorylase or debranching enzyme activity, further
supporting that the RN allele increases glycogen avail-
ability without altering the rate of utilization. Thus, the
mutant RN genotype generates pork of inferior technolog-
ical quality by increasing glycolytic potential and extending
postmortem pH decline, resulting in pork with a low pHu.
4. Rate limiting enzymes
Clearly, glycogen degradation through glycolysis
impacts pork quality development. Enzymes catalyzing
the reactions of glycolysis influence the rate and extent of
pH decline by directly controlling the conversion of metab-
olites through the pathway (Fig. 2). The properties of the
rate limiting enzymes glycogen phosphorylase, phospho-
fructokinase, and pyruvate kinase have been studied in
an effort to explain the aberrant glycolysis leading to PSE
development.
Glycogen phosphorylase and glycogen debranching
enzyme catalyze the complete degradation of glycogen.
Glycogen phosphorylase cleaves its substrate by addition
of inorganic phosphate at a-1,4 linkages of the outer chains
of the glycogen molecules, yielding glucose 1-phosphate.
Then, phosphoglucomutase catalyzes the isomerization of
glucose 1-phosphate to glucose 6-phosphate, which can
then proceed through glycolysis. Once phosphorylase
reaches the fourth glucose from the branch point, a trans-
ferase shifts the maltotriosyl group to the main chain, and
glycogen debranching enzyme breaks the a-1,6 linkages,
releasing free glucose. The main chain and remaining outer
branches are again susceptible to breakdown by
phosphorylase.
Sayre et al. (1963) found similar levels of phosphorylase
in swine exhibiting fast and slow rates of glycolysis
postmortem. However, in skeletal muscle, glycogen
phosphorylase exists in two forms: the less active, non-
phosphorylated form (GP b) and the more active, phos-
phorylated form (GP a). Hence, differences in the fraction
of phosphorylase in the a form could plausibly explain the
abnormal glycolysis observed in PSE muscle. Early in vitro
studies indicated a strong relationship between pHu and
the amount of phosphorylase in the a form (Scopes,
1974). In support, Ono, Topel, Christian, and Althen
(1977) demonstrated an enhanced GP a activity in PSE tis-
sue, and Schwagele and Honikel (1988) determined a
 T.L. Scheffler, D.E. Gerrard / Meat Science 77 (2007) 7–16 11

Glycogen activity strongly decreased when the temperature decreased

from 39 and 42 °C to 4 and 15 °C. Therefore, the activity of
glycogen phosphorylase debranching enzyme does not block rapid glycolysis and
pH decrease when the temperature is high, but rapid cool-
Glucose Glucose 1 - Phosphate ing could limit the activity and thus limit glycogenolysis.
Phosphofructokinase (PFK) catalyzes the conversion of
hexokinase phosphoglucomutase
fructose 1-phosphate and ATP to fructose 1,6-bisphos-
Glucose 6-Phosphate phate and ADP, which is the committed step in the glyco-
lytic pathway. In an in vitro glycolytic system, PFK was
phosphoglucose isomerase
still active under the extreme PSE conditions of 37 °C
Fructose 6-Phosphate and pH 5.35 (Scopes, 1974). Moreover, total and specific
PFK activities were not significantly different in pigs with
phosphofructokinase (PFK)
45 min longissimus muscle pH values ranging from 5.3 to
Fructose 1,6- Bisphosphate 6.8 (Schwagele & Honikel, 1988). Therefore, activation of
PFK does not readily explain the differences in glycolytic
aldolase
rates between normal and PSE muscles. Interestingly, Alli-
son, Bates, Booren, Johnson, and Doumit (2003) observed
Glyceraldehyde Dihydroxyacetone an inverse relationship between PFK capacity and fluid
3-Phosphate phosphate loss, and reasoned that PFK may become partially dena-
Triose phosphate
isomerase tured and inactivated by 20 min postmortem in muscles
that experience a rapid pH decline. Therefore, it appears
glyceraldehyde 3-phosphate dehydrogenase
that increased PFK activity is not the reason for enhanced
Bisphosphoglycerate glycolysis in PSE muscle.
Pyruvate kinase catalyzes the irreversible conversion of
phosphoglycerate kinase phosphoenolpyruvate and ADP to pyruvate and ATP. Sch-
3-Phosphoglycerate wagele, Haschke, Honikel, and Krauss (1996b) demon-
strated that, compared to normal muscle, pyruvate kinase
phosphoglycerate mutase
isolated from PSE muscle of halothane susceptible pigs
2-Phosphoglycerate exhibited a tenfold increase in phosphoenolpyruvate utili-
zation. Interestingly, the activity of pyruvate kinase from
enolase normal muscle was low at pH 5.5, whereas the enzyme
Phosphenolpyruvate from PSE muscle maintained 70% of its maximum activity
under these acidic conditions. Schwagele et al. (1996b)
pyruvate kinase determined that this shift in pyruvate kinase activity was
Pyruvate due to phosphorylation of the enzyme, resulting in an addi-
tional, more acid stable isoform. It is not clear whether the
lactate dehydrogenase additional, more stable isoform is present in live animals or
Lactate if it is generated under postmortem conditions. Neverthe-
less, in pigs free of the HAL gene, pyruvate kinase capacity
Fig. 2. Enzymes and metabolic intermediates of the glycolytic pathway.
was not correlated with longissimus pH, purge, drip loss or
paleness, and in all samples, pyruvate kinase lost more than
higher total activity (GP a and GP b) in PSE-prone mus- 88% of its activity at pH 5.5 compared to pH 7.0 (Allison
cles. However, Ensinger, Rogdakis, Muller, and Faber et al., 2003). Although the role of pyruvate kinase in HAL
(1982) could not establish differences in the activities of sensitive pigs is not completely understood, it appears that
GP a and GP b in muscles of normal and PSE muscles. enhanced pyruvate kinase capacity is not responsible for
More recently, Schwagele, Buesa, and Honikel (1996a) altered postmortem metabolism in PSE muscle from pigs
reported no significant differences in the structural and free of the HAL gene.
kinetic characteristics of GP a and GP b from muscles of For the most part, the inherent properties of the rate
normal versus HAL sensitive pigs. Thus, the characteristics limiting enzymes are not different in normal versus PSE
of glycogen phosphorylase are not likely to be the main muscles, and thus other mechanisms must be responsible
factors facilitating altered metabolism in PSE muscle. for the aberrant glycogenolysis and glycolysis observed in
Because postmortem glycogenolysis may stop in the PSE muscles. Muscle possesses a complex system of both
presence of residual glycogen, it has been speculated that feedforward and feedback regulation in order to precisely
glycogen debranching enzyme may influence the rate and modulate metabolism, and differences in relative levels of
extent of glycogenolysis and glycolysis. Kyla-Puhju, Ruus- glycolytic regulators could contribute to altered postmor-
unen, and Puolanne (2005) reported that the activity of tem glycolysis. Specifically, the modulators of rate limiting
debranching enzyme was only weakly affected by pH, but enzymes could manipulate enzyme activity according to
12 T.L. Scheffler, D.E. Gerrard / Meat Science 77 (2007) 7–16
energy demand of the cell, and this would affect flux
through the glycolytic pathway and contribute to altered
postmortem metabolism.
Rapid rates of postmortem glycolysis are associated
with low levels of glycogen and high levels of glucose
6-phosphate, indicative of increased phosphorylase activity
(Briskey et al., 1966; Kastenschmidt, Hoekstra, & Briskey,
1968; Moesgaard, Quistorff, Christensen, Therkelsen, &
Jorgensen, 1995). Glycogen phosphorylase activity is regu-
lated both by phosphorylation and allosteric mechanisms.
Phosphorylase kinase exists in two forms, phosphorylase
kinase a and b (PK a and PK b), which are both capable
of converting GP b to the active GP a form. Maximal acti-
vation of phosphorylase kinase is achieved through phos-
phorylation of PK b to PK a and Ca2+ binding. The
hormones epinephrine and glucagon induce increases in
cAMP that lead to phosphorylation of PK b to PK a,
and PK a is active at the concentration of Ca2+ in resting
muscle. Exposure or susceptibility to stress would permit
epinephrine mediated activation of PK a and subsequent
activation of GP a, resulting in increased glycogenolysis.
The cAMP cascade provides a link between epinephrine
and glycogen degradation, whereas Ca2+ couples muscle
contraction with glycogenolysis (Drummond, Harwood,
& Powell, 1969). Compared to PK a, PK b requires a
higher Ca2+ concentration for activity (Connett & Sahlin,
1996). The Ca2+ concentration in the sarcoplasm during
contraction is sufficient to activate PK b, resulting in con-
version of GP b to the more active GP a form. Preslaughter
stress places an increased demand on muscles for contrac-
tion, and this could also be responsible for increased glyco-
genolysis by mediating the conversion of GP b to GP a by
means of Ca2+. Similarly, in the case of the HAL gene, the
enhanced Ca2+ release may trigger activation of phosphor-
ylase and stimulate glycogenolysis.
The relative concentration of several modulators related
to cellular energy charge and substrate availability regulate
the activity of GP b. The levels of inhibitors, including
ATP, ADP, and glucose 6-phosphate, are usually sufficient
to inhibit GP b in resting muscle, but their effect may be
overcome by increases in AMP and inosine monophos-
phate (IMP), and lead to GP b activation (Connett & Sah-
lin, 1996). Furthermore, the sensitivity of GP b for each
factor is dependent on the concentration of substrate (gly-
cogen and inorganic phosphate) and product (glucose 1-
phosphate). In contrast, GP a is active in the absence of
AMP, although activity is enhanced by low concentrations
of AMP and high concentrations of IMP. The other allo-
steric modulators have little effect on GP a activity.
In muscles exhibiting normal rates of glycolysis, glucose
6-phosphate levels tend to decrease during the first hour
postmortem and increase thereafter (Kastenschmidt et al.,
1968; Hammelman et al., 2003). This suggests that glyco-
gen phosphorylase may be unable to supply adequate glu-
cose 6-phosphate for the subsequent reactions of the
glycolytic pathway, resulting in a rapid utilization of the
glucose 6-phosphate pool. If animals exhibiting ‘‘slow’’
rates of glycolysis were assumed to experience minimal
antemortem stress, glycogen phosphorylase would exist
predominantly in the b form. Moreover, following exsan-
guination, the relatively high levels of ATP and low levels
of AMP (Kastenschmidt et al., 1968) and IMP (Klont &
Lambooy, 1995) would be insufficient to activate GP b.
After the first hour postmortem, changes in the relative
concentrations of allosteric activators may permit GP b
activation. Therefore, the accumulation of G6P after one
hour may be the result of increased glycogen breakdown
due to glycogen phosphorylase, or a decrease in the activity
of PFK or other enzymes downstream.
The regulation of the other rate limiting enzyme, PFK is
less complex. Similar to GP b, PFK activity depends largely
on the energy status of the muscle cell. As the ATP/AMP
ratio decreases, PFK activity is stimulated. PFK is also
activated by hexose bisphosphates and inorganic phos-
phate. Moreover, ATP is required for the transfer of a
phosphate group from ATP to fructose 6-phosphate.
Rapid depletion of ATP in fast glycolyzing muscles (Kas-
tenschmidt et al., 1968) could compromise the ability of
PFK to catalyze the formation of fructose 1,6-bisphos-
phate. This would suggest PFK is exerting some glycolytic
control. However, in the case of slow glycolyzing muscles,
the imbalance between fructose 6-phosphate and fructose
1,6-bisphosphate becomes apparent after 60 min, and Kas-
tenschmidt et al. (1968) suggested this may be due to
decreasing muscle pH and partial inactivation of PFK.
In contrast, flux through pyruvate kinase appears to be
largely governed by the concentration of the substrates
phosphoenolpyruvate and ADP and product ATP. The
high levels of ATP present in slow glycolyzing muscles
may exert some control over PK early postmortem. How-
ever, in fast glycolyzing muscles, lactate production ceased
in the presence of residual glycogen, glucose 1-phosphate,
glucose 6-phosphate (G6P), and fructose 6-phosphate,
whereas subsequent intermediates were present at low lev-
els (Kastenschmidt et al., 1968). This implies PFK is more
likely than PK to be a site of glycolytic control.
Altogether, previous studies suggest that the relative lev-
els of glycolytic regulators contribute to altered postmor-
tem glycolysis. Specifically, the modulators of rate
limiting enzymes manipulate enzyme activity according to
cellular energy status, and this affects flux through the
pathway. Glucose 6-phosphate levels fall during the first
60 min, and subsequently increase, indicating an imbalance
between glycogenolysis and glycolysis. Therefore, different
enzymes may be rate limiting at different times during the
conversion of muscle to meat.
5. Control of postmortem glycolysis by AMPK
AMPK is particularly attractive for augmenting post-
mortem metabolism primarily because in vivo studies show
AMPK is activated in ischemic cardiac muscle and hypoxic
skeletal muscle (Kim, Solis, & Cartee, 2004). As outlined
above, when muscle (cardiac or skeletal) is placed in a
 T.L. Scheffler, D.E. Gerrard / Meat Science 77 (2007) 7–16
low oxygen environment, the primary source of energy pro-
duction defaults to anaerobic glycolysis, thereby increasing
the AMP:ATP ratio rapidly. This, in turn, allows AMP to
bind to AMPK making it a better substrate for phosphor-
ylation by its upstream kinase, AMPKK. Activated
AMPK can impact glycolysis in two manners. First,
AMPK can activate phosphorylase kinase, which then
activates glycogen phosphorylase and promotes glycogen-
olysis. Second, AMPK can phosphorylate phosphofructo-
kinase-2 which catalyzes the formation of fructose
2,6-bisphosphate. This product is an allosteric activator
of PFK-1, a key rate-limiting enzyme of glycolysis. Thus,
AMPK activation indirectly increases flux through glycol-
ysis making it a strong candidate for modulating meat
quality development.
Shen and Du (2005) first investigated the involvement of
AMPK in regulating, or modulating postmortem metabo-
lism by studying exercised wild-type and AMPK-knockout
mice treated with or without the AMP analog 5-amino-4-
imidazolecarboxamide riboside (AICAR). These scientists
showed heavily exercised mice and those exercised mice
treated with AICAR had significantly lower (more rapid)
muscle pH values at 24 h postmortem (Fig. 3). Conversely,
mice lacking a functional AMPK gene, whether exercised
or not, had higher ultimate muscle pH values, even higher
than wild-type, non-exercised mice. These data closely
reflected trends in AMPK activity suggesting that aug-
mented postmortem glycolysis in mouse skeletal muscle is
partially controlled by the status of AMPK phosphoryla-
tion antemortem. To eliminate the effects of AMPK activa-
tion in living muscle on postmortem metabolism, Du et al.
(personal communication) infused muscle with compound
C immediately antemortem. In good agreement with their
aforementioned data, postmortem glycolysis and muscle
pH decline was inhibited. These data strongly support the
role of AMPK in modulating postmortem muscle
metabolism.
At a glance, these data appeared to shed much light on
the 50 year debate as to what may be controlling extended
postmortem glycolysis in pig muscle. However, conditions
whereby mouse and pig muscle undergo postmortem
Fig. 3. Muscle pH values at various times postmortem from mice treated
differently prior to euthanasia (Shen & Du, 2005).
13
metabolism are quite different. To that end, this same
group (Shen et al., 2006a) used pre-slaughter transporta-
tion as a stressor and monitored AMPK activity and
energy status of pig muscle postmortem. Similar to the
mouse data, AMPK activity reached maximum levels
quicker in transported pig muscle than in control pigs, or
those rested after transport. Consistent with these findings,
the (AMP + IMP):ATP ratio was greater in the trans-
ported pigs arguing AMPK activation through cytosolic
AMP concentrations. These researchers subsequently
showed an association between early postmortem AMPK
activation and PSE development in commercial pigs and
suggested that the rapid glycolysis occurring in muscle of
HAL pigs was in part from increases in AMPK activation
early postmortem (Shen et al., 2006b). These data clearly
show an association between AMPK activation and aggra-
vated metabolism postmortem. However, much more work
is needed to understand exactly how AMPK works in
dying muscle tissue.
We have attempted to study muscle metabolism further
using the HAL and RN mutations separately and in com-
bination – rapid metabolism together with abundant gly-
cogen supply (HAL/RN mutant) – to provide additional
insight into the control of glycogenolysis and glycolysis
(Copenhafer, Richert, Schinckel, Grant, & Gerrard,
2006). As expected, the previously documented effects of
both genes were evident: the HAL mutation elicited has-
tened metabolism, as evidenced by lower ATP levels at
0 and 30 min, and rapid degradation of glycogen and
accumulation of lactate early postmortem compared to
control; whereas the RN mutation resulted in much
greater glycolytic potential due to elevated glycogen
levels.
Additional investigation yielded more insight regarding
biochemical events within the glycolytic pathway. Free glu-
cose, released by the action of debranching enzyme on gly-
cogen, was independently affected by HAL and RN
genotype. The HAL mutants exhibited increased glucose
accumulation indicative of more rapid glycogen degrada-
tion and early postmortem metabolism, whereas the
increases in the RN mutants occurred to a greater extent
and after 60 min postmortem. Rapid rates of glycolysis
were associated with decreased glycogen as well as high
concentrations of glucose 6-phosphate, indicating
increased phosphorylase activity. G6P concentrations
decreased in the HAL mutants during the first 30 min
and remained at a similar level thereafter. Meanwhile, in
the HAL/RN mutant, G6P concentrations tended to
increase in the first 30 min, then decrease through 60 min,
indicating the combined effects of the HAL and RN max-
imize activation of glycogen phosphorylase by enhancing
Ca2+ concentration and glycogen availability. Despite large
differences in G6P between HAL and HAL/RN mutants,
pH decline and lactate accumulation were similar early
postmortem. Glycogen phosphorylase and debranching
enzyme were capable of aggressive glycogenolysis to supply
adequate G6P for the reactions of glycolysis. Additionally,
14 T.L. Scheffler, D.E. Gerrard / Meat Science 77 (2007) 7–16
high glycogen levels did not aggravate rapid postmortem
metabolism.
From 60 min to 24 h postmortem, G6P increased only in
normal and RN mutant genotypes. The accumulation of
G6P after 60 min confirmed that glycogen phosphorylase
activity was adequate to meet the demands of glycolysis
and that either glycogen phosphorylase activity was
increased or phosphofructokinase activity was reduced.
Similar residual glycogen levels as well as the slowing of
glycolysis with time, indicated that increased G6P is likely
due to reduced phosphofructokinase activity.
Interestingly, RN mutants possessed higher phosphocre-
atine concentrations than all other genotypes (Fig. 4).
Ponticos et al. (1998) observed that a fall in PCr:creatine
ratio leads to increased AMPK activity, which results in
a concomitant decrease in muscle CK activity. The pro-
posed inactivation of CK by AMPK may be an energy con-
servation mechanism that prevents CK from consuming
ATP for the rephosphorylation of creatine. Presence of
the mutant RN genotype also resulted in an increase in
ATP concentrations. The constitutively active AMPK in
the RNmutant genotype appears to decrease CK activity
and preserve energy levels more efficiently in the first
30 min postmortem. This may have several ramifications
in extending postmortem proteolysis. First, it could pro-
long glycolysis by having greater adenosine levels in the tis-
sue longer. Second, this could delay maximal rate of
glycolysis, which would likely occur at a reduced carcass
temperature allowing for extended glycolysis. And third,
the increase in inorganic phosphate may increase the buf-
fering capacity of the muscle, and again, allow for extended
glycolysis.
The RN allele is typically associated with increased
glycolytic potential and lower ultimate pH. However, at
24 h postmortem, RN and HAL/RN mutant pigs exhibited
12
Normal
RN
10 a
HAL
HAL/RN
Phosphocreatine (umol/g)
8 by the fact that we observed greater glycolytic potential
b Given that increased GLUT-4 production, the primary
6
b b
b
4 how greater glycogen is accumulated in the muscle of
c HAL/RN mutants. At present, the only possibility is that
c c
2
0 Taken together, the aforementioned review shows that
0 min 30 min
Time
Fig. 4. LS means of postmortem longissimus muscle phosphocreatine
concentrations in halothane and RN genotypes. (Copenhafer et al., 2006).
lower pH than the normal genotype but not the HAL
mutants. Previous studies have also indicated that the
mutant HAL allele tends to lower ultimate pH. Therefore,
substrate availability alone does not explain variation in
ultimate pH. Data from this study show that elevated gly-
cogen does not aggravate rapid early postmortem glycoly-
sis, nor does it fully explain low ultimate pH. The
diminished ATP in muscles undergoing rapid glycolysis
suggests phosphofructokinase may be rate limiting during
aggressive early postmortem metabolism. Moreover, the
accumulation of glucose and glucose 6-phosphate in mus-
cles with normal rates of glycolysis suggests phosphofruc-
tokinase may be rate limiting after 1 h postmortem.
Thus, depending on the pace of postmortem metabolism,
enzymes may become rate limiting at different times during
postmortem metabolism. Regardless, this is a truly valu-
able model for studying adverse pork quality development.
Further analyses of muscle tissues from the aforemen-
tioned study showed that AMPK activation was greater
in muscle from RNpigs compared to all other genotypes
(Park et al., submitted). Curiously, classic data from Milan
et al. (2000) reported that AMPK activation in RN-pigs is
greatly reduced compared to wild type pigs. The reason for
this discrepancy is not readily apparent but may be related
to sampling time or muscle type. In our studies, samples
were collected immediately post-stunning from the longiss-
imus muscle, however, details regarding sampling proce-
dures in the Milan study are somewhat sketchy. Our
data, however, are corroborated by data from the AMPK
RN mutant mice, where increased AMPK activation is
observed in that muscle as well. Clearly, establishing
whether AMPK is activated in RN mutated pig muscle,
in either living or dying muscle is critical for our under-
standing of how AMPK may modulate postmortem muscle
metabolism.
Another intriguing observation merits discussion is that
AMPK activation is blunted in the presence of the HAL
gene (Park et al., submitted). Thus, if it is true that AMPK
modulates postmortem energy metabolism as reported by
Shen et al. (2006), then the state of AMPK prior to slaugh-
ter may be more important than what happens to the
enzyme after death. This scenario is further complicated
in the double mutant, in the absence of AMPK activation.
means of glucose uptake in the muscle, is only observed
in muscle with activated AMPK, it is difficult to imagine
phosphorylation of AMPK may not be the only mecha-
nism by which AMPK modulates energy balance in muscle.
Additional studies are needed to support this hypothesis.
postmortem metabolism in skeletal muscle is quite compli-
cated and largely affected by the state of the tissue prior to
death. Regardless of the complexity, however, numerous
genetic models, molecular techniques and commercially
 T.L. Scheffler, D.E. Gerrard / Meat Science 77 (2007) 7–16
available compounds exist that make studying this event
quite rewarding. If technologies are going to be developed
that allow the pork industry to predict quality, then a com-
prehensive understanding of the biochemical and molecu-
lar events controlling postmortem metabolism must be
known.
References
Aalhus, J. L., Jones, S. D. M., Robertson, W. M., Tong, A. K. W., &
Sather, A. P. (1991). Growth characteristics and carcass composition
of pigs with known genotypes for stress susceptibility over a weight
range of 70 to 120 kg. Animal Production, 52, 347–353.
Allison, C. P., Bates, R. O., Booren, A. M., Johnson, R. C., & Doumit, M.
E. (2003). Pork quality variation is not explained by glycolytic enzyme
capacity. Meat Science, 63, 17–22.
Andersson, L. (2003). Identification and characterization of AMPK
gamma 3 mutations in the pig. Biochemical Society Transactions, 31,
232–235.
Barnes, B. R., Marklund, S., Steiler, T. L., Walter, M., Hjalm, Goran,
Amarger, V., et al. (2004). The 5’-AMP-activated protein kinase
gamma 3 isoform has a key role in carbohydrate and lipid metabolism
in glycolytic skeletal muscle. Journal of Biological Chemistry, 279,
38441–38447.
Bate-Smith, E. C., & Bendall, J. R. (1949). Factors determining the time
course of rigor mortis. Journal of Physiology, 110, 47–65.
Bendall, J. R. (1951). The shortening of rabbit muscles during rigor
mortis: Its relation to the breakdown of adenosine triphosphate and
creatine phosphate and to muscular contraction. Journal of Physiology,
114, 71–88.
Bendall, J. R. (1973). Post mortem changes in muscle. In G. H. Bournes
(Ed.), The structure and function of muscle (pp. 227–274). New York:
Academic Press.
Briskey, E. J. (1964). Etiological status and associated studies of pale, soft,
exudative porcine musculature. Advances in Food Research, 13, 89–178.
Briskey, E. J., Kastenschmidt, L. L., Forrest, J. C., Beecher, G. R., Judge,
M. D., Cassens, R. G., et al. (1966). Biochemical aspects of post-
mortem changes in porcine muscle. Journal of Agricultural and Food
Chemistry, 14, 201–207.
Briskey, E. J., & Wismer-Pedersen, J. (1961). Biochemistry of pork muscle
structure. I. Rate of anaerobic glycolysis and temperature change
versus the apparent structure of muscle tissue. Journal of Food Science,
26, 297–305.
Carr, T. R., Kauffman, R. G., Meeker, D. L., & Meisinger, D. L. (1997).
Factors reducing pork value Pork Industry Handbook, PIH-135. West
Lafayette, IN: Purdue University Cooperative Extension Service, pp.
1–4.
Cassens, R. G. (2000). Historical perspectives and current aspects of pork
meat quality in the USA. Food Chemistry, 69, 357–363.
Ciobanu, D., Bastiaansen, J., Malek, M., Helm, J., Woollard, J., Plastow,
G., et al. (2001). Evidence for new alleles in the protein kinase
adenosine monophosphate-activated gamma3-subunit gene associated
with low glycogen content in pig skeletal muscle and improved meat
quality. Genetics, 159, 1151–1162.
Connett, R. J., & Sahlin, K. (1996). Control of glycolysis and glycogen
metabolism. In L. B. Rowell & J. T. Shepherds (Eds.), Handbook of
physiology. Exercise: Regulation and integration of multiple systems
(pp. 871–911). New York: Oxford University Press.
Copenhafer, T. L., Richert, B. T., Schinckel, A. P., Grant, A. L., &
Gerrard, D. E. (2006). Augmented postmortem glycolysis does not
occur early postmortem in AMPKc3-mutated porcine muscle of
Halothane positive pigs. Meat Science, 73, 590–599.
Drummond, G. I., Harwood, J. P., & Powell, C. A. (1969). Studies on
the activation of phosphorylase in skeletal muscle by contraction
and by epinephrine. Journal of Biological Chemistry, 244,
4235–4240.
15
Eikelenboom, G., & Minkema, D. (1974). Prediction of pale, soft,
exudative muscle with a non-lethal test for the halothane-induced
porcine malignant hyperthermia syndrome. Tijdschr Diergeneesk, 99,
421–426.
Ensinger, V. U., Rogdakis, E., Muller, E., & Faber, H. (1982). Parameters
of metabolism in halothane-negative pigs. I. Concentration of lactate,
cAMP, glycogen, and activity of phosphorylase, phosphofructokinase,
and pyruvate kinase in the m. Longissimus dorsi. Zeitschrift fur
Tierzuchtung und Zuchtungsbiologie, 99, 26–32.
Essen-Gustavsson, B., Karlstrom, K., & Lundstrom, K. (1992). Muscle
fibre characteristics and metabolic response at slaughter in pigs of
different halothane genotypes and their relation to meat quality. Meat
Science, 31, 1–11.
Estrade, M., Ayoub, S., Talmant, A., & Monin, G. (1994). Enzyme
activities of glycogen metabolism and mitochondrial characteristics in
muscles of RN-carrier pigs (Sus scrofa domesticus). Comparative
Biochemistry and Physiology Part B: Biochemistry and Molecular
Biology, 108, 295–301.
Estrade, M., Vignon, X., Rock, E., & Monin, G. (1993). Glycogen
hyperaccumulation in white muscle fibres of RN-carrier pigs. A
biochemical and ultrastructural study. Comparative Biochemistry and
Physiology Part B: Biochemistry and Molecular Biology, 104, 321–326.
Fernandez, X., Neyraud, E., Astruc, T., & Sante, V. (2002). Effects of
halothane genotype and pre-slaughter treatment on pig meat quality.
Part 1. Post mortem metabolism, meat quality indicators and sensory
traits of m. Longissimus lumborum. Meat Science, 62, 429–437.
Fujii, J., Otsu, K., Zorzato, F., de Leon, S., Khanna, V. K., Weiler, J. E.,
et al. (1991). Identification of a mutation in porcine ryanodine
receptor associated with malignant hyperthermia. Science, 253,
448–450.
Hammelman, J. E., Bowker, B. C., Grant, A. L., Forrest, J. C., Schinckel,
A. P., & Gerrard, D. E. (2003). Early postmortem electrical stimula-
tion simulates PSE pork development. Meat Science, 63, 69–77.
Herfort Pedersen, P., Oksbjerg, N., Karlsson, A. H., Busk, H., Bendixen,
E., & Henckel, P. (2001). A within litter comparison of muscle fibre
characteristics and growth of halothane carrier and halothane free
crossbreed pigs. Livestock Production Science, 73, 15–24.
Honikel, K. O., & Kim, C.-J. (1986). Causes of the development of PSE
pork. Fleischwirtsch., 66, 349–353.
Irving, T. C., Swatland, H. J., & Millman, B. M. (1989). X-ray diffraction
measurements of myofilament lattice spacing and optical measure-
ments of reflectance and sarcomere length in commercial pork loins.
Journal of Animal Science, 67, 152–156.
Joo, S. T., Kauffman, R. G., Kim, B. C., & Park, G. B. (1999). The
relationship of sarcoplasmic and myofibrillar protein solubility to
colour and water-holding capacity in porcine longissimus muscle.
Meat Science, 52, 291–297.
Kastenschmidt, L. L., Hoekstra, W. G., & Briskey, E. J. (1968). Glycolytic
intermediates and co-factors in ‘‘fast-’’ and slow-glycolyzing’’ muscles
in pig. Journal of Food Science, 33, 151–158.
Kim, J., Solis, R.S. Arias, E.B. & Cartee, G.D. (2004) Postcontraction
insulin sensitivity: relationship with contraction protocol, glycogen
concentiration and 5’ AMP-activiated protein kinase phosphorylation.
Journal of Applied Physiology. 96, 575–583.
Klont, R. E., & Lambooy, E. (1995). Influence of preslaughter muscle
temperature on muscle metabolism and meat quality in anesthetized
pigs of different halothane genotypes. Journal of Animal Science, 73,
96–107.
Kurth-Kraczek, E. J., Hirshman, M. F., Goodyear, L. J., & Winder, W.
W. (1999). 5’ AMP-activated protein kinase activation causes GLUT4
translocation in skeletal muscle. Diabetes, 48, 1667–1671.
Kyla-Puhju, M., Ruusunen, M., & Puolanne, E. (2005). Activity of
porcine muscle glycogen debranching enzyme in relation to pH and
temperature. Meat Science, 69, 143–149.
Leach, L. M., Ellis, M., Sutton, D. S., McKeith, F. K., & Wilson, E. R.
(1996). The growth performance, carcass characteristics, and meat
quality of halothane carrier and negative pigs. Journal of Animal
Science, 74, 934–943.
16 T.L. Scheffler, D.E. Gerrard / Meat Science 77 (2007) 7–16
Lebret, B., Le Roy, P., Monin, G., Lefaucheur, L., Caritez, J. C., Talmant,
A., et al. (1999). Influence of the three RN genotypes on chemical
composition, enzyme activities, and myofiber characteristics of porcine
skeletal muscle. Journal of Animal Science, 77, 1482–1489.
Le Roy, P., Elsen, J. M., Caritez, J. C., Talmant, A., Juin, H., Sellier, P.,
et al. (2000). Comparison between the three porcine RN genotypes for
growth, carcass composition and meat quality traits. Genetics Selection
Evolution, 32, 165–186.
Le Roy, P., Naveau, J., Elsen, J. M., & Sellier, P. (1990). Evidence for a
new major gene influencing meat quality in pigs. Genetical Research,
55, 33–40.
Lundstrom, K., Essen-Gustavsson, B., Rundgren, M., Edfors-Lilja, I., &
Malmfors, G. (1989). Effect of halothane genotype on muscle
metabolism at slaughter and its relationship with meat quality: A
within litter comparison. Meat Science, 25, 251–263.
MacLennan, D. H., & Phillips, M. S. (1992). Malignant hyperthermia.
Science, 256, 789–794.
Mickelson, J., Gallant, E., Litterer, L., Johnson, K., Rempel, W., & Louis,
C. (1988). Abnormal sarcoplasmic reticulum ryanodine receptor in
malignant hyperthermia. Journal of Biological Chemistry, 263,
9310–9315.
Mickelson, J. R., Gallant, E. M., Rempel, W. E., Johnson, K. M., Litterer,
L. A., Jacobson, B. A., et al. (1989). Effects of the halothane sensitivity
gene on sarcoplasmic reticulum function. American Journal of Phys-
iology – Cell Physiology, 26, C787–C794.
Milan, D., Jeon, J. T., Looft, C., Amarger, V., Robic, A., Thelander, M.,
et al. (2000). A mutation in PRKAG3 associated with excess glycogen
content in pig skeletal muscle. Science, 288, 1248–1251.
Moesgaard, B., Quistorff, B., Christensen, V. G., Therkelsen, I., &
Jorgensen, P. F. (1995). Differences of post-mortem ATP turnover in
skeletal muscle of normal and heterozygote malignant-hyperthermia
pigs: Comparison of 31P-NMR and analytical biochemical measure-
ments. Meat Science, 39, 43–57.
Monin, G., Brard, C., Vernin, P., & Naveau, J. (1992). Effects of the RN-
gene on some traits of muscle and liver in pigs. In: 38th ICoMST,
Clermont-Ferrand, France. pp. 391–394.
Monin, G., & Sellier, P. (1985). Pork of low technological quality with a
normal rate of muscle pH fall in the immediate post-mortem period:
The case of the Hampshire breed. Meat Science, 13, 49–63.
O’Brien, P. J. (1986). Porcine malignant hyperthermia susceptibility:
hypersensitive calcium release mechanism of skeletal muscle sarco-
plasmic reticulum. Canadian Journal of Veterinary Research, 50,
318–328.
Offer, G., Knight, P. K., Jeacocke, R., Almond, R., Cousins, T., Elsey, J.,
et al. (1989). The structural basis of the water-holding, appearance
and toughness of meat and meat products. Food Microstructure, 8,
151–170.
Ono, K., Topel, D. G., Christian, L. L., & Althen, T. G. (1977).
Relationship of cyclic-AMP and phosphorylase a in stress susceptible
and control pigs. Journal of Food Science, 42, 108–110.
Ponticos, M., Lu, Q. L., Morgan, J. E., Hardie, D. G., Partridge, T. A., &
Carling, D. (1998). Dual regulation of the AMP-activated protein
kinase provides a novel mechanism for the control of creatine kinase in
skeletal muscle. EMBO Journal, 17, 1688–1699.
Rempel, W. E., Lu, M. Y., Mickelson, J. R., & Louis, C. F. (1995). The
effect of skeletal muscle ryanodine receptor genotype on pig perfor-
mance and carcass quality traits. Animal Science, 60, 249–257.
Sayre, R. N., Briskey, E. J., & Hoekstra, W. G. (1963). Comparison of
muscle characteristics and post-mortem glycolysis in three breeds of
swine. Journal of Animal Science, 22, 1012–1020.
Schwagele, F., & Honikel, K. O. (1988). Studies in postmortem metab-
olism of PSE-prone pork muscles. In: 34th International Congress of
Meat Science and Technology, Brisbane, Australia.
Schwagele, F., Buesa, P. L. L., & Honikel, K. O. (1996a). Enzymological
investigations on the causes for the PSE-syndrome, II. Comparative
studies on glycogen phosphorylase from pig muscles. Meat Science, 44,
41–53.
Schwagele, F., Haschke, C., Honikel, K. O., & Krauss, G. (1996b).
Enzymological investigations on the causes for the PSE-syndrome, I.
Comparative studies on pyruvate kinase from PSE- and normal pig
muscles. Meat Science, 44, 27–40.
Scopes, R. K. (1974). Studies with a reconstituted muscle glycolytic
system. The rate and extent of glycolysis in simulated postmortem
conditions. Biochemical Journal, 142, 79–86.
Sellier, P., & Monin, G. (1994). Genetics of pig meat quality: A review.
Journal of Muscle Foods, 5, 187–219.
Shen, Q. W., & Du, M. (2005). Role of AMP-activated protein kinase in
the glycolysis of postmortem muscle. Journal of the Science of Food and
Agriculture, 85, 2401–2406.
Shen, Q. W., Means, W. J., Thompson, S. A., Underwood, K. R., Zhu, M.
J., McCormick, R. J., et al. (2006a). Pre-slaughter transport, AMP-
activated protein kinase, glycolysis, and quality of pork loin. Meat
Science, 74, 388–395.
Shen, Q. W., Means, W. J., Underwood, K. R., Thompson, S. A., Zhu, M.
J., McCormick, R. J., et al. (2006b). Early post-mortem AMP-
activated protein kinase (AMPK) activation leads to phosphofructo-
kinase-2 and -1 (PFK-2 and PFK-1) phosphorylation and the
development of pale, soft, and exudative (pse) conditions in porcine
longissimus muscle. Journal of Agricultural and Food Chemistry, 54,
5583–5589.
Stetzer, A. J., & McKeith, F. K. (2003). Benchmarking value in the pork
supply chain: Quantitative strategies and opportunities to improve
quality. Savoy, IL: American Meat Science Association.
Winder, W. W. (2001). Energy-sensing and signaling by AMP-activated
protein kinase in skeletal muscle. Journal of Applied Physiology, 91,
1017–1028.
Yu, H., Fujii, N., Hirshman, M. F., Pomerleau, J. M., & Goodyear, L. J.
(2004). Cloning and characterization of mouse 5’-AMP-activated
protein kinase gamma3 subunit. American Journal of Physiology – Cell
Physiology, 286, C283–C292.
 MEAT
SCIENCE
Meat Science 77 (2007) 17–27
www.elsevier.com/locate/meatsci

Predictive microbiology: Quantitative science delivering

quantifiable benefits to the meat industry and other food industries
T.A. McMeekin
Australian Food Safety Centre of Excellence, Tasmanian Institute of Agricultural Research, University of Tasmania, Private Bag 54,
Hobart, Tasmania 7001, Australia

Received 22 March 2007; received in revised form 2 April 2007; accepted 2 April 2007

Abstract

Predictive microbiology is considered in the context of the conference theme ‘‘chance, innovation and challenge’’, together with the
impact of quantitative approaches on food microbiology, generally. The contents of four prominent texts on predictive microbiology
are analysed and the major contributions of two meat microbiologists, Drs. T.A. Roberts and C.O. Gill, to the early development of
predictive microbiology are highlighted. These provide a segue into R&D trends in predictive microbiology, including the Refrigeration
Index, an example of science-based, outcome-focussed food safety regulation.
Rapid advances in technologies and systems for application of predictive models are indicated and measures to judge the impact of
predictive microbiology are suggested in terms of research outputs and outcomes. The penultimate section considers the future of pre-
dictive microbiology and advances that will become possible when data on population responses are combined with data derived from
physiological and molecular studies in a systems biology approach.
Whilst the emphasis is on science and technology for food safety management, it is suggested that decreases in foodborne illness will
also arise from minimising human error by changing the food safety culture.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Predictive microbiology; Quantitative microbial ecology; Current R&D trends; Applications; Advances in technology; Human error and the
food safety culture

1. Introduction accorded this status as it provides essential support to

1.1. Predictive microbiology in the context of ICOMST 53
The theme of the 53rd International Congress of Meat
Science and Technology is ‘‘chance, innovation, challenge’’,
each element of which has had a part to play in the devel-
opment of predictive microbiology, from the concept stage
to that of a rapidly maturing subdiscipline of food micro-
biology (McMeekin, Olley, Ratkowsky, & Ross, 2002).
This led to a description of predictive microbiology, per-
haps prematurely at that time, by Labuza (1994) as the
new paradigm of food safety management. It is now
E-mail address: Tom.McMeekin@utas.edu.au
underpin widely accepted food safety management systems
such as HACCP, quantitative microbial risk assessment
and evolving systems, such as food safety objectives.
The essence of predictive microbiology is that microbial
population behaviour in foods can be predicted by mea-
surement of the effects of environmental factors, assem-
bling the data into a database and synthesising the data
into a mathematical model to be incorporated into predic-
tive software. When combined with appropriate monitor-
ing technology, shelf life or microbial safety can be
estimated without recourse to traditional microbiological
enumeration techniques. In effect, predictive microbiology
has enabled traditional microbiology to break out of the
condition described by Sharpe (1980) as ‘‘the time warp
of agar-based methods’’. Nevertheless, we must not forget

0309-1740/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.005
18 T.A. McMeekin / Meat Science 77 (2007) 17–27
the vital role of traditional methods in reaching this point,
and that they will continue to be used to develop and val-
idate predictive models.
The concept of predictive microbiology may, therefore,
be described as an important innovation which appears to
have been first clearly enunciated by Scott (1937) as follows:
‘‘A knowledge of the rates of growth of certain microor-
ganisms at different temperatures is essential to studies
of the spoilage of chilled beef. Having these data, it
should be possible to predict the relative influence on
spoilage exerted by the various organisms at each stor-
age temperature. Further, it would be feasible to predict
the possible extent of the changes in populations which
various organisms may undergo during the initial cool-
ing of sides of beef in the meatworks when the meat sur-
faces are frequently at temperatures very favourable to
microbial proliferation.’’
The Scott (1937) paper is one of a trilogy on the growth
of microorganisms on ox muscle published in the period
1936–1938. Scott (1936) discussed the influence of water
content of substrate and Scott (1938) examined the effect
of carbon dioxide. Combined with investigations on chilled
beef (Empey & Scott, 1939; Scott & Vickery, 1939), these
studies provided a sound scientific basis to enable the ship-
ment of chilled meat quarters from Australia to the United
Kingdom.
If the above analysis of Scott’s work is correct, viz.
that the concept developed as a logical consequence of
many years research on meat microbiology by the now
Commonwealth Scientific and Industrial Research Orga-
nisation (CSIRO)/Food Science Australia (FSA) and
not by chance, it represents the beginning of an innova-
tive concept, despite the lack of essential tools, such as
computing power, through which the concept could be
turned into reality. The steps required for this transfor-
mation to occur represented a challenge for the various
disciplines and skill sets required, including microbiology,
mathematics and statistics, and information systems and
technology.
Through efforts in each of these disciplines, predictive
microbiology has had a large influence on the philosophy
underpinning food microbiology with a gradual change,
still underway, where a qualitative/semi-quantitative sci-
ence is being transformed into a quantitative science. This
is a development that would have been applauded by Lord
Kelvin (William Thomson, 1824–1907) who wrote:
‘‘When you can measure what you are speaking about,
and express it in numbers, you know something about
it, but when you cannot, your knowledge is of a meagre
and unsatisfactory kind . . . it may be the beginning of
knowledge, but you have scarcely in your thoughts
advanced to the state of Science.’’
The ultimate beneficiaries of this change in philosophy
will be consumers when the technology is embraced by
both the food industry and regulatory authorities and this
is a particular challenge for research providers. That
challenge, however, is consistent with Codex Alimentarius
Commission principles that the hygienic equivalence of
foods in international trade must be based on sound scien-
tific findings. Perhaps the most persuasive arguments are
that a decision about the safety or shelf life of food can
be made in real time, rather than retrospectively, and that
the benefits of outcome-based regulations, rather than pre-
scriptive regulations, provide industry with flexibility in
assuring safe outcomes.
1.2. Consequences of a change from qualitative to
quantitative approaches to predictive microbiology
Common definitions of chance include: a possibility; a
risk; an undesigned occurrence; fortune/luck; and fate.
Each of the definitions suggests an element of uncertainty
and are clearly not descriptive terms appropriate to deci-
sions of food safety or to development of food safety man-
agement plans. By adopting a quantitative approach,
uncertainty is reduced, possibilities become probabilities,
risks and undesigned occurrences are lessened. Outcomes
decided by luck (not skill), analogous to a game of chance,
and those arising as a result of unalterable events (fate),
have no place in any food safety management system,
whether qualitative or quantitative.
The move to probability brings with it the concept of a
probability distribution function, mathematically described
as ‘‘a function of a continuous random variable whose inte-
gral over any interval gives the probability that the value
specified by it will lie within that interval’’ or, in lay terms,
a probability distribution represents the range of possible
values, or outcomes, and the probability of each of them
occurring.
A consequence of being able to assign a distribution
to define time-based microbial responses, such as genera-
tion times, lag times or death kinetics, is that the vari-
ability of those responses can be incorporated in
calculations. In turn, this provides opportunity to place
confidence limits on predictions with better-defined cer-
tainty about a food safety outcome. For detailed discus-
sion, see Ratkowsky, Ross, McMeekin, and Olley (1991)
and McMeekin, Olley, Ratkowsky, and Ross (1993,
chap. 4), in which the paramount importance of the sto-
chastic (error) assumption to compare the efficacy of
models is also emphasised.
The Gamma distribution (with an extended right hand
tail), in which the variability is proportional to the square
of the population mean, is commonly observed to
describe population kinetic data. A consequence of this
is that a point is reached where variability becomes so
large that kinetic models are inappropriate and instead
it is more useful to consider the probability that growth
is likely to occur at all. This type of approach is now
widely used to define conditions representing the bound-
ary between growth and no growth, thus providing a
means to quantify the hurdle concept (McMeekin et al.,
 T.A. McMeekin / Meat Science 77 (2007) 17–27
2000; Presser, Ross, & Ratkowsky, 1998; Ratkowsky &
Ross, 1995; Tienungoon, Ratkowsky, McMeekin, &
Ross, 2000).
Probability distributions are also the basis of Monte
Carlo simulation approaches which are finding increasing
utility in quantitative microbial risk assessments (Cassin,
Lammerding, Todd, Ross, & McColl, 1998; Koutsou-
manis, Taoukis, & Nychas, 2005; Rasmussen, Ross, Olley,
& McMeekin, 2002).
2. Classifying the content of significant texts on predictive
microbiology
Three text books and a 196-page report have been pub-
lished on predictive modelling of microbial behaviour in
foods, viz: McMeekin et al. (1993); Ross (1999) (who spe-
cifically reported on predictive models for the meat indus-
try); McKellar and Lu (2004), Brul, Zwietering, and Van
Gerwen (2007).
Inspection of the chapter titles in these texts identifies
those areas of predictive microbiology that may be categor-
ised as
 basic to the underpinning science, but require incremen-
tal updating:
 McMeekin et al. (1993): ‘‘Basic concepts and meth-
ods’’ (Chapter 2); ‘‘Modelling temperature effects’’
(Chapter 3); ‘‘Comparing Belehradek-type and
Arrhenius-type models using real data’’ (Chapter 4);
‘‘Modelling the combined effect of temperature, water
activity and other factors on microbial growth rate’’
(Chapter 5).
 Ross (1999): ‘‘Literature review’’ (Chapter 1)
describes the origins of predictive microbiology, the
microbial ecology of foods, limitations to the applica-
tion of predictive microbiology, model availability
and assessing model performance.
 McKellar and Lu (2004): ‘‘Experimental design and
data collection’’ (Chapter 1); ‘‘Primary models’’
(Chapter 2); ‘‘Secondary models’’ (Chapter 3);
‘‘Model fitting and uncertainty’’ (Chapter 4).
 Brul et al. (2007): ‘‘Experimental design, data pro-
cessing and model fitting in predictive microbiology’’
(Chapter 3).
These chapters enunciate and reinforce the rules of predic-
tive modelling, supporting the conclusion that the ‘‘front
end’’ of the modelling process has a sound scientific basis
(McMeekin, 2004).
 those aspects that continue to cause concern and require
innovative approaches to aid the advancement of
knowledge:
 Lag phase and fluctuating conditions (Ross, 1999;
Smelt & Brul, 2007).
 Food heterogeneity and microbial interactions (Broc-
klehurst, 2004; Leroy & De Vuyst, 2007).
19
 Uncertainty and variability (Nauta, 2007).
 validation of applications for predictive models in food
safety management:
 Expanding horizons for the application of predictive
models in food safety management, particularly risk
assessment, were dealt with by Lammerding and
McKellar (2004) and Zwietering and Nauta (2007).
 emergence of new technologies and systems, mostly elec-
tronic, bringing greater precision and more rapid
reporting:
 Rapid developments in technology and systems have
been reported by Tamplin, Baranyi, and Paoli (2004),
McMeekin, Szabo, and Ross (2005), McMeekin et al.
(2006a) and Legan (2007). These are mainly based on
electronic logging devices (of increasing sophistica-
tion) and information systems. However, a biologi-
cally based indicator strip is now available from
Cryolog (see www.cryolog.com).
 and futuristic concepts which will expand the bound-
aries of knowledge through the incorporation of physio-
logical and molecular information into predictive
models (Brul & Westerhoff, 2007).
3. Significant roles of two meat microbiology research groups
in the early development of predictive microbiology
In the introductory remarks to this paper, the concept of
predictive food microbiology was attributed to W.J. Scott,
an Australian meat microbiologist also renowned for initi-
ating the water activity concept in food microbiology. And,
the tradition has been continued as several research groups
with a research focus on meat microbiology have made sig-
nificant contributions to the early development of predic-
tive microbiology.
3.1. Dr. T.A. Roberts and colleagues at the Meat Research
Institute in Bristol and the Institute of Food Research,
Reading, United Kingdom
The term ‘predictive modelling’ appears to have first
been used by Roberts and Jarvis (1983) and early exper-
iments on the probability of growth and production of
toxins by Clostridium botulinum were reported by Rob-
erts, Gibson, and Robinson (1981), with Roberts subse-
quently making major contributions to predictive
modelling through his central role in the United King-
dom program funded by the Ministry of Agriculture,
Fisheries and Forestry (MAFF) and resulting in Food
MicroModel. This activity was paralleled in the United
States by Dr. R.L. Buchanan’s group at the United
States Department of Agriculture, Philadelphia, which
gave rise to the Pathogen Modeling Program. Together,
these are now the basis for ComBase, an international
predictive models database, which will be discussed later
in this contribution.
A significant appointment to the group by Roberts in
1990 was Dr. Jozsef Baranyi, a Hungarian mathematician
20 T.A. McMeekin / Meat Science 77 (2007) 17–27
whose prowess as a quantitative thinker expanded the
group’s output into more fundamental studies and had a
significant and rapid effect on the field. Dr. Baranyi is
now the driving force in the UK for ComBase (http://
www.combase.cc).
3.2. Dr. C.O. Gill and colleagues at the Meat Industry
Research Institute, New Zealand (MIRINZ), and
Agri-Food Canada, Alberta, Canada
 Dr. Colin Gill, made an immediate impact on the field
of meat microbiology when appointed to MIRINZ.
Most of his early work (1979–1989) was concerned
with the spoilage biota and spoilage mechanisms of
meat and meat products. His first modelling paper
was published in 1984 and, at the 39th International
Congress of Meat Science and Technology meeting
held in Calgary in 1993, the following modelling appli-
cations were attributed to Gill and colleagues (McMe-
ekin & Ross, 1993):
 offal cooling (Gill, 1984; Gill & Jones, 1992b);
 computer programs for process hygiene evaluation
(Gill, Phillips, Loeffen, & Bishop, 1988);
 meat thawing procedures (Lowry, Gill, & Pham,
1989);
 spray cooling of carcasses (Gill, Jones, & Tong,
1991);
 hot boning processes (Reichel, Phillips, Jones, & Gill,
1991);
 cooling of pig carcasses (Gill & Jones, 1992a).
The studies above used temperature loggers to record
temperature fluctuations which, when integrated, gave
an estimate of growth of the target organism at a refer-
ence temperature [see McMeekin et al. (1993, chap. 6),
for detail and discussion of the relative rate concept].
The overriding principle for temperature function integra-
tion (TFI) as an index of process hygiene was stated
clearly by Gill and Phillips (1993a): ‘‘TFI must evaluate
the hygienic equivalence of a process: it cannot be used
to assess the absolute hygienic status of individual units
leaving the process’’.
Further applications by Gill and his colleagues in
Alberta, Canada, were developed to estimate the shelf life
of meat during transcontinental transportation (Gill &
Phillips, 1993b) and for beef and pork during retail dis-
play (Gill, Greer, & Dilts, 1998). Additional, and confir-
matory, process hygiene evaluations were carried out for
offal cooling (Gill, Taylor, & Tong, 1995), processed
meats (Gill, Friske, & Tong, 1995), beef cooling processes
(Gill & Bryant, 1997), air cooling of lamb carcasses (Gill
& Jones, 1997) and sheep carcass dressing (Gill & Baker,
1998).
Thus, Gill and his colleagues had firmly established the
utility of TFI in the meat industry and the concept of the
process hygiene index (PHI) as a surrogate measure to
assess the adequacy of processes.
4. Significant research and development trends in predictive
microbiology
This section of the paper was originally conceived as a
series of subsections with a few paragraphs on each, but,
following several iterations and identification of multiple
linkages between the subsections, it became apparent that
a more coherent section should be written to draw out
the interactions between:
 database development (particularly ComBase) and
access to data;
 systematic analysis of the literature;
 quantitative microbial risk assessment;
 variability and uncertainty;
 stochastic simulation models and
 estimation of lag phase duration.
The processes involved in developing and evaluating
predictive models are now well-established, comprising
building a database from which a mathematical model is
constructed (McMeekin, 2004). In turn, the model is incor-
porated into predictive software by which it is applied to
estimate the microbiological consequences of a process
through monitoring the environment.
ComBase is an international database of microbial
growth or death rates that originated through pooling the
data collected for the US Pathogen Modeling Program
and the UK Growth Predictor (formerly Food Micro-
Model) to which date from the literature was added to pro-
vide 39,000 records in late 2005. These were supplemented
by a further 8000 records provided in early 2006 by the
Australian Food Safety Centre of Excellence (AFSCoE),
comprising 5500 kinetic datasets and 2500 boundary
model datasets. During 2006, AFSCoE also became the
third full partner of the ComBase consortium with the
US and the UK. More detailed information on ComBase
can be found in Baranyi and Tamplin (2004), Tamplin
et al. (2004) and Legan (2007) and at www.combase.cc.
A significant recent development, likely to lead to rapid
expansion of ComBase, is the initiative of the Journal of
Food Protection to provide a mechanism for datasets in
papers published in that journal to be added to ComBase.
This opportunity is also likely to be offered soon by the
International Journal of Food Microbiology and, possibly,
by other leading food microbiology journals.
Access to models was discussed by Legan (2007) who
observed that, while many models have been published
in refereed journals, only a small proportion are available
in electronic form which has the great advantage of con-
venience. Legan (2007) also provided a very useful list of
electronic contact details for models and modelling tools.
Related lists with similar contact information can be
found in McMeekin et al. (2005) and McMeekin et al.
(2006a).
McMeekin, Mellefont, and Ross (2007) identified a
trend in the food safety literature indicating an increasing
 T.A. McMeekin / Meat Science 77 (2007) 17–27
number of studies based on systematic and critical analysis
of the available literature, rather than a simple chronolog-
ical account of developments in a particular area. Specific
examples cited included an analysis of 4066 thermal inac-
tivation curves for foodborne pathogens (van Asselt &
Zwietering, 2006) and Campylobacter control strategies
for the UK broiler industry (Adkin, Hartnett, Jordan,
Newell, & Davison, 2006). More generally, Vialette,
Pinon, Leporq, Dervin, and Membré (2005) reported on
a meta-analysis of food safety information based on a
combination of a relational database and a predictive
modelling tool.
A meat industry example was provided by Ross and
Shadbolt (2001), who systematically reviewed the inactiva-
tion of E. coli in fermented meat manufacture. The outcome
was modification of the relevant legislation by the Australia
and New Zealand Food Authority (now Food Standards
Australia New Zealand), as follows: ‘‘the process of fermen-
tation and any other subsequent processes must reduce
prior to sale from the processing factory by 99.9% or greater
than the number of Escherichia coli organisms potentially
present in an uncooked, comminuted meat product’’.
The ‘take home’ message was that analysis of individual
studies yielded specific information not readily transferable
to other studies. However, by combining all the available
data, patterns of microbial inactivation emerged which
could be modelled and which identified time–temperature
combinations during the process as, by far, the major influ-
ence on the rate of decline. Interestingly, funding of further
experimental work by Meat and Livestock Australia
(MLA) failed to improve the model developed from pub-
lished data.
The trail now leads from systematic analysis of the liter-
ature to Quantitative Microbial Risk Assessment which
was endorsed by the Codex Alimentarius Commission
(Hathaway, 1993; Hathaway & Cook, 1997), spawning a
great deal of activity in the area, much of which has been
reported in the FAO/WHO Microbiological Risk Assess-
ment Series (http://www.who.int/foodsafety/publications/
micro/en/index.html).
Again, meat microbiologists were early contributors to a
rapidly developing field with Cassin et al. (1998) reporting
a quantitative risk assessment for Escherichia coli O157:H7
in ground beef hamburgers.
And, thence, the debate returns to variability and uncer-
tainty, the intricacies of which were very clearly laid out by
Nauta (2007). Variability and uncertainty lead to imprecise
point estimates by predictive modelling software. This dif-
ficulty can be overcome by using stochastic simulation
(Monte Carlo) models such as @Risk (Palisade Corpora-
tion, Ithaca, NY, USA), AnalyticaÒ (Lumina Decision Sys-
tems Inc., Los Gatos, CA, USA) and Crystal BallÒ
(Decisioneering Inc., Denver, CO, USA), as demonstrated
by Fazil et al. (2002), Rasmussen et al. (2002), Ross and
McMeekin (2003) and Membré et al. (2006).
Perhaps the most integrated approach is the Safety
Monitoring and Assurance System (SMAS) developed for
21
chilled food products by Koutsoumanis et al. (2005). The
basis of SMAS is a combination of mathematical models
describing commonly occurring meat microorganisms and
the response of an appropriate time–temperature integra-
tor (TTI) which integrates the temperature history of the
product as described before. The TTI response is related
to the microbiological quality status of the food by collect-
ing information at predetermined points of the chill chain.
To evaluate the effectiveness of the SMAS system, a large
number of chill chain scenarios were run using a Monte
Carlo simulation, the predictions of which were compared
with results using the traditional first in, first out (FIFO)
system. The Monte Carlo simulation model includes prob-
ability distributions for the variable storage and cooking
conditions to which meat will be subjected prior to con-
sumption and the outputs are a probability distribution
of the number of pathogenic bacteria at the time of con-
sumption and the probability of illness.
The simulation results for the risk of listeriosis demon-
strated a shift of the SMAS distribution compared with
the FIFO distribution, indicating a lower probability of ill-
ness and a significant decrease in high risk units when the
product was shipped in an export chill chain. When the dis-
tribution of the remaining shelf life at 0 °C was examined,
unacceptable products were reduced from 12.5% to 4.3%
(Koutsoumanis et al., 2005). It is interesting to note that
one temperature profile can be interpreted both for public
health significance and shelf life estimation and that,
although developed for meat products, the authors
expressed confidence that it can be applied to any product
in the chill chain.
Incorporating lag times into predictive model calcula-
tions has been a long-standing irritation for predictive
modellers as it is well known that estimates of lag times
are much more variable than generation times. This issue
was addressed in Chapter 5 (Ross, 1999) who reviewed
the unpredictability of bacterial lag times and the limita-
tions this imposes on the application of predictive models.
Using a combination of systematically analysed literature
data and de novo experimental data, he introduced the
innovative concept of relative lag times or ‘‘generation time
equivalents’’, i.e. the ratio of lag time to generation time.
This approach revealed that there is a common pattern
of distribution for a wide range of species across a wide
range of conditions which has a sharp peak in the range
4–6 generation time equivalents. Thus, based on a very sub-
stantial body of information, it is possible to justify inclu-
sion of lag times in calculations of the effect of different
meat processing and handling procedures.
4.1. The Refrigeration Index (RI) – an example of science-
based, outcome-focused food safety regulation
Earlier, the work of Gill and colleagues at MIRINZ,
leading to the Process Hygiene Index (PHI), was consid-
ered. Despite being based on limited data for temperature
only, the model appeared to give a reasonable estimate of
22 T.A. McMeekin / Meat Science 77 (2007) 17–27
the potential for E. coli growth during chilling. In Austra-
lia, Smith (1987) produced a similar model for E. coli
growth and, many years later, the model used for the RI
was based on >1000 observations of the effects of temper-
ature, water activity, pH and lactate concentration on
E. coli growth (Ross, Ratkowsky, Mellefont, & McMeekin,
2003). The model was rigorously evaluated by Mellefont,
McMeekin, and Ross (2003) and compared very well with
other published models for E. coli growth, particularly on
meat carcasses. This goodness of fit was attributed to inclu-
sion of a term for the effect of lactate concentration which
could be varied according to that measured or anticipated
in different meat cuts.
The science behind the development and evaluation of
the RI was considered sufficiently extensive and accurate
to cause the Australian Quarantine Inspection Service
(AQIS) to revise the Export Control (Meat and Meat Prod-
ucts) Orders in 2005 (AQIS, 2005). The outcome is that the
chilling of meat carcasses in Australian export abattoirs is
based on an E. coli growth model which is used to interpret
cooling profiles measured by temperature loggers. These
are placed in a strictly prescribed manner to estimate the
potential for E. coli growth expressed as the RI (http://
www.mla.com.au).
5. Significant recent advances in technology for the
application of predictive models
For many years, the choice of technology to monitor the
temperature of foods in distribution lay between chemical
or physical monitors or electronic temperature loggers,
with the latter becoming increasingly selected. However,
there are potential limitations with data recovery if the log-
gers are used to monitor shipments of product, rather than
in a process at a fixed location, e.g. non-return of loggers,
sometimes deliberate, with loss of all the information col-
lected; retrospective analysis of the microbiological conse-
quences of a temperature profile; and manual
examination for ‘progress reports’ as the shipment moves
along the supply chain (McMeekin et al., 2005).
In recent years, radio frequency identification (RFID-
based) technologies have been applied in an attempt to
overcome these problems and a further promising develop-
ment is wireless network technology, such as the Smart-
Traceä system developed by Ceebron Pty Ltd, Sydney,
Australia (www.ceebron.com). This is based on innovative
platform ad hoc wireless technology in which disposable
sensor tags, placed on individual pallet loads of meat, relay
messages from one to the next and on to an external reader.
McMeekin, Smale, Jenson, Ross, and Tanner (2006b)
reported combining SmartTraceä with the RI predictive
model to assess the significance, in microbiological terms,
of a refrigerated shipment of lamb from Colac, Victoria,
to Brisbane, Queensland, Australia, a road journey of
2000 km which took 65 h.
During the journey, 280,000 temperature recordings
were taken (by Food Science Australia’s Supply Chain
Innovation Group) using 18 SmartTraceä tags on pallets.
Interpretation of the data by the model showed that the
minimum RI was 0.0 (no growth of E. coli) and the maxi-
mum RI was 0.28 (i.e. less than one generation of E. coli).
The regulatory requirements for RI through the entire pro-
cess (not transport only) are on average <2.0 and a maxi-
mum of 2.5 leading to the conclusion that the transport
phase made a negligible contribution to the RI for the
entire process.
Domestic regulatory requirements in Australia stipulate
that carcasses are chilled to, and maintained at, less than
7 °C and that cut products are maintained at <5 °C (Aus-
tralian Standard 4694, 2002). Strict interpretation of the
regulations would identify some situations during this ship-
ment which exceeded the prescriptive requirements of the
current regulations. However, when the microbiological
(and public health) implications are assessed on the basis
of the RI (AQIS, 2005), the negligible effect of the trans-
port phase demonstrates a strong argument for outcome-
based regulations. A similar argument was sustained by
AQIS in mandating the RI to monitor carcass chilling in
the revised export meat orders discussed earlier. Prior to
that, the regulation stated that the surface temperature of
a carcass must be reduced to 7 °C or less in a 24 h period.
However, one can envisage cooling profiles that would lead
to minimal potential proliferation of E. coli (rapid initial
cooling, followed by a slower rate of cooling to 7 °C) to
those resulting in 4–5 generations of E. coli (slow initial
rate of cooling, particularly at temperatures above 30 °C,
followed by rapid cooling towards the end of the process).
A further advantage is that, while the RI was developed
for conventional air chilling of carcasses in Australian
export abattoirs, the same criteria can be applied to alter-
native processes such as hot boning. Thus, different sectors
of the export meat industry have a ‘‘level playing field’’
and, importantly, flexibility in meeting an outcome-based
food safety regulation.
Earlier, it was indicated that electronic devices for
recording temperature history had largely taken over from
chemical or physical indicators. In part, this may be due to
the indicator strips being based on non biological reaction
kinetics. McMeekin and Ross (1996) speculated on the
potential of ‘‘bioindicators’’, citing the example of the
product developed at the Swedish Institute for Research
to monitor sterilisation processes (Rönner, 1990). They
hypothesised that: ‘‘To select a ‘bioindicator’ for shelf life
studies, the organism would be required to display the same
temperature characteristics as the specified spoilage organ-
ism.. . . The organism selected could be the specified spoilage
organism itself or a ‘food grade’ organism such as a lactic
acid bacterium if there was concern about contamination of
the product.’’
The concept proposed in 1996 is now a reality through
the work of a French company, Cryolog (http://www.cry-
log.com) and its research partner ADRIA (http://www.
adria.tm.fr) located in Quimper, Brittany, France, where
the European predictive models database Sym’Previus is
 T.A. McMeekin / Meat Science 77 (2007) 17–27
also located (http://www.symprevius.net). One Cryolog
product, ‘‘TraceoÒ’’, is essentially a barcode which is
transparent with the barcode visible when the product is
fresh, but not visible when the product is no longer edible.
Another product, ‘‘(eO)Ò’’, is designed to be attached to
retail packages, e.g. of sandwiches. In this case, the dis-
play incorporates some Gallic flair with a green flower
when the product is fresh, which turns red at the end of
shelf life.
6. Measures of scientific and technological significance
Scientists tend to be focussed on research outputs, exem-
plified, in the main, by publication in peer-reviewed jour-
nals, distributed to an international audience and with a
high impact factor. This is the stuff of which research rep-
utations, continued research funding streams and research
careers are made.
Research funding agencies in the food safety field will
also value publication, as a tangible indicator of research
carried out on their behalf, which provides evidence for a
sound scientific basis for new regulation and a measure
of the hygienic equivalence of food in international trade,
as required by the Codex Alimentarius Commission. As
observed earlier, predictive microbiology now underpins
the major systems of food safety management: HACCP,
Quantitative Microbial Risk Assessment (QMRA) and
Food Safety Objectives (FSO). For these reasons, food
safety research is often viewed as an activity that should
be funded primarily by government, supported by industry.
The development and maintenance of a nation’s food
industry is closely linked to a high level of food safety
assurance and is inextricably linked to public good out-
comes in any jurisdiction, i.e. both industry and the popu-
lation of a nation will gain benefit from food safety
management systems based on sound science.
In Australia, the meat industry, through MLA and rel-
evant government departments, such as AQIS, have collab-
orated in predictive microbiology research and
development with quantifiable benefits to both the industry
and the nation through the RI and a model for inactivation
of E. coli in fermented meat manufacture (Ross & Shad-
bolt, 2001). These fall into the category of research out-
comes by which practical application of the science
contained in a research output is achieved. Conservatively,
the benefits calculated on the basis of a AUD$3.8 million
(2006 value) investment in R&D and implementation by
MLA, represent an industry benefit–cost ratio in excess
of 10:1. The increase in Australia’s GDP is estimated to
be over AUS$150 million and social benefits of over
AUS$250 million over the next 30 years (I. Jenson, pers.
comm.).
Around the world, most predictive modelling research
has been on foodborne pathogens rather than modelling
spoilage organisms, the latter usually undertaken by an
industry sector or individual food company to gain a
commercial advantage. For this reason, government agen-
23
cies are less likely to invest in spoilage/shelf life studies.
Once again, a research link between studies into shelf life
and foodborne illness needs to be taken into account as
microbial interactions may have a significant outcome
on the microbial ‘‘winners’’ competing for the same
resources in a food ecosystem. Technological changes to
a process to extend shelf life also need to be approached
with caution to ensure that a pathogen cannot reach an
infective dose or dangerous level of toxin production by
extending the use-by-date for the product. A particularly
dangerous situation is possible where an antimicrobial
agent causes greater inhibition of the spoilage biota than
of a pathogen.
Thus, the twin goals of guaranteed safety and adequate
shelf life depend on knowledge of the microbial ecology of
the food and the chance of a successful outcome on both
fronts is greatly increased if quantitative knowledge is
available.
7. Futurology
Futurology is defined as ‘‘the prediction of future as a
result of systematic analysis, especially by study of present
day trends in human affairs’’. Predictive microbiology
appears to represent a subset of human activity by which
future events are predicted on the basis of systematic anal-
ysis of microbial behaviour in foods.
In his excellent chapter, Legan (2007) considered future
trends in predictive microbiology, quoting McMeekin
(2004) who wrote ‘‘one continues to have the sense that
the predictive models and databases on which they are
based have not really reached their potential’’ and pre-
dicted the future as ‘‘moving beyond building models to
developing and verifying applications’’. Legan (2007) com-
mented, ‘‘this is where models can add considerable value
to product and process evaluation and essential decision
making in industry and government.. . . However, this pro-
cess will, itself, take a long time (one hopes less than
another 20 years).’’
I hope that Dr. Legan is correct, as I may not be extant
in the year 2027. And, I think he will be proved correct as
several of the barriers he identified as preventing uptake of
predictive microbiology are already ‘‘crumbling’’.
 ‘‘Traditional scientific publishing provides a considerable
barrier to the wider application of many perfectly useful
models’’.
 ‘‘. . .often only derivatives (summary statistics) of the raw
data are provided in the interests of saving space’’.
These barriers will be lessened soon due to the innova-
tion first introduced by the Journal of Food Protection
and soon to be followed by the International Journal of
Food Microbiology which publishes 25% of predictive
modelling papers (http://www.isiknowledge.com). The
instigator of this important innovation was Professor Mark
Tamplin who requested that raw data be made available
24 T.A. McMeekin / Meat Science 77 (2007) 17–27
voluntarily before he was aware of the comment: ‘‘It is con-
ceivable that submission of raw data to such a database
[ComBase] will eventually become a requirement for publica-
tion of papers addressing microbiological growth, death or
survival’’ (Legan, 2007).
 ‘‘Another barrier to greater uptake is that models are so
conservative as to be impractical’’.
Again, progress has been made, e.g. with the relative lag
time concept (Ross, 1999) allowing the most variable of
response times (lag phase duration) to be factored into
model calculations. It is also anticipated that models will
become more precise when physiological and molecular
data can be included (Brul & Westerhoff, 2007).
Legan (2007) concluded that: ‘‘It is also clear that food
microbiology is a much more quantitative discipline now
[2007] than it was in the 1970s and 1980s’’. This epitomises
the aspirations of the many scientists who have spent 30
years or more making food microbiology a quantitative sci-
ence, such as my colleagues Professors June Olley and
David Ratkowsky, who published their first paper in the
field in 1973 (Olley & Ratkowsky, 1973) and their latest
in 2005 (Ratkowsky, Olley, & Ross, 2005).
However, we now have many more tools at our disposal
which describe events at the cellular level and the subcellu-
lar level listed by Brul and Westerhoff (2007) as:
 microbial physics;
 genomics;
 microbial physiology;
 models in microbiology,
which, when combined, constitute Systems Biology.
In relation to food safety, Brul and Westerhoff (2007)
observed that ‘‘gene expression profiles have helped us elu-
cidate the genes operative in resistance against antimicro-
bials.. . . Currently, an integration of such data with
state-of-the-art physiological data allows us to apply true
systems biology approaches where metabolic and molecular
data are jointly analysed and translated into cellular mod-
els of homeostasis.’’ Similar opportunities exist in micro-
bial food spoilage research, particularly in relation to
microbial survival in food which may be considered as
the outcome of persistence, adaptation and resistance
to food preservation methods and other rigours encoun-
tered in the food environment.
Near the beginning of this treatise, the views of Lord
Kelvin (otherwise titled Baron Kelvin of Largs, H. Swat-
land, pers. comm.) on quantitative science were quoted
and Brul and Westerhoff (2007) and H. Swatland (pers.
comm.) quoted Lord Rutherford as saying ‘‘All science is
either physics or stamp collecting’’. Modern biology, whilst
still dependent on the laws of physics and chemistry, has
moved a long way from the stamp collecting arena and is
now the branch of science experiencing quantum leaps in
understanding life on earth (and beyond). The physicists
in the eras of Kelvin and Rutherford were in their heyday
and, perhaps, because they did not have to deal with the
variability and uncertainty inherent in biological responses,
physics reached the state of quantitative science much ear-
lier than biology.
One is tempted to predict correctly, with a high degree
of probability, that, in the <20 year time frame suggested
by Legan (2007), predictive food microbiology, armed with
the knowledge derived from a systems biology approach as
described by Brul and Westerhoff (2007), will have become
a truly quantitative science.
8. The challenge of changing the food safety culture
In this contribution, as is the norm, the emphasis has
been on science and technology, specifically, the science
underpinning the development and evaluation of predictive
models and electronic and other technologies to allow
practical application of the vast amount of knowledge
stored in databases and condensed into models. It is also
well established that predictive models have a particular
role in supporting food safety management systems such
as HACCP, Quantitative Risk Assessment and Food
Safety Objectives.
Thus, science and technology, appropriately combined,
are essential pillars in the quest to improve food safety out-
comes. The third factor is sociological rather than scien-
tific. Specifically, the apparent propensity of humans to
ignore or misunderstand, even the basic rules for produc-
tion, storage, distribution and preparation of foods. This
frailty is manifest both in employees involved in the food
production chain, from procurement or harvest of the
raw product, to food service and to consumers. In Austra-
lia, the last group continue to be responsible for 20–25% of
foodborne illness, despite the best efforts of agencies such
as the Food Safety Information Council (www.foodsafety.
asn.au).
The importance of the human factor and how changes in
the prevailing culture have led to significant decreases in
adverse events have been demonstrated in the aviation
industry (Maurimo, Reason, Johnson, & Lee, 1995) and
in hospital practice (S. Nightingale pers. comm.,
sdn@burntsidepartners.com) who reported a reduction in
in-house mortality at a major hospital from an unaccept-
able 4.2% to 2.8% (similar to the low end of the range
for comparable hospitals) within a month of introducing
a revised care system.
The first public commentary on the challenge of chang-
ing the food safety culture seems to be that of Yiannas
(2007), who advocated an innovative sociological approach
to change the food safety culture which, he believes, will
significantly decrease the chance of human error contribut-
ing to the incidence of foodborne illness. Given the enor-
mous investment in food industry R&D, in plant and
equipment and in personnel at all levels, plus the cost of
failure as a result of human error, changing the mindset
about food safety of those involved in the food supply
 T.A. McMeekin / Meat Science 77 (2007) 17–27
chain (including consumers) would appear to be a worth-
while investment. This is an area of food safety research
likely to become much more prominent with a monograph
by Frank Yiannas, the current President of the Interna-
tional Association of Food Protection (frank.yiannas@
disney.com), due to be published in 2007.
Acknowledgements
The author is indebted to Meat and Livestock Australia
(http://www.mla.com.au) for continued funding of re-
search in predictive microbiology and, in particular, to
Ian Jenson of that organisation for securing tangible out-
comes for the Australian meat industry on the basis of
research outputs in predictive microbiology. Grateful
thanks are also due to Associate Professor Tom Ross
and Dr. Susan Dobson for critical evaluation of the man-
uscript, and to Sally Jones, without whose assistance in
preparation of the text, the submission deadline could
not have been reached or even approached.
References
Adkin, A., Hartnett, E., Jordan, L., Newell, D., & Davison, H. (2006).
Use of a systematic review to assist the development of Campylobacter
control strategies in broilers. Journal of Applied Microbiology, 40,
306–315.
AQIS (Australian Quarantine Inspection Service). (2005). Export Control
(Meat and Meat Products) Orders 2005 (http://www.comlaw.gov.au/
ComLaw/Legislation/LegislativeInstrumentCompilation1.nsf/).
Australian Standard 4694. (2002). The hygienic production and transpor-
tation of meat and meat products for human consumption, Standards
Australia, 84pp.
Baranyi, J., & Tamplin, M. L. (2004). ComBase: A combined database on
microbial responses to food environments. Journal of Food Protection,
67, 1967–1971.
Brocklehurst, T. (2004). Challenge of the food environment. In R. C.
McKellar & X. Lu (Eds.), Modeling microbial responses in foods
(pp. 233–242). Boca Raton, USA: CRC Press.
Brul, S., & Westerhoff, H. V. (2007). Systems biology and food science. In
S. Brul, M. Zwietering, & S. Van Gerwen (Eds.), Modelling microor-
ganisms in food (pp. 250–286). Cambridge, England: Woodhead
Publishing.
Brul, S., Zwietering, M., & Van Gerwen, S. (Eds.). (2007). Modelling
microorganisms in food (pp. 289). Cambridge, England: Woodhead
Publishing.
Cassin, M. H., Lammerding, A. M., Todd, E. C. D., Ross, W., & McColl,
R. S. (1998). Quantitative risk assessment for Escherichia coli O157:H7
in ground beef hamburgers. International Journal of Food Microbiol-
ogy, 41, 21–44.
Empey, W. A., & Scott, W. J. (1939). Investigations on chilled beef. Part 1:
Microbial contamination acquired in the meat works. Council of
Scientific and Industrial Research (Australia), Bulletin no. 126.
Fazil, A. M., Ross, T., Paoli, G., Vanderlinde, P., Desmarchelier, P., &
Lammerding, A. (2002). A probabilistic analysis of Clostridium
perfringens growth during food service operations. International
Journal of Food Microbiology, 73, 315–330.
Gill, C. O. (1984). Prevention of early spoilage of livers. Proceedings of the
30th European Meeting of Meat Research Workers, Bristol, U.K.,
240–241.
Gill, C. O., & Baker, L. P. (1998). Assessment of the hygienic performance
of a sheep carcass dressing process. Journal of Food Protection, 61(3),
329–333.
25
Gill, C. O., & Bryant, J. (1997). Assessment of the hygienic performances
of two beef carcass cooling processes from product temperature
history data or enumeration of bacteria on carcass surfaces. Food
Microbiology, 14(6), 593–602.
Gill, C. O., Friske, M., & Tong, A. K. Q. (1995). Assessment of the
hygienic characteristics of a process for the distribution of processed
meats, and of storage conditions at retail outlets. Food Research
International, 28(2), 131–138.
Gill, C. O., Greer, G. G., & Dilts, B. D. (1998). Predicting the growth
of Escherichia coli on displayed pork. Food Microbiology, 15(2),
235–242.
Gill, C. O., & Jones, T. (1992a). Assessment of the hygienic efficiencies of
two commercial processes for cooling pig carcasses. Food Microbiol-
ogy, 90, 335–342.
Gill, C. O., & Jones, T. (1992b). Evaluation of a commercial process for
collection and cooling of beef offals by a temperature function
integration technique. International Journal of Food Microbiology, 15,
131–143.
Gill, C. O., & Jones, T. (1997). Assessment of the hygienic characteristics
of a process for dressing pasteurised pig carcasses. Food Microbiology,
14(1), 81–91.
Gill, C. O., Jones, S. D. M., & Tong, S. K. W. (1991). Application of a
temperature function integration technique to assess the hygienic
adequacy of a process for spray chilling beef carcasses. Journal of Food
Protection, 54, 731–736.
Gill, C. O., & Phillips, D. M. (1993). Hygienically appropriate time/
temperature parameters for raw meat processing. In Congress pro-
ceedings: 36th international congress of meat science and technology,
Havana, Cuba (pp. 458–470).
Gill, C. O., & Phillips, D. M. (1993b). The efficiency of storage during
distant continental transportation of beef sides and quarters. Food
Research International, 26(4), 239–245.
Gill, C. O., Phillips, D. M., Loeffen, M. P. F., & Bishop, C. (1988). A
computer program for evaluating the hygienic efficiency of meat
processing procedures from product temperature history data. In
Congress proceedings: 34th international congress of meat science and
technology, Brisbane, Australia (pp. 531–532).
Gill, G. O., Taylor, C. M., & Tong, A. K. W. (1995). Use of a
temperature function integration technique to assess the maintenance
of control over an offal cooling process. Fleischwirtschaft, 75(5),
719–721.
Hathaway, S. C. (1993). Risk assessment procedures used by the Codex
Alimentarius Commission and its subsidiary and advisory bodies.
Joint FAO/WHO Food Standards Programme, In 20th Session of the
Codex Alimentarius Commission, Geneva.
Hathaway, S. C., & Cook, R. L. (1997). A regulatory perspective on the
potential uses of microbial risk assessment in international trade.
International Journal of Food Microbiology, 36, 127–133.
Koutsoumanis, K., Taoukis, P. S., & Nychas, G. J. E. (2005). Develop-
ment of a safety monitoring and assurance system for chilled food
products. International Journal of Food Microbiology, 100, 253–260.
Labuza, T. P. (1994). Microbial modeling – A new paradigm in food
science. Food Technology, 48(6), 16.
Lammerding, A., & McKellar, R. (2004). Predictive microbiology in
quantitative risk assessment. In R. C. McKellar & X. Lu (Eds.),
Modeling microbial responses in foods (pp. 263–284). Boca Raton,
USA: CRC Press.
Legan, D. (2007). Application of models and other quantitative micro-
biology tools in predictive microbiology. In S. Brul, M. Zwietering, &
S. Van Gerwen (Eds.). Modelling microorganisms in food (pp. 82–109).
Cambridge, England: Woodhead Publishing.
Leroy, F., & De Vuyst, L. (2007). Modelling microbial interactions in
foods. In S. Brul, M. Zwietering, & S. Van Gerwen (Eds.), Modelling
microorganisms in food (pp. 214–227). Cambridge, England: Wood-
head Publishing.
Lowry, P. D., Gill, C. O., & Pham, Q. T. (1989). A quantitative method of
determining the hygienic efficiency of meat thawing processes. Food
Australia, 41, 1080–1082.
26 T.A. McMeekin / Meat Science 77 (2007) 17–27
Maurimo, D., Reason, J., Johnson, J., & Lee, R. B. (1995). Beyond
aviation human factors. U.K.: Ashgate Publishing (p. 169).
McKellar, R. C., & Lu, X. (Eds.). (2004). Modeling microbial responses in
food (pp. 343). Boca Raton, FL, USA: CRC Press.
McMeekin, T. A. (2004). An essay on the unrealised potential of
predictive microbiology. In R. C. McKellar & X. Lu (Eds.), Modeling
microbial responses in food (pp. 321–335). Boca Raton, FL, USA: CRC
Press.
McMeekin, T. A., & Ross, T. (1993). Use of predictive microbiology in
relation to meat and meat products. In Proceedings of the 39th
international congress of meat science and technology, Calgary, Alberta,
Canada (pp. 257–274).
McMeekin, T. A., Baranyi, J., Bowman, J. P., Dalgaard, P., Kirk, M.,
Ross, T., et al. (2006a). Information systems in food safety
management. International Journal of Food Microbiology, 112, 181–
194.
McMeekin, T. A., Szabo, L., & Ross, T. (2005). Connecting science with
technology to improve microbial food safety management. Interna-
tional Union of Food Science and Technology (IUFoST), Winter
2005/2006, pp. 125–130.
McMeekin, T. A., Mellefont, L. A., & Ross, T. (2007). Predictive
microbiology: Past, present and future. In A. Brul, M. Zwietering, & S.
Van Gerwen (Eds.), Modelling microorganisms in food. Woodhead
Publishing (pp. 7–21).
McMeekin, T. A., Olley, J., Ratkowsky, D. A., & Ross, T. (1993).
Predictive microbiology: Theory and application. Taunton, UK: John
Wiley & Sons (p. 340).
McMeekin, T. A., Olley, J., Ratkowsky, D. A., & Ross, T. (2002).
Predictive microbiology: Towards the interface and beyond. Interna-
tional Journal of Food Microbiology, 73, 395–407.
McMeekin, T. A., Presser, K., Ratkowsky, D. A., Ross, T., Salter, M., &
Tienungoon, S. (2000). Quantifying the hurdle concept by modelling
the bacterial growth/no growth interface. International Journal of Food
Microbiology, 55, 93–98.
McMeekin, T. A., & Ross, T. (1996). Shelf life prediction: status and
future possibilities. International Journal of Food Microbiology, 33,
65–83.
McMeekin, T. A., Smale, N., Jenson, I., Ross, T., & Tanner, D.
(2006b). Combining microbial growth models with near real-time
temperature monitoring technologies to estimate the shelf life and
safety of foods during processing and distribution. Proceedings of
the 2nd International Workshop on Cold Chain Management, Bonn,
71–78.
Mellefont, L. A., McMeekin, T. A., & Ross, T. (2003). Performance
evaluation of a model describing the effects of temperature, water
activity, pH and lactic acid concentration on the growth of
Escherichia coli. International Journal of Food Microbiology, 82,
45–58.
Membré, J-M., Amezquit, A., Bassett, J., Giavedone, C., Blackburn, C.
deW., & Gorris, L. G. M. (2006). A probabilitic modelling approach in
thermal inactivation: Estimation of postprocess Bacillus cereus spore
prevalence and concentration. Journal of Food Protection, 698,
118–129.
Nauta, M. J. (2007). Uncertainty and variability in predictive models of
microorganisms in foods. In S. Brul, M. Zwietering, & S. Van Gerwen
(Eds.), Modelling microorganisms in food (pp. 44–46). Cambridge,
England: Woodhead Publishing.
Olley, J., & Ratkowsky, D. A. (1973). Temperature function integra-
tion and its importance in the storage and distribution of flesh
foods above the freezing point. Food Technology in Australia, 25,
66–73.
Presser, K. A., Ross, T., & Ratkowsky, D. A. (1998). Modelling
the growth limits (growth/no growth interface) of Escherichia
coli as a function of temperature, pH, lactic acid concentration,
and water activity. Applied and Environmental Microbiology, 64,
1773–1779.
Rasmussen, S. K. J., Ross, T., Olley, J., & McMeekin, T. A.
(2002). A process risk model for the shelf life of Atlantic
salmon fillets. International Journal of Food Microbiology, 73,
47–60.
Ratkowsky, D. A., Olley, J., & Ross, T. (2005). Unifying
temperature effects on the growth rate of bacteria and the
stability of globular proteins. Journal of Theoretical Biology, 233,
351–362.
Ratkowsky, D. A., & Ross, T. (1995). Modelling the bacterial growth/no
growth interface. Letters in Applied Microbiology, 2, 29–33.
Ratkowsky, D. A., Ross, T., McMeekin, T. A., & Olley, J. (1991).
Comparison of Arrhenius-type and Belehradek-type models to
predict bacterial growth in foods. Journal of Applied Bacteriology,
15, 13–32.
Reichel, M. P., Phillips, D. M., Jones, R., & Gill, C. O. (1991). Assessment
of the hygienic adequacy of a commercial hot boning process for beef
by a temperature function integration technique. International Journal
of Food Microbiology, 14, 27–42.
Roberts, T. A., Gibson, A. M., & Robinson, A. (1981). Prediction of toxin
production by Clostridium botulinum in pasteurised pork slurry.
Journal of Food Technology, 16, 337–355.
Roberts, T. A., & Jarvis, B. (1983). Predictive modelling of food
safety with particular reference to Clostridium botulinum in model
cured meat systems. In T. A. Roberts & F. A. Skinner (Eds.),
Food microbiology: Advances and prospects (pp. 89–95). New York:
Academic Press.
Rönner, U. (1990). Bioindicator for control of sterility. Food Laboratory
News No. 22 (6), 4, 51–54.
Ross, T. (1999). Predictive food microbiology models in the meat industry
(MSRC003). Meat and Livestock Australia, Sydney, 190 pp.
Ross, T., & Shadbolt, C. (2001). Predicting Escherichia coli inactivation in
uncooked comminuted fermented meat products. Report prepared for
Meat and Livestock Australia, 59 pp.
Ross, T., & McMeekin, T. A. (2003). Modelling microbial growth within
food safety risk assessments. Risk Analysis, 23, 179–197.
Ross, T., Ratkowsky, D. A., Mellefont, L. M., & McMeekin, T. A. (2003).
Modelling the effects of temperature, water activity, pH and lactic acid
concentration on the growth rate of Escherichia coli. International
Journal of Food Microbiology, 82, 33–43.
Scott, W. J. (1936). The growth of microorganisms on ox muscle.I. The
influence of water content of substrate on rate of growth at 1 °C.
Journal of the Council of Scientific and Industrial Research, Australia, 9,
177–190.
Scott, W. J. (1937). The growth of microorganisms on ox muscle.II. The
influence of temperature. Journal of the Council of Scientific and
Industrial Research, Australia, 10, 338–350.
Scott, W. J. (1938). The growth of microorganisms on ox muscle.III.
The influence of 10% CO2 on rates of growth at 1 °C. Journal of
the Council of Scientific and Industrial Research, Australia, 11,
266–277.
Scott, W. J., & Vickery, J. R. (1939). Investigation on chilled beef. Part II.
Cooling and storage in the meat works. Council of Scientific and
Industrial Research, Australia Bulletin no. 129.
Sharpe, A. N. (1980). Food microbiology: A framework for the future.
Springfield, IL: C.C. Thomas (p. 224).
Smelt, J. P. P. M., & Brul, S. (2007). Modelling lag-time in predictive
microbiology with special eference to lag phase of bacterial spores. In
S. Brul, M. Zwietering, & S. Van Gerwen (Eds.), Modelling microor-
ganisms in food. Cambridge, England: Woodhead Publishing (pp. 67–
81).
Smith, M. G. (1987). Calculations of the expected increases of
coliform organisms, Escherichia coli and Salmonella typhimurium,
in raw blended mutton tissue. Epidemiological Infections, 99,
323–331.
Tamplin, M., Baranyi, J., & Paoli, G. (2004). Software programs to
increase the utility of predictive microbiology information. In R. C.
McKellar & X. Lu (Eds.), Modeling microbial responses in food
(pp. 233–242). Boca Raton, FL, USA: CRC Press.
Tienungoon, S., Ratkowsky, D. A., McMeekin, T. A., & Ross, T. (2000).
Growth limits of Listeria monocytogenes as a function of temperature,
 T.A. McMeekin / Meat Science 77 (2007) 17–27 27

pH, NaCl, and lactic acid. Applied and Environmental Microbiology,
66, 4979–4987.
van Asselt, E. D., & Zwietering, M. (2006). A systematic approach to
determine global thermal inactivation parameters for various food
pathogens. International Journal of Food Microbiology, 107, 73–82.
Vialette, M., Pinon, A., Leporq, B., Dervin, C., & Membré, J. M. (2005).
Meta-analysis of food safety information based on a combination of a
relational database and a predictive modelling tool. Risk Analysis, 25,
75–83.
Yiannas, F. (2007). Point of view. Food Protection Trends, 27, 164.
Zwietering, M. H., & Nauta, M. J. (2007). Predictive models in food risk
assessment. In S. Brul, M. Zwietering, & S. Van Gerwen (Eds.),
Modelling microorganisms in food (pp. 110–126). Cambridge, England:
Woodhead Publishing.
 MEAT
SCIENCE
Meat Science 77 (2007) 28–35
www.elsevier.com/locate/meatsci

Managing safety and quality through the red meat chain

a,*
P. Desmarchelier , N. Fegan a, N. Smale b, A. Small a

a
Food Science Australia, P.O. Box 3312, Tingalpa DC, Qld 4173, Australia
b
Food Science Australia, P.O. Box 52, North Ryde, NSW 1670, Australia

Received 21 March 2007; received in revised form 30 April 2007; accepted 30 April 2007

Abstract

To successfully manage food safety and quality risks in meat production, a holistic approach is required. The ideal would be a fully
integrated assurance system, with effective controls applied at all stages. However, the red meat industry is by nature somewhat frag-
mented, and a truly integrated system is not at present achievable in all but a few operations. This paper describes a variety of assurance
initiatives, and explores how targeted research and development can be used to augment assurance programmes by providing underpin-
ning knowledge, using the Australian beef and lamb industry as an example.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Meat safety; Quality assurance; Supply chain

1. The Australian beef, lamb and livestock industry
Australia is among the world’s largest and most success-
ful producers of commercial livestock. Expansive range-
lands, a variety of climatic and environmental conditions
and excellent animal health mean that many breeds of live-
stock thrive in Australia. When the First Fleet landed at
Sydney Cove in 1788 it carried a handful of sheep, cattle
and goats as a source of wool, milk and fresh meat for
the new colony. The Australian environment was relatively
hostile to these European animals and selection of animals
for their resilience and ability to thrive under the difficult
conditions was necessary.
Today, Australia is one of the world’s largest red meat
and livestock exporters with stock in excess of 130 million
head and the industry contributing $15 billion to the econ-
omy. Each year, over 2 million cattle and 14 million sheep
are sold in livestock markets (Table 1), and weekly slaugh-
tering ranges from 133,000 cattle and 308,600 sheep (MLA,
2007a). The national sheep flock of 100 million is down
*
Corresponding author. Tel.: +61 7 32142000; fax: +61 7 33214 2150.
E-mail address: Patricia.Desmarchelier@csiro.au (P. Desmarchelier).
from 117 million in 1960 due to changes in the fibre market
and land use and now drought. The beef cattle herd size is
down from 30 million in the 1970s to 27 million; dairy cat-
tle and goats each approximately 3 million.
Australian red meat is sold around the world, mainly to
Japan and the USA (Table 2). Significant quantities of
sheepmeat are exported to North America, Asia and the
Middle East (MLA, 2007a). Approximately AU$300 mil-
lion worth of edible offal, known as ‘fancy meats’, was
exported during 1999/2000 and cattle hides and sheep skins
are valuable co-products.
The beef cattle feedlot industry is now an important
value-adding component of the Australian beef industry
in response to demand for consistent beef quality. Feedlots
are concentrated in the major agricultural regions with
access to adequate supplies of store cattle, grain and other
feedstuffs. There are currently about 600 accredited feed-
lots representing a total capacity of over a million cattle
(Anon., 2007a). In the past 5 years, the composition of cat-
tle fed for specific markets has moved from 20% domestic/
80% export, to 40% domestic/60% export. Supermarkets
are currently drawing 40–50% of their beef supplies from
feedlots as feedlot animals provide a continuous supply
of consistent quality product (MLA, 2007a).

0309-1740/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.027
 P. Desmarchelier et al. / Meat Science 77 (2007) 28–35 29

Table 1 requirements of the Food Standards Australia and New

Numbers of sheep and cattle sold at livestock markets in 2006 and average Zealand (FSANZ) Food Standards Code (FSANZ, 2002)
weekly kill
and in conjunction with the specific Australian Standard
Species Sold at saleyards Slaughtered per week for the Hygienic Production and Transportation of Meat
Cattle 2,434,000 133,000 and Meat Products for Human Consumption (AS
Lambs 9,003,000 179,000 4696:2007) (Anon., 2007b). The prime objective is to ensure
Older Sheep 5,299,000 201,600
that meat and meat products for human consumption com-
ply with food safety requirements and are wholesome. A
Table 2 whole of food chain approach is taken and incorporates
Australian red meat exports 2006 (tonnes) also the need for accurate identification, traceability, effec-
Importing Beef and Lamb Mutton tive recall and animal welfare objectives.
country veal The Standard reflects the shared responsibility between
USA 319,800 39,800 27,319 (North industry and governments for food safety and quality sys-
America) tems based on risk assessment. Meat processors are
Canada 9,600 NRa required to put into place approved HACCP-based process
Japan 405,800 11,900 32,268 (Asia) management systems, with the processing practices vali-
Korea 149,700 NR
dated by underpinning scientific research. For export, the
Taiwan 28,600 NR
Other Asia 23,882 11,800 processors must also comply with the specific requirements
(China) of the importing country. These processes are monitored
EU 8500 11,800 6700 and certified by the Australian Quarantine and Inspection
Middle East NR 17,700 43,100 Service (AQIS).
Africa NR 15,600 25,600
Other 32,518 45,400 27,913
3. Managing food safety and quality
Total 953,900 146,700 162,900
a
NR: Not individually reported. A number of industry strategies have been implemented
to assure safety and quality throughout the production
Australia is a world leader in the export of commercial chain, and targeted research and development (R&D) is
livestock, exporting more than 4 million sheep, 570,000 cat- used to maximise the industry benefit from science. R&D
tle and up to 50,000 goats annually (MLA, 2007a). Leader- in the Australian red meat industry is funded through a
ship is attributed to Australian animal health status, world system of producer and processor levies and government
leading export standards and ability to produce both tem- expenditure, managed by Meat and Livestock Australia
perate and tropical breeds. (MLA). Outcomes of industry R&D are communicated
to the industry though a number of media, including trade
2. Regulatory control of the Australian red meat industry journals and some specific industry initiatives.
This paper reviews some particularly important initia-
The Australian red meat industry operates under an out- tives and recent research projects in food safety and quality
comes-based food safety programme underpinned by the management in the Australian red meat industry (Fig. 1).

Research and Development Inputs

Pathogen surveys
Risk Assessment

Innovative technologies Cold Chain

Process hygiene investigations

Livestock Production Processing Distribution Consumer

Assurance Programs: • HACCP-based approved arrangements • Advice and

• Farm • E. coli & Salmonella monitoring education
• Feedlot
• Livestock Transport
• Residues monitoring programs

Legislation
Risk Mitigation Strategies

Fig. 1. Key research outcomes and risk mitigation strategies.

30 P. Desmarchelier et al. / Meat Science 77 (2007) 28–35

4. Key research outcomes supporting red meat production 4.2. Microbiology of Australian beef and lamb

4.1. Through-chain risk profile National baseline surveys of the microbiology of Aus-
tralian red meats were undertaken in 1993–94 (Vanderl-
MLA completed a risk profiling exercise encompassing inde, Fegan, Mills, & Desmarchelier, 1999; Vanderlinde,
the entire red meat production chain. The aim was to Shay, & Murray, 1998, 1999), in 1998 (Phillips, Sumner,
attach a risk rating to specific hazard-product pairings Alexander, & Dutton, 2001a, 2001b) and 2004 (Fegan
and thus inform future research and development strategies et al., 2005, Phillips et al., 2006; Phillips et al., 2005). When
to mitigate the risks identified domestically or on interna- differences in sampling and laboratory analysis were taken
tional trade (Sumner, Ross, Jenson, & Pointon, 2005). into account, there were improvements in the microbiolog-
Both qualitative and semi-quantitative risk rating tech- ical criteria of beef carcasses, cartoned beef and cartoned
niques were used to analyse the information, and uncer- sheep meat. Between 1993 and 2004, the mean total viable
tainties encountered were used to identify data gaps that count (TVC) in frozen boneless beef had fallen from
required further research and development. Outcomes of 2.77 log10 cfu/g to 1.18 log10 cfu/g, and sheepmeat from
this profiling are shown in Table 3. Some food safety 3.47 log10 cfu/g to 1.8 log10 cfu/g. Where comparisons are
research priorities MLA has identified include the follow- possible, it appears that Australia’s meat products have
ing (MLA, 2007b): pathogen levels equal to, or lower than, those in other
countries (Phillips, Jordan, Morris, Sumner, & Jenson,
 Benchmarking the levels of contamination of retail and 2005). For example, in 2003, 0.35% of samples from Aus-
export meat and meat products for microorganisms of tralian beef carcasses yielded Salmonella spp., compared
public health concern. with 0.8% in the USA.
 Determining the ecology and virulence of Enterohaem-
orrhagic Escherichia coli (EHEC) and Salmonella spp. 4.3. Surveys of the prevalence of specific pathogens
in cattle and meat processing and the impact of control
strategies; understanding the development of microbial 4.3.1. Shiga toxin producing E. coli (STEC)
antibiotic resistance. The most notable group of human pathogenic STEC are
 Development of effective controls for Listeria monocyt- the enterohaemorrhagic E. coli (EHEC). Several EHEC
ogenes in processed meats. serotypes have been identified in human disease in Austra-
lia including O157, O26 and O111, contrasting to the
northern hemisphere and New Zealand, where E. coli
O157:H7 is the major human EHEC serotype.
These infections are relatively rare in Australia with
Table 3
notification rates for STEC regardless of serotype at
Hazard-product pairings identified as high, medium and low public health
risk for the Australian red meat industry (adapted from Sumner et al., around 0.4/100,000 (OzFoodNet, 2006). In 2001, OzFood-
2005) Net recorded 39% (15/38) of STEC notifications were due
Risk Hazard Product Circumstance to the O157 serotype where serotype information was avail-
rating able. Human and animal isolates of E. coli O157 have dis-
High Clostridium Meals provided by tinct similarities, the majority containing all the virulence
perfringens caterers markers associated with human illness (Fegan & Desmar-
Salmonella Kebabs Cross- chelier, 2002). Although the isolates appear phenotypically
contamination similar, disease incidence is low and this may be associated
in drip trays
with the evolution of different lineages globally (Kim et al.,
Salmonella Meals served in the Cross-
home contamination 2001). The importance of E. coli O157:H7, in the northern
hemisphere, has led to the Australian industry having to
Medium Listeria Ready-to-eat meats Extended
monocytogenes shelf-life
monitor this particular serotype over others to support
Listeria Terrines international trade. In recent surveys of beef cattle from
monocytogenes farms and feedlots through to slaughter, E. coli O157 was
Enterohaemorrhagic Uncooked detected in 13% of cattle faeces, with no significant differ-
Escherichia coli comminuted ences between cattle from feedlots or farms (grass-fed)
(EHEC) and fermented meat
Salmonella (UCFM e.g. Salami)
(Fegan et al., 2004a; Fegan et al., 2005). The O157 concen-
EHEC Undercooked trations, however, were quite variable: 67% of positive
hamburgers samples yielded less than 10 cfu/g, and 8% positive samples
Low Salmonella Cooked Sausage had levels between 103 and 105 cfu/g. High shedding or
Listeria UCFM ‘‘super shedders’’ were shown to be associated with higher
monocytogenes contamination rates at slaughter. In a subsequent study,
EHEC Hamburgers Well-cooked where one animal carried 7.5 · 105 E. coli O157 cfu/g of
in Australia
faeces, 15% of pen floor faecal samples, 24% of oral cavity
 P. Desmarchelier et al. / Meat Science 77 (2007) 28–35
swabs, 44% of hide samples and 6% of pre-chill carcasses
were contaminated although no post-chill carcasses were
contaminated with E. coli O157. In general, the level of
the organism on hides was less than 5 per cm2 (range 0–
22), and on pre-chill carcasses less than 0.12 per cm2. The
results indicated that the practices at the abattoir were in
general sufficient to control the risk of carcass
contamination.
In a survey of retail ground beef and lamb cuts for
STEC no O157, O111 or O26 (the common EHEC sero-
types in Australia) were detected, although STEC were
detected in 16% of ground beef and 40% of lamb samples
(Barlow, Gobius, & Desmarchelier, 2006). The most com-
mon serotypes were O174 and O91 in ground beef, and
O128 and O91 in lamb samples, but none of the isolates
carried the eae gene commonly found in EHEC. The pres-
ence of these STEC, of which only O91 has been associated
with illness, in retail meats highlights the need for a clearer
understanding of STEC in order to interpret their public
health significance.
4.3.2. Salmonella species
Salmonella is one of the more common causes of bacte-
rial foodborne illness in Australia (Hall et al., 2005,
OzFoodNet, 2003). There are approximately 92,000 cases
of salmonellosis per annum, 81,000 of which (87%) are con-
sidered foodborne. Recent work on Salmonella prevalence
in beef cattle presented for slaughter in Australia has dem-
onstrated 6.8% faecal carriage in cattle (Fegan, Vanderl-
inde, Higgs, & Desmarchelier, 2004b). The majority of
samples had less than 10 salmonellas/g, with a few samples
yielding up to 3,000 cfu/g, suggesting that there would be a
low risk of carcass contamination. An ‘‘in herd’’ study was
conducted in which 68% of cattle hides were contaminated
with Salmonella at levels of up to 4.2 cfu/cm2, although
only 2% of the subsequently produced carcasses were con-
taminated (Fegan et al., 2005), demonstrating that
although the pathogen is present in cattle, routine process-
ing in Australian plants effectively minimises the risk of
meat contamination. On positive carcasses, a maximum
concentration of salmonellas was 0.31 per cm2. Similar
studies are underway in the Australian sheep processing
industry.
4.3.3. Staphylococcus aureus
Staphylococcus aureus in meat come from the skin of the
animal and from handling and environmental contamina-
tion during processing. Transmission was studied at an
abattoir. The incidence of coagulase positive staphylococci
on cattle hides at slaughter was found to range from 20% to
68% (Desmarchelier, Higgs, Mills, Sullivan, & Vanderl-
inde, 1999). The subsequent carcass incidence was low
immediately after evisceration (6.5–16.7%), although this
could increase to between 46% and 83% following 72 h
chilling at 10–12 C post-slaughter. Likewise, the numbers
of Staphylococcus increased from below 50 cfu/cm2 before
chilling to 64–112 cfu/cm2 post-chilling. The hands of the
31
workers and static boot dips at the evisceration point were
found to be contaminated. Further investigation of the iso-
lates using Pulsed Field Gel Electrophoresis, PFGE, anal-
ysis was used to demonstrate the isolates from the hands
of workers on the evisceration platform were indistinguish-
able to those on the carcasses prepared, and different from
the isolates collected from cattle hides, non-evisceration
workers or office personnel (Vanderlinde et al., 1999).
However, in the Australian national baseline study over
98% of frozen boneless beef samples tested had below the
target maximum level of 50 cfu/g S. aureus (Phillips
et al., 2005), demonstrating an overall high standard of
meat hygiene. Interventions with the use of gloves and
hand hygiene have been successful in reducing this bacte-
rium on carcasses.
4.3.4. Process hygiene
Under the outcomes-based legislation in Australia, the
red meat industry is permitted to implement ‘‘alternative
procedures’’ to those traditionally used, as long as they
can be validated as having an equivalent food safety out-
come. As such, there has been considerable interest in
developing or adopting ‘‘alternative procedures’’ that will
improve market access, improve product safety, reduce
the impact on the environment or achieve increased effi-
ciency. Some examples of alternative procedures explored
by the Australian meat industry are outlined below.
4.3.5. Hot boning
Traditionally, Australian meat production has followed
the pattern of slaughter, dressing and chilling to a deep
muscle temperature of 7 C prior to boning into wholesale
cuts. Use of a short chilling period (warm-boning), or bon-
ing immediately after dressing (hot-boning), increases the
flow of product, reduces handling and the cost of chilling,
due to the smaller portions and thus faster cooling. Using
predictive microbiology techniques to demonstrate control
of microbial growth, many Australian processors have
been able to implement hot-boning or warm-boning in
their plants. The application of predictive microbiology
has resulted in the development of the Refrigeration Index
that has become part of the approved arrangements under
the Export Control (Meat and Meat Products) Orders and
allows companies to demonstrate whether their refrigera-
tion process is effective (Anon., 2005). Other technologies,
such as electrical stimulation to ensure a rapid fall of mus-
cle pH, are needed to optimise the eating quality of the
product. (Sumner & Krist, 2002, Toohey & Hopkins,
2006).
4.3.6. Sanitation of knives
For many years meat processors worldwide have con-
templated the necessity to dip hand tools such as knives
into hot water at 82 C (180 F). While accepting tools
should be kept clean and sanitised regularly, the relevance
of 82 C as the ideal temperature is questioned in light of
the modern risks to meat safety and there is substantial
32 P. Desmarchelier et al. / Meat Science 77 (2007) 28–35
Table 4
Predicted durations of off-power events for a significant quantity of frozen manufacturing beef to reach 5 C in a well insulated, fully loaded 12 m
refrigerated container (adapted from Smale et al., 2007)
Meat initial temperature (C) Ambient temperature
10 15
20 C 310 h 227 h
18 C 269 h 194 h
16 C 226 h 159 h
14 C 180 h 123 h
12 C 132 h 87 h
environmental cost to maintaining the water-baths. A sur-
vey of abattoirs in 2002 suggested that a medium-sized
plant utilises 300 kL water per day, but if the waterbath
temperature was reduced to 73 C, this would halve (Midg-
ley & Eustace, 2003). Rinsing of the knife under warm run-
ning water led to a significant reduction in microbial counts
on the knife, subsequent dipping in 82 C water gave little
further reduction, and immersing the knife blade in 72 C
for 15 s, or in 75 C for 10 s gave an equivalent micro-
bial reduction (3.2–3.5 log10 E. coli) to immersion in
82 C for 10 s. Alternating two knives, rinsing them in
hand wash water, then immersing them in 60 C water
between use has also been shown to be equivalent (Eustace
et al., 2007).
4.3.7. Carcass decontamination
The use of chemical decontamination for carcasses is
common practice in certain parts of the world, for example
the USA, while other countries also have approved the use
of certain chemicals such as organic acids or acidified
sodium chlorite. Carcass decontamination by chemical
methods or hot water is not widely practised in Australia;
however, processors exporting to markets with stringent
port-of-entry requirements have shown interest in carcass
decontamination technologies, or ‘‘interventions’’, and
MLA have responded by funding an internet-based infor-
mation package for member companies about interven-
tions for the red meat industry, and assisting processors
in conducting in-plant trials.
4.3.8. Maintenance of the cold chain during frozen meat
transport
Much of Australian red meat is exported, very often as
frozen boneless meat in cartons. Operations during trans-
port of refrigerated containers mean that off power events
can occur. With a product such as frozen meat where small
fluctuations in temperature are not critical and there is a
considerable volume of cold product inside the container,
substantial durations without power may not adversely
affect the meat quality. The meat transport industry cur-
rently takes advantage of this, often transporting contain-
ers from abattoirs to ports without active refrigeration.
Guidelines for maximum acceptable durations for such
transport have previously been published (Anon., 1978;
Middlehurst, Parker, & Coffey, 1969); but advances in con-
tainer construction, temperature measurement and model-
20 25 30 35 40
175 h 140 h 115 h 97 h 83 h
147 h 117 h 95 h 80 h 68 h
119 h 93 h 75 h 63 h 53 h
91 h 70 h 57 h 47 h 40 h
63 h 48 h 39 h 32 h 27 h
ling techniques make revisiting these recommendations
sensible.
Recently a series of full scale experiments were con-
ducted in both fully loaded 12 m (40 0 ) and 6 m (20 0 ) con-
tainers measuring meat temperature rises during off
power events (Smale, East, Eddy, & Kang, 2007). The con-
tainers were held under controlled, constant temperature
conditions (10 C, 25 C and 40 C) in an environmental
chamber, purpose built for testing of transport equipment.
Heat transfer was modelled and validated through compar-
ison with experimental data. The validated model was then
used to provide recommendations for maximum off-power
durations as a function of ambient and initial temperatures
and container size and quality. For example, Table 4 gives
the estimated time it would take for containerised beef to
reach a temperature of 5 C from a specified initial prod-
uct temperature (left axis) under a particular external ambi-
ent temperature (top axis) (Smale et al., 2007).
5. Transferring research outcomes to industry
Research programmes can only achieve benefit if the
outcomes of the research are translated and delivered to
the target audience (in this case, the Australian red meat
industry) in a form that is understood and able to be used.
Therefore, MLA and Food Science Australia have funded
a number of knowledge management and technology trans-
fer initiatives.
MINTRAC is a company, owned by the Meat Industry,
which represents the industry on training matters. In the
1990’s it was realised that the Australian meat industry
would benefit from increasing the technical knowledge of
QA managers, and also the skill base of the workers.
Now, MINTRAC, in conjunction with a number of train-
ing providers, manage an Australia-wide accredited train-
ing programme from entry level through to senior
management. Workers in the meat industry as in many
other industries can follow a structured career path. The
training packages are based on underpinning science and
are updated regularly to keep abreast of current thinking
and international requirements.
Meat Industry Services (MIS) is a programme involving
a team of experienced meat scientists, based at Food Sci-
ence Australia’s research laboratory in Queensland, with
expertise in a broad range of topics, including animal wel-
fare, meat quality, refrigeration, meat safety and rendering.
 P. Desmarchelier et al. / Meat Science 77 (2007) 28–35
The team manage an internet-based information service for
the Australian meat industry and produce six newsletters
each year, updating industry on scientific research that
can be utilised in modern production. They are also avail-
able to meat processors to help in problem solving, design-
ing in-plant trials, and validating their HACCP plans.
6. Key risk mitigation strategies in red meat production
6.1. Farm and feedlot
The potential for contamination of meat with microbial
and chemical contaminants begins on the farm, and thus a
truly holistic meat safety programme involves risk mitiga-
tion strategies involving the live animal from its concep-
tion. Meat quality also is significantly influenced by the
health and welfare of the live animal, and a good meat
quality assurance program addresses the events leading
up to the point of slaughter as well as processing variables.
Australia has a number of such assurance schemes for live-
stock, such as Livestock Production Assurance and the
National Feedlot Accreditation Scheme. These programs
require livestock producers to develop a QA system for
their property, based on HACCP principles, and compliant
with current codes of practice on welfare, environment pro-
tection and use of medicines or chemicals (Horchner, Brett,
Gormley, Jenson, & Pointon, 2006).
When animals are traded, they are identified, and
accompanied by a National Vendor Declaration (NVD,
2007), providing relevant ‘‘chain information’’ for safe
meat production, such as medicines administered to the
mob, source, supplementary feeding and grazing history.
Cattle are identified with a radio-frequency identification
device (RFID) which may be incorporated in the ear tag,
or as a rumen bolus, registered on a central database
(NLIS, 2007). This database is being extended to include
sheep and goats.
6.2. Livestock transportation
Australian livestock can travel long distances between
properties and to slaughter. The welfare of these animals
is of prime importance to the Australian meat industry,
and a quality assurance programme for livestock transport,
TruckCare, has been developed in order to optimise animal
welfare and hence meat quality and safety (ARTA, 2007).
6.3. Chemical residues
The Australian Pesticides and Veterinary Medicines
Authority (APVMA) approves the use of medications
and other chemicals (e.g. pesticides, weedkillers) in live-
stock production, in order to minimise the risk of residues
being present in meat (APVMA, 2007). The National Res-
idue Survey (NRS) has been in place since the 1960’s,
although the range of foods and chemicals monitored has
increased dramatically since inception (NRS, 2007; Row-
33
land, Evans, & Walcott, 1997). The NRS is focussed on
trade issues, and monitors livestock feeds as well as human
foods for a range of substances determined following risk
assessment. These may include antimicrobials, anthelmin-
tics, hormonal growth promotants, acaricides, fungicides,
herbicides, fumigants and metals. The NRS does not, for
example, undertake testing for antimicrobial resistance,
which is a health rather than a trade concern at this time.
The NRS is supplemented by the National Antibacterial
Residue Minimisation (NARM) and National Organo-
chlorine Residue Minimisation (NORM) Programs, the
objectives of which are to raise the awareness of the risk
to trade associated with antibiotic or organochlorine resi-
dues above maximum residue limit (MRL) in meat and
assist the red meat industry in minimising such residues.
Results of the surveillance programs demonstrate a high
rate of compliance with the requirements and good practice
in the use of agricultural and veterinary chemicals. In the
2005–2006 period, only 5 incidences of any chemical above
MRL were found in meat, a proportion of less than 0.003%
of analyses (Anon., 2006).
7. Abattoir assurance programs
7.1. HACCP-based approved arrangements
Export registered abattoirs are required to develop a
quality manual for their process, based on HACCP princi-
ples. This manual includes detailed operating procedures,
and each procedure must be validated by reference to pub-
lished scientific literature and verified in plant. These proce-
dures are then agreed with AQIS, and the plant is expected
to operate within the bounds of this Approved Arrange-
ment (AFFA, 2006). Any change to the process requires
a full review, and approval of the updated arrangement.
The aim is to have an integrated system to assure the safety
and integrity of meat and meat products produced for
human consumption.
7.2. E. coli and salmonella monitoring program (ESAM)
To verify its performance-based monitoring systems,
which are integral to food safety assurance, AQIS intro-
duced the generic E. coli and Salmonella Monitoring pro-
gram (ESAM) in 1998. This is a national program of
microbiological monitoring of carcass surfaces, compliant
with the requirements of the US Pathogen Reduction Rule.
Under the program, carcass surfaces of all species of live-
stock slaughtered in Australia are tested for generic E. coli
and Salmonella (Vanderlinde, Jenson, & Sumner, 2005).
7.3. Eating quality
The Australian meat industry uses a science-based grad-
ing programme called Meat Standards Australia (MSA).
MSA uses a palatability assurance programme, based on
HACCP principles (PACCP) to label meat with a guaran-
34 P. Desmarchelier et al. / Meat Science 77 (2007) 28–35
teed grade and best cooking method to maintain premium
eating quality (Polkinghorne, 2006).
8. Future challenges
Market access requirements continue to be a challenge
for industry, with recognition of Australian systems diffi-
cult to negotiate with overseas countries. A significant chal-
lenge affecting the Australian meat industry in current
years is an ongoing drought situation, and much effort is
being made to identify means of conserving water and
minimising unnecessary use. In a benchmarking exercise
in 2004, comparing data from meat processing plants with
similar data from 1998, the key finding was an overall
reduction in water usage, generation of less wastewater
and a reduction in noise and odour complaints (MLA,
2007a). While average energy usage per tonne of meat pro-
duced has remained steady, average raw water use has
decreased by 11%, which equates to an additional 2.5 Bil-
lion l of water being made available to the environment
each year. The survey results also indicated a significant
shift in management attitude towards environmental sus-
tainability. It is now seen as a key aspect of competitive
advantage for meat processors, and there is widespread
adoption of environmental reporting as a part of normal
management reporting on company performance.
9. Conclusion
The Australian red meat industry uses a number of
approaches to manage food safety and quality, including
both voluntary and mandatory assurance schemes. The
utilisation of outcomes-based legislation gives the industry
the opportunity to move forward as science and under-
standing progress. Knowledge transfer programs help to
maximise the industry benefit of research outcomes, assist-
ing Australian processors to maintain their position in the
global marketplace.
Acknowledgements
The authors thank Mr. I. Jenson, MLA, and Mr. P.
Vanderlinde, AQIS, for there critical comments in prepara-
tion of the review.
References
AFFA (2006). Approved arrangement guideline. Australian Government
Department of Agriculture Fisheries and Forestry.
Anon. (1978). Holding of frozen cartoned meat in insulated shipping
containers. Meat Research News Letter, 78, 3.
Anon. (2005). Export control (meat and meat products) orders. AQIS
Notice Number: Meat 2005/04. <http://www.affa.gov.au/corpo-
rate_docs/publications/pdf/quarantine/mid/2005_04.pdf>.
Anon. (2006). Report on results: National residue survey 2005–2006.
Canberra: Department of Agriculture Fisheries and Forestry.
Anon. (2007a). Australian lot feeders association, ALFA. <http://
www.feedlots.com.au>.
Anon. (2007b). Australian standard for the hygienic production and
transportation of meat and meat products for human consumption.
Primary Industries Report Series 80. Canberra: CSIRO Publishing.
APVMA (2007). Australian pesticides and veterinary medicines authority.
<http://www.apvma.gov.au/index.asp>.
ARTA. (2007). Australian road train association. Truck care. <http://
www.arta.org.au/info.asp>.
Barlow, R. S., Gobius, K. S., & Desmarchelier, P. (2006). Shiga toxin-
producing Escherichia coli in ground beef and lamb cuts: results of a
one-year study. International Journal of Food Microbiology, 111, 1–5.
Desmarchelier, P. M., Higgs, G. M., Mills, L., Sullivan, A. M., &
Vanderlinde, P. B. (1999). Incidence of coagulase positive Staphylo-
coccus on beef carcasses in three Australian abattoirs. International
Journal of Food Microbiology, 47, 221–229.
Eustace, I., Midgley, J., Giarrusso, C., Laurent, C., Jenson, I., & Sumner, J.
(2007). An alternative process for cleaning knives used on meat slaughter
floors. International Journal of Food Microbiology, 113, 23–27.
Fegan, N., & Desmarchelier, P. P. (2002). Comparison between human
and animal isolates of shiga toxin-producing Escherichia coli O157
from Australia. Epidemiology and Infection, 128, 357–362.
Fegan, N., Higgs, G., Vanderlinde, P., & Desmarchelier, P. (2005). An
investigation of Escherichia coli O157 contamination of cattle during
slaughter at an abattoir. Journal of Food Protection, 68, 451–457.
Fegan, N., Vanderlinde, P., Higgs, G., & Desmarchelier, P. (2004a). The
prevalence and concentration of Escherichia coli O157 in faeces of
cattle from different production systems at slaughter. Journal of
Applied Microbiology, 97, 362–370.
Fegan, N., Vanderlinde, P., Higgs, G., & Desmarchelier, P. (2004b).
Quantification and prevalence of Salmonella in beef cattle presenting at
slaughter. Journal of Applied Microbiology, 97, 892–898.
Fegan, N., Vanderlinde, P., Higgs, G., & Desmarchelier, P. (2005). A
study of the prevalence and enumeration of Salmonella enterica in
cattle and on carcasses during processing. Journal of Food Protection,
68, 1147–1153.
FSANZ (2002). Australian and New Zealand food standards
code. <http://www.foodstandards.gov.au/thecode/foodstandardscode.
cfm>.
Hall, G., Kirk, M. D., Becker, N., Gregory, J. E., Unicomb, L., Millard,
G., et al. (2005). Estimating foodborne gastroenteritis, Australia.
Emerging Infectious Diseases, 11, 1257–1264.
Horchner, P. M., Brett, D., Gormley, B., Jenson, I., & Pointon, A. M. (2006).
HACCP-based approach to the derivation of an on-farm food safety
program for the Australian red meat industry. Food Control, 17, 497–510.
Kim, J., Nietfeldt, J., Ju, J., Wise, J., Fegan, N., Desmarchelier, P., et al.
(2001). Ancestral divergence, genome diversification, and phylogeo-
graphic variation in subpopulations of sorbitol-negative, b-glucuron-
idase-negative enterohemorrhagic Escherichia coli O157. Journal of
Bacteriology, 183, 6885–6897.
Middlehurst, J., Parker, N. S., & Coffey, M. (1969). The transport of food
in I.S.O. containers. CSIRO Food Preservation Quarterly, 29, 21–26.
Midgley, J., & Eustace, I. (2003). Investigation of alternatives to 82 C
water for knife and equipment sterilisation. North Sydney: Meat and
Livestock Australia.
MLA (2007a). Meat and livestock Australia. Market information. <http://
www.mla.com.au/TopicHierarchy/MarketInformation/default.htm>.
MLA (2007b). Meat and livestock Australia. Food safety. <http://
www.mla.com.au/TopicHierarchy/InformationCentre/FoodQuality
SafetyAndvalueadding/FoodSafety/default.htm>.
NLIS (2007). Meat and livestock Australia. National livestock identifica-
tion system. <http://www.mla.com.au/NR/exeres/2EA7C6DB-F81A-
412D-9A9D-DF4AB9E89EED.htm>.
NRS (2007). National residue survey. <http://www.affa.gov.au/
content/output.cfm?ObjectID=D2C48F86-BA1A-11A1-A2200060B0-
A0 5746>.
NVD (2007). Meat and livestock Australia. National vendor declaration.
<http://www.mla.com.au/topichierarchy/industryprograms/livestock
qualitysystems/nationalvendordeclarations/nvd%2Binformation%2
Brecording%2Bsystems%2B(nirs).htm>.
 P. Desmarchelier et al. / Meat Science 77 (2007) 28–35
OzFoodNet (2003). Foodborne disease in Australia: incidence, notifica-
tions and outbreaks. Annual report of the OzFoodNet network, 2002.
Communicable Disease Intelligence, 27, 209–243.
OzFoodNet (2006). Burden and causes of foodborne disease in Australia:
Annual report of the OzFoodNet network, 2005. Communicable
Disease Intelligence, 30, 300–378.
Phillips, D., Jordan, D., Morris, S., Sumner, J., & Jenson, I. (2005).
Microbiological quality of Australian beef and sheepmeat – Results of
the industry’s third national abattoir survey. North Sydney: Meat and
Livestock Australia.
Phillips, D., Jordan, D., Morris, S., Jenson, I., & Sumner, J. (2006).
Microbiological quality of Australian sheep meat in 2004. Meat
Science, 74, 261–266.
Phillips, D., Sumner, J., Alexander, J. F., & Dutton, K. M. (2001a).
Microbiological quality of Australian beef. Journal of Food Protection,
64, 692–696.
Phillips, D., Sumner, J., Alexander, J. F., & Dutton, K. M. (2001b).
Microbiological quality of Australian sheep meat. Journal of Food
Protection, 64, 697–700.
Polkinghorne, R. J. (2006). Implementing a palatability assured critical
control point (PACCP) approach to satisfy consumer demands. Meat
Science., 74, 180–187.
Rowland, P., Evans, G., & Walcott, J. (1997). In The environment and food
quality. In L. Resources (Ed.). Australia: State of the environment
technical paper series. Central Queensland University Publishing Unit.
35
Smale, N., East, A., Eddy, A., & Kang, S. P. (2007). Modelling meat
temperatures during off-power events in refrigerated shipping containers.
Beijing, China: International Congress of Refrigeration.
Sumner, J., & Krist, K. (2002). The use of predictive microbiology by the
Australian meat industry. International Journal of Food Microbiology,
73, 363–366.
Sumner, J., Ross, T., Jenson, I., & Pointon, A. (2005). A risk microbi-
ological profile of the Australian red meat industry: Risk ratings of
hazard-product pairings. International Journal of Food Microbiology,
105, 221–232.
Toohey, E. S., & Hopkins, D. L. (2006). Eating quality of commercially
processed hot boned sheep meat. Meat Science, 72, 660–665.
Vanderlinde, P., Fegan, N., Mills, L., & Desmarchelier, P. (1999). Use of
pulse field gel electrophoresis for the epidemiological characterisation
of coagulase positive Staphylococcus isolated from meat workers and
beef carcasses. International Journal of Food Microbiology, 48, 81–85.
Vanderlinde, P., Jenson, I., & Sumner, J. (2005). Using national
microbiological data to set meaningful performance criteria for
slaughter and dressing of animals at Australian export abattoirs.
International Journal of Food Microbiology, 104, 155–159.
Vanderlinde, P. B., Shay, B., & Murray, J. (1998). Microbiological quality
of Australian beef carcass meat and frozen bulk packed beef. Journal
of Food Protection, 61, 437–443.
Vanderlinde, P. B., Shay, B., & Murray, J. (1999). Microbiological status
of Australian sheep meat. Journal of Food Protection, 62, 380–385.
 MEAT
SCIENCE
Meat Science 77 (2007) 36–45
www.elsevier.com/locate/meatsci

Application of genomic technologies to the improvement

of meat quality of farm animals
Yu Gao, Ran Zhang, Xiaoxiang Hu, Ning Li *

State Key Laboratory for Agrobiotechnology, China Agricultural University, Beijing 100094, China

Received 15 March 2007; received in revised form 27 March 2007; accepted 27 March 2007

Abstract

Meat quality is of economic importance in farm animals. It is controlled by multigenes and the environment. During the past few
decades, advances in molecular genetics have led to the identification of genes, or markers associated with genes, that affect meat quality.
Work on sequencing farm animal genomes will help us to understand how genes function in various organisms and might be applied in
the field to study the molecular control of meat quality. Candidate gene and genome scans are two main strategies to identify loci asso-
ciated with the trait of meat quality. Several genes that influence meat quality have already been, or are close to being, identified. Some of
them have been applied to the breeding of farm animals by marker-assisted selection. This will accelerate cumulative and permanent
genetic improvement of herds.
 2007 Published by Elsevier Ltd.

Keywords: Genomic; QTL; Farm animals; Meat; Quality; Marker-assisted selection

1. Introduction such as feed, handling details, slaughtering and meat pro-

Much progress in farm animal breeding has been made
in the last few decades, but achieving greater understanding
in the improvement of meat quality was very slow before
molecular markers became an accessible technology with
wide applications in breeding schemes.
The trait which we will call ‘‘meat quality’’ is one of eco-
nomic importance in farm animals. From the appearance
of the raw material, quality requires analysis and classifica-
tion of fat content, composition, tenderness, water-holding
capacity, color, oxidative stability and uniformity. It is
influenced by several factors, such as breed, genotype, feed-
ing, fasting, preslaughter handling, stunning, slaughter
methods, chilling and storage conditions (Rosenvold &
Andersen, 2003). In brief, all the factors affecting meat
quality can be divided into two aspects on the genetic basis
and on management systems. For management systems,
*
Corresponding author. Tel.: +86 10 6273 1146; fax: +86 10 6273 3904.
E-mail address: ninglcau@cau.edu.cn (N. Li).
cessing, there are so many market specifications existing
which have been emphasized for many years (Mullen, Sta-
pleton, Corcoran, Hamill, & White, 2006). For the genetic
basis, the correct selection of breeds or lines is very impor-
tant because the genetic influence on meat quality is very
different among breeds as well as among animals in the
same breed. Selective breeding has been carried out in large
populations for thousands of years. Such strong selection,
especially in recent centuries, has resulted in the accumula-
tion of new mutations with favorable phenotypic effects
(Andersson, 2001). These new mutations can provide
greater options, especially when molecular technology is
used in breeding schemes.
In farm animal breeding, selection is effective for most
traits, with many achievements through a process that
has been slowly accelerating in recent years, particularly
in relation to the trait of meat quality. Indeed, the trait
of meat quality is difficult to improve by traditional selec-
tion because the heritability of meat quality is low and
the measure for the quality trait is difficult, expensive,
and only possible after slaughter. Moreover, meat quality

0309-1740/$ - see front matter  2007 Published by Elsevier Ltd.

doi:10.1016/j.meatsci.2007.03.026
 Y. Gao et al. / Meat Science 77 (2007) 36–45
is determined polygenically. There exists little knowledge of
genes and their interactions that are involved in meat prop-
erties. The understanding of meat quality on a genetic basis
is scanty, and needs to be addressed at first. With a better
knowledge of meat quality genetics, molecular breeding
aspects will be developed. At the present time, the main
task of genetics is to identify factors in the molecular or
biological components of meat quality that will be useful
for marker-assisted selection in breeding, i.e. giving
‘‘designer meat’’ by genetic and molecular methods. The
finding of potential genes or chromosome regions responsi-
ble for meat quality will benefit the producers. In these
years, a lot of work has been carried out in this field to find
potential genes or chromosome regions associated with the
meat quality trait in different farm animals, such as pig,
cattle, sheep and chicken. It will be an exciting period when
the molecular breeding method is used to ‘‘design the
meat’’ on the base of genomics knowledge and technolo-
gies. The aim of this paper is to introduce the developing
field of farm animal genomics, to describe the strategies
and technologies to map and characterize meat quality trait
loci, to review genes and their function that are involved in
determining meat quality, and to discuss marker-assisted
selection and its use in farm animal breeding.
2. The developing field of farm animal genomics
Genome research in farm animals progressed rapidly in
recent years, moving from linkage maps to genome
sequence. The work on farm animal genome sequencing
began in the early 1990s, and assists in the understanding
of how genomics function in various organisms (Fadiel,
Anidi, & Eichenbaum, 2005). It will be applied in different
fields, such as to study the molecular components of meat
quality and ways of improving it.
In March 2004, the first draft of the chicken genome was
released (Antin & Konieczka, 2005). In May 2006, the
Genome Sequencing Center submitted an improved 6.6X
draft chicken genome assembly. The chicken genome has
a haploid size of 1200 Mb. It is not only a food animal that
comprises 41% of the meat produced in the world, but also
a model organism for studies of disease and biology (Dequ-
eant & Pourquie, 2005). With the chicken genome
sequence, especially the genome-wide screening in three
chicken breeds yielding a set of 2.8 million SNP markers
(Wong et al., 2004), chicken breeders will have a frame-
work for investigating polymorphisms of informative
quantitative traits to continue the directed evolution of
these species (Fadiel et al., 2005).
In October 2004, the first draft of the bovine genome
sequence was deposited in a free public database. In June
2005, the Bovine Genome Sequencing Project released the
second version of the bovine genome, Btau_2.0, which
was a 6.2X whole genome shotgun assembly. In August
2006, the Bovine Genome Sequencing Project released the
third version of the bovine genome assembly, Btau_3.1,
which was a 7.15X mixed assembly that combines whole
37
genome shotgun sequence with BAC sequence (http://
www.ncbi.nlm.nih.gov/genome/guide/cow/).
Plans for the porcine genome project are underway. The
whole-genome sequencing for pig began in 2005. What is
more, the ‘‘Sino-Danish Pig Genome Project’’ has pub-
lished the pig genome sequence with <1X coverage (Wern-
ersson et al., 2005). In the near future, the sequencing of
the porcine genome will allow gene markers for specific
traits to be identified, assisting breeders in generating pig
stocks selection.
The focal point of interest in sheep is the quest to max-
imize sheep meat and wool production. The third genetic
linkage map of sheep included well over 1000 markers
and BAC libraries have been produced (Maddox et al.,
2001) and contigs assembled around several regions of
interest to individual laboratories (Womack, 2005). The
sheep genome sequence will be completed by the Sheep
Genome Project.
Besides the genome sequence database, there are several
major sources of information on farm animal genomics.
The ArkDB (http://www.thearkdb.org) is one of them
and it is available through the Roslin Institute (UK) and
Texas A&M University (USA) (Hu et al., 2001). The
ArkDB provides detailed genomic mapping data on
chicken, cow, pig and sheep, including data on PCR prim-
ers, genetic linkage map assignments of specific loci and
markers, and cytogenetic map assignments (Law & Archi-
bald, 2000). All this genome information, especially the
sequence information will permit cross-species comparison
of the effects of candidate gene allelic polymorphisms on
meat quality. Integration of the linkage map, the RH
map, the BAC fingerprinted contig (FPC) map and the
genome sequence information will help to identify the can-
didate genes affecting meat quality.
3. The strategies and technologies to map and characterize
the loci of the meat quality trait
Meat quality trait has a multifactorial background and
is controlled by an unknown number of quantitative trait
loci (QTL). The main goal of genome research in farm ani-
mals is to map and characterize trait loci. There are two
main strategies to identify trait loci, association tests using
candidate genes and genome scans based on linkage map-
ping DNA markers (Andersson, 2001). The information
of the meat quality trait loci can be applied in breeding pro-
grams by using marker-assisted selection (MAS).
3.1. Candidate gene approach
The candidate gene approach studies the relationship
between the trait of interest and known genes that may
be associated with the physiological pathways underlying
the trait (Andersson, 2001). If the candidate gene is a true
causative gene, this approach can be very powerful and can
detect loci having even small effects. The implementation of
a candidate gene approach consists of the following steps:
38 Y. Gao et al. / Meat Science 77 (2007) 36–45
(1) construction or collection of a resource population, (2)
phenotyping of the specific components of the trait(s), (3)
selection of genes or functional polymorphisms that poten-
tially could affect the traits, (4) genotyping of the resource
population for the selected genes or functional polymor-
phisms, (5) statistical analysis of the phenotypic and geno-
typic data (Da, 2003). This is an effective way to find the
genes associated with the trait. So far, a number of genes
have been investigated.
Although great progress has been made by using candi-
date gene approach, the limitation of this approach is obvi-
ous. The candidate gene tests must be interpreted with
caution because spurious results can occur due to linkage
disequilibria to linked or non-linked ‘‘causative’’ genes, or
because the significance thresholds have not been adjusted
properly when testing multiple candidate genes (Anders-
son, 2001). It also requires prior knowledge of the physiol-
ogy of the specific trait, which is not always available. On
the other hand, there are sometimes many candidate genes
for the trait; it will take a long time to evaluate all of them.
Furthermore, some genes that are not part of the known
physiological pathways may contribute to the trait under
investigation. Before the genome was sequenced, the selec-
tion of candidate genes was based upon cross-species gene
comparison. It was difficult to analyse and compare genes
from different species. The genome sequence, especially
the SNP map, would solve a lot of questions on the selec-
tion of candidate genes. For example, the chicken SNP
map provides abundant and useful SNP information to find
the causative mutations among broiler, layer, silkie and red
jungle fowl, which will accelerate the research.
3.2. Genome scans approach
The genome scan approach studies the relationship
between a trait and markers selected across the genome to
identify chromosomal locations associated with the trait
(Andersson, 2001). The genome scan will find out the map
location of a trait locus with a major effect. It involves the
following steps: (1) design and construction of a resource
population, (2) phenotyping traits of the resource popula-
tion, (3) selection of genetic markers, (4) genotyping of the
population for selected markers, (5) construction of linkage
maps, (6) statistical analysis of the phenotypic and geno-
typic data derived from the resource population (Da, 2003).
The design of a resource population is the first step in
genome scanning, which will decide whether QTL can be
found. A resource population is a population generated
for a particular research purpose, with phenotypic informa-
tion and sufficient DNA supply for genotyping; for exam-
ple, an intercross between two divergent breeds of farm
animal or a population containing particularly interesting
phenotypic data. Because the design of the intercross
between two divergent populations of farm animals has a
more powerful approach for QTL mapping, it is used in
most resource populations, e.g. the wild boar and large
white pigs (Andersson-Eklund et al., 1998; Andersson-Ekl-
und, Uhlhorn, Lundeheim, Dalin, & Andersson, 2000;
Knott et al., 1998; Nii et al., 2005, 2006), Asian and Euro-
pean breeds of pig (Jungerius et al., 2004; Stratil et al.,
2006), Bos taurus (Angus) and Bos indicus (Brahman) cattle
(Kim, Farnir, Savell, & Taylor, 2003), the broiler and the sil-
kie (Gao et al., 2006), the jungle fowl and the layer (Wright
et al., 2006). In this design, the F1 animals show a high het-
erozygosity at marker loci and, in particular, at those loci
that account for phenotypic differences between the two
populations. If the construction of a resource population
is too costly, a population containing particularly interest-
ing phenotypic data can be used, e.g. a half-sib families com-
prising >1000 progeny from a single male can be collected in
a species, such as cattle (Mizoshita et al., 2004). And selec-
tive genotyping and DNA pooling are cost-saving strategies
that apply to any of the above designs (Carleos, Baro,
Canon, & Corral, 2003; Taylor & Phillips, 1996).
Using the genome scan, the large amount of QTLs can be
obtained in farm animals. It can provide a useful bridge to
link genome information with phenotype. There is a data-
base, Animal quantitative trait loci (QTL) database (Ani-
malQTLdb), which contains all publicly available QTL
data on farm animal species for the past decade (Hu, Fritz,
& Reecy, 2007). It comprises QTL location (chromosome,
location, location span), flanking markers, peak markers,
test statistics, QTL effects and traits. And it is easy to locate
and make comparisons within and between species with this
database. To date (January 19, 2007), there are 1675 QTL
from 110 publications representing 281 different traits for
pig, 846 QTL from 55 publications representing 91 different
traits for cattle, 657 QTL from 45 publications representing
112 different traits for chicken (http://www.animalge-
nome.org/QTLdb/). Several groups have worked on the
identification of QTL controlling meat quality and most
of them are about pork quality. QTL with significant influ-
ences on meat quality were located on almost every porcine
chromosome. In PigQTLDB (Hu et al., 2005), there are 12
types for meat quality and total 1405 QTL for meat quality,
such as 595 QTL for Anatomy, 18 QTL for Chemical, 25
QTL for Conductivity, 1 QTL for Enzyme Activity, 64
QTL for Fat Composition, 439 QTL for Fatness, 26 QTL
for Flavor, 79 QTL for Meat Color, 5 QTL for Odor, 66
QTL for PH, 3 QTL for Stiffening and 84 QTL for Texture
at present (January 19, 2007) (http://www.animalge-
nome.org/QTLdb/pig.html). Although, a genome scan
can give full genome coverage for that trait, it will fail to
detect trait loci with smaller effects if they do not reach
the stringent significance of the thresholds.
3.3. Fine mapping
The ultimate goal of a genome scan approach is to iden-
tify the genes that underlie polygenic traits and gain a bet-
ter understanding of their physiological and biochemical
functions. In fact, a region of QTL often spans 5–30 cM,
and it is too large to find the target genes, so fine mapping
needs to be done. It is a step towards restricting the region
 Y. Gao et al. / Meat Science 77 (2007) 36–45
of interest and the number of potential candidate genes.
The goal of fine mapping is mapping a QTL to a narrow
chromosome region so that the physical QTL affecting
the phenotype can be identified and cloned.
Fine mapping of the QTL of interest began in the rele-
vant region by adding genetic markers and increasing the
marker density to the linkage map. As new maps were
obtained, new QTL analyses were performed (Grapes,
Dekkers, Rothschild, & Fernando, 2004). However, an
extremely valuable tool is the full DNA sequence of the
region, where specific QTL are located. And high-resolu-
tion mapping of QTL can be obtained by backcross exper-
iments using animals that carry recombinant chromosome
(Horvat & Medrano, 1995). The development of advanced
intercross lines is another useful approach for improving
the resolution of the map in intercross experiments (Darv-
asi & Soller, 1995; Vitarius, Sehayek, & Breslow, 2006).
3.4. Position cloning
Followed by fine mapping to narrow the region, one of
meat quality trait loci will be position cloned. Mapping the
location of the meat quality trait is an extremely important
step, and is an integral part of the investigation that sub-
stantially limits the number of candidate genes that may
be identified (Flaherty, Herron, & Symula, 2005).
The cloning of QTL is a challenge for several reasons. A
major hurdle is the poor precision in the location of QTL,
because the relationship between the genotype and the phe-
notype is more complex than it is for a monogenic trait;
therefore one cannot directly identify recombinants
between markers and trait loci (Andersson, 2001). How-
ever, it is possible to determine the genotype at a QTL indi-
rectly by progeny testing.
Pure positional cloning is rarely used in farm animals. In
practice, the candidate gene approach is often combined
with the genome scan strategy, which is positional candi-
date cloning. Position candidate cloning is the work to be
progressed as the main strategy for this purpose. In farm
animals, it often relies heavily on the exploitation of com-
parative data and will become even more powerful with the
completion of the human map and the generation of infor-
mative databases on gene function and gene expression
patterns. The positional cloning of PRKAG3 gene is a typ-
ical example in the farm animal mapping research. To
clone this gene, the linkage mapping, linkage disequilib-
rium mapping, radiation mapping, construction of a
BAC contig, BAC sequencing and bioinformatic analysis
were used. Eventually the result was the identification of
the causative missense mutation in PRKAG3 genes. This
mutation is associated with the ‘‘acid meat’’ phenotype in
pigs (Milan et al., 2000).
3.5. Marker-assisted selection in breeding programs
After identification of QTL, the genes and causative
mutations, a further step is normally required for practical
39
use of this variation in selection and breeding programs.
This information can be applied in breeding programs by
using marker-assisted selection (MAS).
Three types of observable polymorphic genetic loci can
be distinguished: (1) direct markers – loci that are the func-
tional mutations, which causative for the trait of interest;
(2) LD markers – loci that are in linkage disequilibrium
across the population with the functional mutation; (3)
LE markers – loci that are in linkage equilibrium with
the functional mutation in outbred populations (Dekkers,
2004). The three types of marker loci differ not only in
methods of detection, but also in their application in selec-
tion programs. Selection on these three types of markers is
referred to as gene-assisted selection (GAS), LD markers-
assisted selection (LD–MAS), and LE marker-assisted
selection (LE–MAS).
GAS is currently the most practical and commercially
viable system, because GAS gives certainty to the inheri-
tance of the desired trait and so can be used for selection
across the population. To LD–MAS, the extent of linkage
in the genome and the population history decide its utility
(Dekkers, 2004). Because linkage disequilibrium extends
far in cattle breeds (Farnir et al., 2000), it is possible to
use markers that are in linkage equilibrium with the QTL
in the general population (Dekkers, 2004). However, LD
markers are difficult to identify and there are only few
detected in livestock populations to date (Freking et al.,
2002). Although, LE markers are readily identifiable,
LE–MAS is too difficult to use in commercial breeding.
LE studies are currently most useful in the initial stages
of marker identification, such as finding QTLs that segre-
gate between breeds (Mullen et al., 2006).
MAS can lead to decisions that predict improved perfor-
mance in farm animals, and will accelerate cumulative and
permanent genetic improvement of the herd. Because the
meat quality trait is a complex trait controlled by multi-
genes and the environment, it is important to realize that
markers for MAS are only one or few of many genes that
contribute towards that trait. The presence or absence of
the numerous other ‘‘unmarked’’ genes and the production
environment will determine whether an animal actually dis-
plays the desired phenotype (Alison, 2006). Implementa-
tion of MAS requires careful consideration of issues
ranging from sample collection and storage, genotyping
and data analysis (Dekkers, 2004). Furthermore, MAS
should be seen as a tool to assist with, not as a replacement
for traditional selection techniques.
4. Important genes affecting meat quality traits in farm
animals
4.1. Important genes affecting pork meat quality trait
Pork is the major red meat source worldwide. Selection
based on body composition, in particular the relative pro-
portion of muscle to fat tissue, is very important in meat-
producing animals. During the past 50 years, there has
40 Y. Gao et al. / Meat Science 77 (2007) 36–45
been an intensive selection for lean growth in several
breeds. Several genes that influence body composition have
already been identified or are close to being identified. The
Halothane gene, the RN gene and the IGF2 gene have been
reported to have a direct influence on pork quality. Gene
tests to remove their negative effects have been carried
out by breeding companies.
4.1.1. The Halothane gene
The ‘‘Halothane’’ gene, referred to as the PSS (Porcine
Stress Syndrome) gene, was one of the first trait loci to
be characterized at the molecular level in pigs. It causes
malignant hyperthermia, which can be triggered by stress
or exposure to the anaesthetic gas, halothane. Since the
1960s, the halothane gene has been observed to be closely
associated with the development of PSE (Pale, Soft and
Exudative) meat (Briskey, 1964), which was first described
as ‘‘muscle degeneration’’. PSE meat is caused by an exten-
sive protein denaturation due to the low pH values early
post-mortem in combination with the simultaneously high
temperatures (Briskey, 1964). Christian (1972) suggested
that there was a monogenic variation in stress-susceptibil-
ity. The homozygous pigs for the halothane gene reacted
to halothane gas, which led to the gene being named
accordingly. After a few years, the gene was mapped to
chromosome 6 (Davies, Harbitz, Fries, Stranzinger, &
Hauge, 1988), and the PSS phenotype is caused by an
R614C missense mutation in RYR1 gene (ryanodine recep-
tor 1, an ion channel that regulates the release of Ca2+ in
skeletal muscle; see Fujii et al., 1991). A recessive mutation
of this gene causes susceptibility to malignant hyperther-
mia, which can be triggered by stress or exposure to the
anesthetic gas, halothane.
A number of studies have focused on the effect of the
halothane gene on pork meat quality (Apple et al., 2005;
Hamilton, Ellis, Miller, McKeith, & Parrett, 2000; Kerth
et al., 2001). Pigs homozygous and heterozygous for the
halothane gene have higher carcass yield and lean percent-
age (Herfort Pedersen et al., 2001; Klont, Lambooy, & van
Logtestijn, 1994; Leach, Ellis, Sutton, McKeith, & Wilson,
1996; McPhee & Trout, 1995). Although, the positive effect
of halothane is prominent, the negative effect on WHC
(water-holding capacity) and color is obvious, especially
in porcine stress syndrome. The carriers of the halothane
gene are highly susceptible to stress. Even during careful
handling, the stress accompanying preslaughter treatment
is sufficient to trigger a higher rate of post-mortem glycol-
ysis in pigs that are both homozygous and heterozygous for
the halothane gene, being most severe in the homozygous
pigs (Lundstrom, Essen-Gustavsson, Rundgren, Edfors-
Lilja, & Malmfors, 1989).
4.1.2. The RN gene
The Rendment Napole (RN) gene has been discussed
over the last few decades as well as the effects of the halo-
thane gene. In 1986, Naveau suggested the existence of a
single major gene affecting the meat quality trait, the
Napole technological yield or in French, Rendement Napole
(Rosenvold & Andersen, 2003). Further studies supported
the hypothesis of a major gene with two alleles for RN
trait, a dominant mutant allele RN and a recessive normal
RN+ allele. The RN gene identified in the Hampshire breed
is associated with reduced Napole yield and leaner car-
casses. Meat from carriers is associated with poor process-
ing quality when producing ham, and low pH in the meat
because of post-mortem degradation of glycogen, referred
to as ‘‘acid meat’’.
As in the previous description, the work to identify the
RN gene was typically positional cloning studies in farm
animals. The RN gene was mapped to chromosome 15
by Milan et al. (1996b). They mapped the RN gene on
SSC 15q2.4–2.5 between the flank markers SW2053 and
SW936 (Milan et al., 1996a). The RN gene was located
exactly at SSC 15q2.5 by RH mapping (Milan et al.,
1998). In 2000, Milan et al. (2000) reported that the RN
phenotype was caused by an R225Q mutation in the
PRKAG3 gene, which encodes a muscle-specific isoform
of the regulatory c-subunit of adenosine monophosphate
activated protein kinase (AMPK). The distinct phenotype
of the RN-mutation indicates that PRKAG3 plays a key
role in the regulation of energy metabolism in skeletal mus-
cle. The other mutations were found in the PRKAG3 gene
associated with meat quality of pork loin (Lindahl et al.,
2004a, 2004b). Recently, a comparative proteome study
of the RN gene effect showed that the expression profiles
of several enzymes of the glycogen storage pathways were
differentially regulated in a pattern, and the integrated data
from this proteome study indicates that regulation of glu-
cose transport was severely affected in mutant animals
(Hedegaard et al., 2004). Further studies of these mutations
were of great interest in order to explain molecular mecha-
nisms that influence ‘‘drip loss’’ in porcine meat (Otto
et al., 2007).
The RN phenotype is associated with elevated glyco-
gen content in the sarcoplasma, as well as in the lysosomal
compartment, of glycolytic muscle cells. The RN gene has
no effect on early post-mortem pH values, but results in a
lower pH24h value, which again is associated with a higher
reflectance (lighter meat) and inferior WHC (Le Roy,
Naveau, Elsen, & Sellier, 1990). In other words, the posi-
tive effect of the RN gene cannot be ignored. The RN
allele is another example of a mutation that has probably
increased in frequency because of selection for meat con-
tent in pigs. It occurs at a high frequency only in the
Hampshire breed and increases glycogen content in muscle
by 70% (Milan et al., 2000).
4.1.3. The IGF2 gene
IGF2, insulin-like growth factor 2, is implicated in myo-
genesis and lean meat content. A mutation, a single base (A
for G substitution) of IGF2 (position 3072 in intron 3), was
described as quantitative trait nucleotide (QTN), which
was the cause of a major QTL effect on muscle growth
and fat deposition in pigs (Van Laere et al., 2003). This
 Y. Gao et al. / Meat Science 77 (2007) 36–45
QTL affecting muscle growth and fat deposition was first
located on chromosome 2. It explained 15–30% of the phe-
notypic variation in muscle mass and 10–20% of the varia-
tion in back-fat thickness (Jeon et al., 1999; Nezer et al.,
1999). Haplotype sharing refined the location with major
effect on muscle mass to a 250 kb chromosome segment
containing the porcine IGF2 gene (Nezer et al., 2003).
The genotyping of pig populations for IGF2 could be an
important part of breeding programs in the future because
mutation in IGF2 may have potential influence on meat
quality and quantity. The A/G mutation has been acciden-
tally selected for in- breeding schemes based on production
performance and/or lean meat deposition. Carrodeguas
et al. (2005) evaluated a rapid assay based on real time
PCR (RT-PCR) to detection this mutation.
4.1.4. Other genes
Besides the above mentioned major genes, genes in the
leptin pathway are proving profitable in association studies
with growth and backfat, e.g. the MC4R gene (Kim, Lar-
sen, Short, Plastow, & Rothschild, 2000; Meidtner et al.,
2006; Park, Carlborg, Marklund, & Andersson, 2002). It
was clearly associated with reduced feed intake, faster
growth and less backfat. A particular haplotype in the cal-
pastatin (CAST) gene in pork was associated with sensory
tenderness and other meat quality traits (Ciobanu et al.,
2004; Kocwin-Podsiadla, Kuryl, Krzecio, Zybert, &
Przybylski, 2003).
4.2. Important genes affecting beef meat quality trait
Important aspects for beef meat quality maybe include
meat pH, marbling and tenderness. They have been studied
for many years and some SNPs have been found in differ-
ent genes. Several markers for tenderness have been devel-
oped for the inhibitor of calpain gene, the calpastatin and
the calpain I genes (Casas et al., 2006; Lonergan, Ernst,
Bishop, Calkins, & Koohmaraie, 1995; Page et al., 2004;
White et al., 2005). Furthermore, the leptin gene, the thyro-
globulin gene, the DGAT1 gene and the growth hormone
gene were associated with the marbling trait, and the myo-
genic regulatory factors gene family was another group of
important candidate genes for muscle growth (Mullen
et al., 2006).
4.3. Important genes affecting sheep meat quality trait
The requirements in sheep meat quality trait are fewer
than for pork or beef, and the focuses are muscularity
and fat deposition. A QTL with major effects on muscular-
ity was located on chromosome 2 in sheep (Laville et al.,
2004), and accounted for 5–25% of variance. A higher den-
sity map was constructed to refine the map position of this
QTL because the confidence interval of the QTL spanned
10 cM in previous reports. Finally, it was fine-mapped to
a chromosome interval encompassing the myostain
(GDF8) gene. There was a G–A transition in the 3’UTR
41
that created a target site for mir1 and mir206 in the
GDF8 allele of Texel sheep, which caused translational
inhibition of GDF8 gene and contributed to the muscular
hypertrophy of Texel sheep. A new class of regulatory
mutations was identified that might make an important
contribution to the heritability of quantitative traits (Clop
et al., 2006).
The callipyge phenotype has also been studied for many
years, being mapped to ovine chromosome 18 using a bat-
tery of bovine chromosome 21 markers (Cockett et al.,
1994). The effect of the callipyge phenotype on traits
affected the muscle growth and meat tenderness (Freking
et al., 1999; Koohmaraie, Shackelford, Wheeler, Lonergan,
& Doumit, 1995). A single base change causes the callipyge
muscle hypertrophy phenotype, being the only known
example of polar overdominance in mammals (Freking
et al., 2002). Although, it causes hypertrophy in sheep but-
tocks, it yields less tender and palatable meat as a conse-
quence. Because of the high similarity of genomes
between sheep and cattle, many of the markers developed
in cattle may also be useful in sheep, and this will accelerate
the sheep breeding program.
4.4. Important genes affecting chicken meat quality trait
In chicken, more investigations focus on fat deposition,
such as the percentage of hypodermal fat, abdominal fat,
and intramuscular fat in breast and legs. The intramuscular
fat (IMF) was in positive correlation with meat flavor and
succulency, especially tenderness of meat (Le Bihan-Duval,
Millet, & Remignon, 1999). Increasing IMF and control-
ling fatty deposition is an increased interest in improving
meat quality. QTL for fatness, being found in various
crosses between different breeds of chickens. The gene for
extracted extracellular fatty acid binding protein (EX-
FABP) was considered as a candidate locus or linked to
a major gene that significantly affected abdominal fat traits
in chicken. The DNA marker discovered by Wang et al.
(2001) can be used as a molecular marker for assisted selec-
tion on chicken fat trait.
5. Applying MAS for improvement of meat quality
In the past, genetic change has been slow due to selec-
tion technique methodology with low accuracy. In today’s
farm animal breeding systems, directional and actual
change can come more quickly because of the improved
DNA-based technology and genetic markers for selection.
Table 1 shows candidate genes associated with meat quality
in farm animals.
MAS allows for the accurate selection of specific DNA
variations that have been associated with a measurable dif-
ference or effect on meat quality trait. Marker information
can be used to increase the frequency of the marker that is
positively associated with the trait of interest by selecting
for animals carrying two copies of that marker, and against
those carrying no copies of it (Alison, 2006).
42 Y. Gao et al. / Meat Science 77 (2007) 36–45

Table 1
Candidate genes associated with meat quality in farm animals
Animal Candidate genes Traits References
Pig HAL Meat quality/stress Fujii et al., 1991
MC4R Growth and fatness Kim et al., 2000
RN, PRKAG3 Meat quality Milan et al., 2000
AFABP/FABP4 Intramuscular fat Gerbens et al., 1998
HFABP/FABP3 Intramuscular fat Gerbens et al., 1999
CAST Tenderness Ciobanu et al., 2004
IGF2 Growth and fatness Van Laere et al., 2003
Cattle CAST Meat tenderness Lonergan et al., 1995
Leptin/Thyroglobulin Marbling Mullen et al., 2006
Myostation Growth and composition Grobet et al., 1998
DGAT1 Intramuscular fat/marbling Thaller et al., 2003
Sheep Callipyge Muscular hypertrophy Freking et al., 2002
GDF8 Muscular hypertrophy Clop et al., 2006
Chicken EX-FABP Fatness Wang et al., 2001
L-FABP Fatness Wang et al., 2006

Table 2
With pork, MAS has been most successful in the elimi-
List of companies offering commercially available markers for meat
nation of undesirable traits, ensuring more consistent meat quality
quality from the population. Since, 1991 when PIC first
Company Animal Traits
started using a DNA marker test to detect presence of
Biogenetic services Cattle Meat quality
the halothane gene, PIC has pioneered the use of DNA
(http:// Pig Porcine stress syndrome
markers in pig breeding. PICmarqTM is used to make more www.biogeneticservices.com/)
accurate selection decisions in different traits, such as Bovigen Cattle Tenderness
growth, lean, efficiency, meat quality (http://www.pic.- (http://www.bovigen.com/) Quality grade
com/). From 1990, some of the biggest international breed- Marbling
Genmark Cattle Double-muscling phenotype
ing companies decided to remove the halothane gene from
(http://www.genmarkag.com/) Pig Porcine stress syndrome
their selection lines, and some countries eliminated the GeneSeek Cattle Tenderness
presence of the halothane gene from their selection lines (http://www.geneseek.com) Pig RN gene
several years ago, e.g. Denmark, The Netherlands, Sweden Porcine stress syndrome
and Switzerland (Rosenvold & Andersen, 2003). Genetic Visions Cattle Tenderness
(http://www.geneticvisions.net/)
With beef, DNA markers associated with marbling and
Igenity Cattle Quality grade
tenderness have become commercially available, such as (http://www.igenity.com/) Marbling
GeneStarTM Tenderness, which test for favourable SNP’s PIC Pig Porcine stress syndrome/
at major genes known to be involved in meat tenderness RN/. . .
(Alison, 2006). GeneStarTM Tenderness was the first (http://www.pic.com/)
multi-marker single trait test commercially available to
the beef industry. It combines three markers, T1, T2 and
T3, for two important genes, Calpastatin and Calpain. selection programs for cattle, pig, sheep and so on.
The combined effects of these three markers account for Although, there are some commercially available markers
2.5 pounds of WBSF (Warner–Bratzler shear force). for pork meat and beef meat, there are few for sheep meat.
The GeneSTAR Quality Grade is another DNA genetic For sheep breeding, the markers focus on Spider Lamb
marker panel test offered by Bovigen LLC. The marker Syndrome and Scrapie Resistance Tests. The commercially
in this panel is T5, which identifies the presence of the thy- marker for chicken meat quality is the same state as sheep.
roglobulin gene associated with meat quality grade and
marbling. And it is the only quality grade or marbling test 6. Conclusions
to have passed by the NBCEC (National Beef Cattle Eval-
uation Consortium). The IgenityCarcass Composition is To date, advances in molecular genetics have led to the
another panel test for beef, which includes information identification of genes or markers associated with genes
on quality grade, tenderness, marbling and so on. The that affect the meat quality trait. The molecular basis of
DNA test panel of the Geneseek Company includes two meat quality is being revealed by functional genomics
SNP in the CAPN1 gene that have been linked to meat ten- approaches. These will help us to gain further insight into
derness. In the United States, the commercially available the biological components and the development of meat
markers for carcass quality traits have been validated by quality. It gives greater opportunities to enhance genetic
the NBCEC (http://www.nbcec.org/nbcec/index.html). improvement program in farm animal through marker-
Table 2 lists companies that provide markers for breeding assisted selection.
 Y. Gao et al. / Meat Science 77 (2007) 36–45
Acknowledgements
This work was supported by Grants National Major Ba-
sic Research Development Program and National Natural
Science Foundation of China.
References
Alison, V. E. (2006). DNA-Based Technologies. Beef Sire Selection
Manual, National Beef Cattle Evaluation Consortium, 66–73.
Andersson, L. (2001). Genetic dissection of phenotypic diversity in farm
animals. Nature reviews. Genetics, 2(2), 130–138.
Andersson-Eklund, L., Marklund, L., Lundstrom, K., Haley, C. S.,
Andersson, K., Hansson, I., et al. (1998). Mapping quantitative trait
loci for carcass and meat quality traits in a wild boar · large white
intercross. Journal of Animal Science, 76(3), 694–700.
Andersson-Eklund, L., Uhlhorn, H., Lundeheim, N., Dalin, G., &
Andersson, L. (2000). Mapping quantitative trait loci for principal
components of bone measurements and osteochondrosis scores in a
wild boar · large white intercross. Genetical Research, 75(2), 223–230.
Antin, P. B., & Konieczka, J. H. (2005). Genomic resources for chicken.
Developmental Dynamics, 232(4), 877–882.
Apple, J. K., Kegley, E. B., Maxwell, C. V., Jr., Rakes, L. K., Galloway,
D., & Wistuba, T. J. (2005). Effects of dietary magnesium and short-
duration transportation on stress response, post-mortem muscle
metabolism, and meat quality of finishing swine. Journal of Animal
Science, 83(7), 1633–1645.
Briskey, E. J. (1964). Etiological status and associated studies of pale, soft,
exudative porcine musculature. Advances in Food Research, 13, 89–178.
Carleos, C., Baro, J. A., Canon, J., & Corral, N. (2003). Asymptotic
variances of QTL estimators with selective DNA pooling. Journal of
Heredity, 94(2), 175–179.
Carrodeguas, J. A., Burgos, C., Moreno, C., Sanchez, A. C., Ventanas, S.,
Tarrafeta, L., et al. (2005). Incidence in diverse pig populations of an
IGF2 mutation with potential influence on meat quality and quantity:
An assay. Meat Science, 71(3), 577–582.
Casas, E., White, S. N., Wheeler, T. L., Shackelford, S. D., Koohmaraie,
M., Riley, D. G., et al. (2006). Effects of calpastatin and micro-calpain
markers in beef cattle on tenderness traits. Journal of Animal Science,
84(3), 520–525.
Christian, L. L. (1972). A review of the role of genetics in animal stress
susceptibility and meat quality. In R. G. Cassens, F. Giesler, & Q.
Kolb (Eds.), Proceedings of the Pork Quality Symposium (pp. 91–115).
Wisconsin: University Wisconsin.
Ciobanu, D. C., Bastiaansen, J. W., Lonergan, S. M., Thomsen, H.,
Dekkers, J. C., Plastow, G. S., et al. (2004). New alleles in calpastatin
gene are associated with meat quality traits in pigs. Journal of Animal
Science, 82(10), 2829–2839.
Clop, A., Marcq, F., Takeda, H., Pirottin, D., Tordoir, X., Bibe, B., et al.
(2006). A mutation creating a potential illegitimate microRNA target
site in the myostatin gene affects muscularity in sheep. Nature Genetics,
38(7), 813–818.
Cockett, N. E., Jackson, S. P., Shay, T. L., Nielsen, D., Moore, S. S.,
Steele, M. R., et al. (1994). Chromosomal localization of the callipyge
gene in sheep (Ovis aries) using bovine DNA markers. Proceedings of
the National Academy of Sciences of the United States of America,
91(8), 3019–3023.
Da, Y. (2003). Statistical analysis and experimental design for mapping
genes of complex traits in domestic animals. Yi Chuan Xue Bao, 30(12),
1183–1192.
Darvasi, A., & Soller, M. (1995). Advanced intercross lines, an exper-
imental population for fine genetic mapping. Genetics, 141(3),
1199–1207.
Davies, W., Harbitz, I., Fries, R., Stranzinger, G., & Hauge, J. G. (1988).
Porcine malignant hyperthermia carrier detection and chromosomal
assignment using a linked probe. Animal Genetics, 19(3), 203–212.
43
Dekkers, J. C. (2004). Commercial application of marker- and gene-
assisted selection in livestock: Strategies and lessons. Journal of Animal
Science, 82(E-Suppl.), E313–E328.
Dequeant, M. L., & Pourquie, O. (2005). Chicken genome: New tools and
concepts. Developmental Dynamics, 232(4), 883–886.
Fadiel, A., Anidi, I., & Eichenbaum, K. D. (2005). Farm animal genomics
and informatics: An update. Nucleic Acids Research, 33(19),
6308–6318.
Farnir, F., Coppieters, W., Arranz, J. J., Berzi, P., Cambisano, N.,
Grisart, B., et al. (2000). Extensive genome-wide linkage disequilib-
rium in cattle. Genome Research, 10(2), 220–227.
Flaherty, L., Herron, B., & Symula, D. (2005). Genomics of the future:
Identification of quantitative trait loci in the mouse. Genome Research,
15(12), 1741–1745.
Freking, B. A., Keele, J. W., Shackelford, S. D., Wheeler, T. L.,
Koohmaraie, M., Nielsen, M. K., et al. (1999). Evaluation of the ovine
callipyge locus: III. genotypic effects on meat quality traits. Journal of
Animal Science, 77(9), 2336–2344.
Freking, B. A., Murphy, S. K., Wylie, A. A., Rhodes, S. J., Keele, J. W.,
Leymaster, K. A., et al. (2002). Identification of the single base change
causing the callipyge muscle hypertrophy phenotype, the only known
example of polar overdominance in mammals. Genome Research,
12(10), 1496–1506.
Fujii, J., Otsu, K., Zorzato, F., de Leon, S., Khanna, V. K., Weiler, J. E.,
et al. (1991). Identification of a mutation in porcine ryanodine
receptor associated with malignant hyperthermia. Science, 253(5018),
448–451.
Gao, Y., Hu, X. X., Du, Z. Q., Deng, X. M., Huang, Y. H., Fei, J., et al.
(2006). A genome scan for quantitative trait loci associated with body
weight at different developmental stages in chickens. Animal Genetics,
37(3), 276–278.
Gerbens, F., Jansen, A., van Erp, A. J., Harders, F., Meuwissen, T. H.,
Rettenberger, G., et al. (1998). The adipocyte fatty acid-binding
protein locus: Characterization and association with intramuscular fat
content in pigs. Mammal Genome, 9(12), 1022–1026.
Gerbens, F., van Erp, A. J., Harders, F. L., Verburg, F. J.,
Meuwissen, T. H., Veerkamp, J. H., et al. (1999). Effect of genetic
variants of the heart fatty acid-binding protein gene on intramus-
cular fat and performance traits in pigs. Journal of Animal Science,
77(4), 846–852.
Grapes, L., Dekkers, J. C., Rothschild, M. F., & Fernando, R. L. (2004).
Comparing linkage disequilibrium-based methods for fine mapping
quantitative trait loci. Genetics, 166(3), 1561–1570.
Grobet, L., Poncelet, D., Royo, L. J., Brouwers, B., Pirottin, D., Michaux,
C., et al. (1998). Molecular definition of an allelic series of mutations
disrupting the myostatin function and causing double-muscling in
cattle. Mammal Genome, 9(3), 210–213.
Hamilton, D. N., Ellis, M., Miller, K. D., McKeith, F. K., & Parrett, D.
F. (2000). The effect of the Halothane and Rendement Napole genes on
carcass and meat quality characteristics of pigs. Journal of Animal
Science, 78(11), 2862–2867.
Hedegaard, J., Horn, P., Lametsch, R., Sondergaard Moller, H.,
Roepstorff, P., Bendixen, C., et al. (2004). UDP-glucose pyrophos-
phorylase is upregulated in carriers of the porcine RN-mutation in the
AMP-activated protein kinase. Proteomics, 4(8), 2448–2454.
Herfort Pedersen, P., Oksbjerg, N., Karlsson, A. H., Busk, H., Bendixen,
E., & Henckel, P. (2001). A within litter comparison of muscle fibre
characteristics and growth of halothane carrier and halothane free
crossbreed pigs. Livestock Production Science, 73, 15–24.
Horvat, S., & Medrano, J. F. (1995). Interval mapping of high growth
(HG), a major locus that increases weight gain in mice. Genetics,
139(4), 1737–1748.
Hu, Z. L., Dracheva, S., Jang, W., Maglott, D., Bastiaansen, J.,
Rothschild, M. F., et al. (2005). A QTL resource and comparison
tool for pigs: PigQTLDB. Mammalian Genome, 16(10), 792–800.
Hu, Z. L., Fritz, E. R., & Reecy, J. M. (2007). AnimalQTLdb: A livestock
QTL database tool set for positional QTL information mining and
beyond. Nucleic Acids Research, 35(Database issue), D604–D609.
44 Y. Gao et al. / Meat Science 77 (2007) 36–45
Hu, J., Mungall, C., Law, A., Papworth, R., Nelson, J. P., Brown, A.,
et al. (2001). The ARKdb: Genome databases for farmed and other
animals. Nucleic Acids Research, 29(1), 106–110.
Jeon, J. T., Carlborg, O., Tornsten, A., Giuffra, E., Amarger, V.,
Chardon, P., et al. (1999). A paternally expressed QTL affecting
skeletal and cardiac muscle mass in pigs maps to the IGF2 locus.
Nature Genetics, 21(2), 157–158.
Jungerius, B. J., van Laere, A. S., Te Pas, M. F., van Oost, B. A.,
Andersson, L., & Groenen, M. A. (2004). The IGF2-intron3-G3072A
substitution explains a major imprinted QTL effect on backfat
thickness in a Meishan · European white pig intercross. Genetical
Research, 84(2), 95–101.
Kerth, C. R., Carr, M. A., Ramsey, C. B., Brooks, J. C., Johnson, R. C.,
Cannon, J. E., et al. (2001). Vitamin-mineral supplementation and
accelerated chilling effects on quality of pork from pigs that are
monomutant or noncarriers of the halothane gene. Journal of Animal
Science, 79(9), 2346–2355.
Kim, J. J., Farnir, F., Savell, J., & Taylor, J. F. (2003). Detection of
quantitative trait loci for growth and beef carcass fatness traits in a
cross between Bos taurus (Angus) and Bos indicus (Brahman) cattle.
Journal of Animal Science, 81(8), 1933–1942.
Kim, K. S., Larsen, N., Short, T., Plastow, G., & Rothschild, M. F.
(2000). A missense variant of the porcine melanocortin-4 receptor
(MC4R) gene is associated with fatness, growth, and feed intake traits.
Mammalian Genome, 11(2), 131–135.
Klont, R. E., Lambooy, E., & van Logtestijn, J. G. (1994). Effect of
dantrolene treatment on muscle metabolism and meat quality of
anesthetized pigs of different halothane genotypes. Journal of Animal
Science, 72(8), 2008–2016.
Knott, S. A., Marklund, L., Haley, C. S., Andersson, K., Davies, W.,
Ellegren, H., et al. (1998). Multiple marker mapping of quantitative
trait loci in a cross between outbred wild boar and large white pigs.
Genetics, 149(2), 1069–1080.
Kocwin-Podsiadla, M., Kuryl, J., Krzecio, E., Zybert, A., & Przybylski,
W. (2003). The interaction between calpastatin and RYR1 genes for
some pork quality traits. Meat Science, 65(2), 731–735.
Koohmaraie, M., Shackelford, S. D., Wheeler, T. L., Lonergan, S.
M., & Doumit, M. E. (1995). A muscle hypertrophy condition in
lamb (callipyge): Characterization of effects on muscle growth
and meat quality traits. Journal of Animal Science, 73(12), 3596–
3607.
Laville, E., Bouix, J., Sayd, T., Bibé, B., Elsen, J. M., Larzul, C., et al.
(2004). Effects of a quantitative trait locus for muscle hypertrophy
from Belgian Texel sheep on carcass conformation and muscularity.
Journal of Animal Science, 82(11), 3128–3137.
Law, A. S., & Archibald, A. L. (2000). Farm animal genome databases.
Briefings in Bioinformatics, 1(2), 151–160.
Leach, L. M., Ellis, M., Sutton, D. S., McKeith, F. K., & Wilson, E. R.
(1996). The growth performance, carcass characteristics, and meat
quality of halothane carrier and negative pigs. Journal of Animal
Science, 74(5), 934–943.
Le Bihan-Duval, E., Millet, N., & Remignon, H. (1999). Broiler meat
quality: Effect of selection for increased carcass quality and estimates
of genetic parameters. Poultry Science, 78(6), 822–826.
Le Roy, P., Naveau, J., Elsen, J. M., & Sellier, P. (1990). Evidence for a
new major gene influencing meat quality in pigs. Genetical research,
55(1), 33–40.
Lindahl, G., Enfalt, A. C., von Seth, G., Joseli, A., Hedebro-Velander, I.,
Andersen, H. J., et al. (2004a). A second mutant allele (V199I) at the
PRKAG3 (RN) locus-II. Effect on colour characteristics of pork loin.
Meat Science, 66(3), 621–627.
Lindahl, G., Enfalt, A. C., von Seth, G., Josell, A., Hedebro-Velander, I.,
Andersen, H. J., et al. (2004b). A second mutant allele (V199I) at the
PRKAG3 (RN) locus- I. Effect on technological meat quality of pork
loin. Meat Science, 66(3), 609–619.
Lonergan, S. M., Ernst, C. W., Bishop, M. D., Calkins, C. R., &
Koohmaraie, M. (1995). Relationship of restriction fragment length
polymorphisms (RFLP) at the bovine calpastatin locus to calpastatin
activity and meat tenderness. Journal of Animal Science, 73(12),
3608–3612.
Lundstrom, K., Essen-Gustavsson, B., Rundgren, M., Edfors-Lilja, I., &
Malmfors, G. (1989). Effect of halothane genotype on muscle
metabolism at slaughter and its relationship with meat quality: A
within-litter. Meat Science, 25(4), 251–263.
Maddox, J. F., Davies, K. P., Crawford, A. M., Hulme, D. J., Vaiman, D.,
Cribiu, E. P., et al. (2001). An enhanced linkage map of the sheep
genome comprising more than 1000 loci. Genome Research, 11(7),
1275–1289.
McPhee, C. P., & Trout, G. R. (1995). The effects of selection for lean
growth and the halothane allele on carcass and meat quality of pigs
transported long and short distances to slaughter. Livestock Production
Science, 42, 55–62.
Meidtner, K., Wermter, A. K., Hinney, A., Remschmidt, H., Hebebrand,
J., & Fries, R. (2006). Association of the melanocortin 4 receptor with
feed intake and daily gain in F2 Mangalitsa · Pietrain pigs. Animal
Genetics, 37(3), 245–247.
Milan, D., Bidane, J. P., Le Roy, P., Chevalet, C., Woloszyn, N., Caritez,
J. C., et al. (1998). Current status of QTL detection in large
white · Meishan crosses in France. Proceeding of the sixth world
congress on genetics applied to livestock production, Armidale, Australia,
26, 414–417.
Milan, D., Jeon, J. T., Looft, C., Amarger, V., Robic, A., Thelander,
M., et al. (2000). A mutation in PRKAG3 associated with excess
glycogen content in pig skeletal muscle. Science, 288(5469), 1248–
1251.
Milan, D., Woloszyn, N., Giteau, M., Navas, A., Yerle, M., Rogel-
Gaillard, C., et al. (1996a). Toward the identification of RN gene
involved in meat quality in pigs. Animal Genetics, 27, 114.
Milan, D., Woloszyn, N., Yerle, M., Le Roy, P., Bonnet, M., Riquet, J.,
et al. (1996b). Accurate mapping of the ‘‘acid meat’’ RN gene on
genetic and physical maps of pig chromosome 15. Mammalian
Genome, 7(1), 47–51.
Mizoshita, K., Watanabe, T., Hayashi, H., Kubota, C., Yamakuchi,
H., Todoroki, J., et al. (2004). Quantitative trait loci analysis for
growth and carcass traits in a half-sib family of purebred Japanese
Black (Wagyu) cattle. Journal of Animal Science, 82(12), 3415–
3420.
Mullen, A. M., Stapleton, P. C., Corcoran, D., Hamill, R. M., & White,
A. (2006). Understanding meat quality through the application of
genomic and proteomic approaches. Meat Science, 74(1), 3–16.
Nezer, C., Collette, C., Moreau, L., Brouwers, B., Kim, J. J., Giuffra, E.,
et al. (2003). Haplotype sharing refines the location of an imprinted
quantitative trait locus with major effect on muscle mass to a 250 kb
chromosome segment containing the porcine IGF2 gene. Genetics,
165(1), 277–285.
Nezer, C., Moreau, L., Brouwers, B., Coppieters, W., Detilleux, J.,
Hanset, R., et al. (1999). An imprinted QTL with major effect on
muscle mass and fat deposition maps to the IGF2 locus in pigs. Nature
Genetics, 21(2), 155–156.
Nii, M., Hayashi, T., Mikawa, S., Tani, F., Niki, A., Mori, N., et al.
(2005). Quantitative trait loci mapping for meat quality and muscle
fiber traits in a Japanese wild boar · large white intercross. Journal of
Animal Science, 83(2), 308–315.
Nii, M., Hayashi, T., Tani, F., Niki, A., Mori, N., Fujishima-Kanaya, N.,
et al. (2006). Quantitative trait loci mapping for fatty acid composi-
tion traits in perirenal and back fat using a Japanese wild boar · large
white intercross. Animal Genetics, 37(4), 342–347.
Otto, G., Roehe, R., Looft, H., Thoelking, L., Knap, P. W., Rothschild,
M. F., et al. (2007). Associations of DNA markers with meat quality
traits in pigs with emphasis on drip loss. Meat Science, 75(2), 185–
195.
Page, B. T., Casas, E., Quaas, R. L., Thallman, R. M., Wheeler, T. L.,
Shackelford, S. D., et al. (2004). Association of markers in the bovine
CAPN1 gene with meat tenderness in large crossbred populations that
sample influential industry sires. Journal of Animal Science, 82(12),
3474–3481.
 Y. Gao et al. / Meat Science 77 (2007) 36–45
Park, H. B., Carlborg, O., Marklund, S., & Andersson, L. (2002).
Melanocortin-4 receptor (MC4R) genotypes have no major effect on
fatness in a large white · wild boar intercross. Animal Genetics, 33(2),
155–157.
Rosenvold, K., & Andersen, H. J. (2003). Factors of significance for pork
quality – A review. Meat Science, 64(3), 219–237.
Stratil, A., Van Poucke, M., Bartenschlager, H., Knoll, A., Yerle, M.,
Peelman, L. J., et al. (2006). Porcine OGN and ASPN: Mapping,
polymorphisms and use for quantitative trait loci identification for
growth and carcass traits in a Meishan · Pietrain intercross. Animal
Genetics, 37(4), 415–418.
Taylor, B. A., & Phillips, S. J. (1996). Detection of obesity QTLs on mouse
chromosomes 1 and 7 by selective DNA pooling. Genomics, 34(3),
389–398.
Thaller, G., Kuhn, C., Winter, A., Ewald, G., Bellmann, O., Wegner, J.,
et al. (2003). DGAT1, a new positional and functional candidate gene
for intramuscular fat deposition in cattle. Animal Genetics, 34(5),
354–357.
Van Laere, A. S., Nguyen, M., Braunschweig, M., Nezer, C., Collette, C.,
Moreau, L., et al. (2003). A regulatory mutation in IGF2 causes a
major QTL effect on muscle growth in the pig. Nature, 425(6960),
832–836.
Vitarius, J. A., Sehayek, E., & Breslow, J. L. (2006). Identification of
quantitative trait loci affecting body composition in a mouse intercross.
Proceedings of the National Academy of Sciences of the United States of
America, 103(52), 19860–19865.
45
Wang, Q. G., Li, N., Deng, X. M., Lian, Z. X., Li, H., & Wu, C. X. (2001).
Single nucleotide polymorphism analysis on chicken extracelluar fatty
acid binding protein gene and its associations with fattiness trait.
Science in China, 44(4), 429–434.
Wang, Q., Li, H., Li, N., Leng, L., & Wang, Y. (2006). Tissue expression
and association with fatness traits of liver fatty acid-binding protein
gene in chicken. Poultry Science, 85(11), 1890–1895.
Wernersson, R., Schierup, M. H., Jorgensen, F. G., Gorodkin, J., Panitz,
F., Staerfeldt, H. H., et al. (2005). Pigs in sequence space: A 0.66X
coverage pig genome survey based on shotgun sequencing. BMC
Genomics, 6(1), 70.
White, S. N., Casas, E., Wheeler, T. L., Shackelford, S. D., Koohmaraie,
M., Riley, D. G., et al. (2005). A new single nucleotide polymorphism
in CAPN1 extends the current tenderness marker test to include cattle
of Bos indicus, Bos taurus and crossbred descent. Journal of Animal
Science, 83(9), 2001–2008.
Womack, J. E. (2005). Advances in livestock genomics: Opening the barn
door. Genome Research, 15(12), 1699–1705.
Wong, G. K., Liu, B., Wang, J., Zhang, Y., Yang, X., Zhang, Z.,
et al. (2004). A genetic variation map for chicken with 2.8
million single-nucleotide polymorphisms. Nature, 432(7018), 717–
722.
Wright, D., Kerje, S., Lundstrom, K., Babol, J., Schutz, K., Jensen, P.,
et al. (2006). Quantitative trait loci analysis of egg and meat
production traits in a red junglefowl · white leghorn cross. Animal
Genetics, 37(6), 529–534.
 MEAT
SCIENCE
Meat Science 77 (2007) 46–54
www.elsevier.com/locate/meatsci

Identification of pork quality parameters by proteomics

a,* b
Dick F.M. van de Wiel , Wei Li Zhang
a
Animal Sciences Group of Wageningen, University and Research Institute, Animal Breeding and Genomics Centre, ID-Lelystad, P.O. Box
65, 8200AB Lelystad, The Netherlands
b
Anhui Agricultural University, College of Animal Science and Aquatic Science, P.O. Box 241, 130 Changjiang Road, Hefei 230036, PR China

Received 14 March 2007; received in revised form 13 April 2007; accepted 16 April 2007

Abstract

A major parameter for quality of pork is its waterholding capacity (WHC). Prediction of WHC immediately after slaughter would be
of benefit both to slaughterhouses for reasons of better logistics and/or branding of premium-meat, and to consumers for improved qual-
ity of meat products such as ham.
In our pilot study on proteome analysis of porcine muscle by two-dimensional electrophoresis, we have identified at least three and
possibly eight significantly changed proteins that may serve as marker proteins for waterholding capacity. The most clearly identified
proteins are creatine phospho kinase M-type (CPK), desmin, and a transcription activator (SWI/SNF related matrix-associated actin-
dependent regulator of chromatin subfamily A member1, SNF2L1). A possible mechanism of how these proteins may influence
WHC is discussed. An optimised protocol for protein extraction that provides for sufficient amounts of relatively pure proteins has been
developed. Further studies are needed to validate and extend our preliminary results.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Meat quality; Marker proteins; Proteomics; Driploss

1. Introduction meat/food processing companies, are becoming more

Prediction of meat quality in the slaughter line on the
day of slaughter is very much desired, from both a logistic
and an economical point of view. The vast progress in the
production of lean meat during the last decades has largely
taken place at the expense of meat quality, causing serious
problems in the tenderness and the waterholding capacity
(drip loss) of meat. Drip loss may vary between 2% and
12%, which causes a direct economical loss because of
reduced weight. Indirect losses are caused by reduced pos-
sibilities for processing of meat into high quality products
such as ham. Rapid on-site prediction of meat quality,
preferably even before the meat is chilled, may allow for
allocation of carcasses to different production lines or for
branding of meat (premium-meat, etc.). End-users, like
*
Corresponding author. Tel.: +31 320 238526; fax: +31 320 238050.
E-mail addresses: dick.vandewiel@wur.nl (D.F.M. van de Wiel),
zwl123@mail.hf.ah.cn (W.L. Zhang).
aware that ‘‘raw’’ meat proteins must have consistent and
specific quality and functionality characteristics. The latest
consumer demands have already led to product differentia-
tion and greater pressure on the value of meat quality
parameters, especially tenderness, juiciness and flavor of
fresh and value-added meat products (Sosnicki, Pommier,
Klont, Newman, & Plastow, 2003). Moreover, tenderiza-
tion through proteolytic degradation, and waterholding
capacity are interrelated phenomena, as shown by a pub-
lished model for possible causes of driploss (Kristensen &
Purslow, 2001).
1.1. Tenderness
It is well established that meat tenderizes during storage;
however, the underlying biochemical and physiochemical
mechanisms during the tenderization processes are still a
matter of dispute (Lametsch et al., 2003). In general, post-
mortem degradation of muscle proteins is an important

0309-1740/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.017
 D.F.M. van de Wiel, W.L. Zhang / Meat Science 77 (2007) 46–54
factor in the meat tenderization process (Koohmaraie,
1996; Hopkins & Thompson, 2002a), as postmortem degra-
dation of several structural proteins including troponin T,
nebulin, titin, vinculin, desmin and dystrophin has been
demonstrated using one-dimensional SDS–PAGE and
immuno-blotting (Hopkins & Thompson, 2002b). This is
supported by the fact that titin, a giant muscle protein
spanning from the Z-line to M-line region, and nebulin,
which runs parallel with the thin filaments to the Z-line
(Bandman & Zdanis, 1988), both have been shown to
degrade faster in tender meat than in tough meat (Fritz
& Greaser, 1991; Huff-Lonergan, Parrish, & Robson,
1995). Moreover, electron microscopy studies of the myofi-
brils during tenderization have shown that the attachment
of the sarcolemma to myofibrils and the junctions
between myofibrils at the level of the Z-disk and M-line
are disrupted postmortem (Taylor, Geesink, Thompson,
Koohmaraie, & Goll, 1995), which suggests that proteoly-
sis of the costamere proteins desmin, dystrophin, and
vinculin, which attach the myofibrils to the sarcolemma, is
taking place. Strong correlations have been found between
postmortem degradation of troponin T and shear force
(Huff-Lonergan et al., 1996; Steen, Claeys, Uytterhaegen,
DeSmet, & Demeyer, 1997; Wheeler & Koohmaraie, 1999).
However, whether the degradation of troponin T is an essen-
tial part of the tenderization process or just a marker for
postmortem proteolytic activity is still unclear.
Proteome analysis by two-dimensional gel electrophore-
sis (2DE) and matrix-assisted laser desorption/ionization
time-of-flight mass spectrometry (MALDI-TOF MS) fur-
ther extended the number of observed degrading proteins
(in M. longissimus dorsi of pigs during the first 72 h post-
mortem) to 103, which included actin, myosin HC, myosin
LC I and II, CapZ and cofilin (Lametsch et al., 2003).
More recently, a tentative 2DE-based reference map
encompassing 133 proteins was prepared by Hwang, Park,
Kim, Cho, and Lee (2005).
Despite the fact that postmortem degradation of several
structural proteins has been studied in great detail, it is still
far from established whether any of these are directly
responsible for the tenderization of meat.
The observation that desmin and vinculin are degraded
with conditioning in bovine, porcine and chicken meat has
generally led to the hypothesis, that meat tenderization
involves degradation of intracellular cytoskeletal, i.e. inter-
mediate filament (IF) structures. This may take place both
by easing the separation of myofibrils and therefore weak-
ening the lateral strength of meat, and by degradation of
costameres which may weaken the sarcolemma, thereby
increasing calcium influx and calpain activity (Purslow,
Morrison, & Mielche, 1997).
1.2. Waterholding capacity
An accepted model of changes in waterholding capacity
in muscle postmortem involves lateral shrinkage of myofi-
brils as the principle mechanism driving water loss (Kris-
47
tensen & Purslow, 2001). Lateral shrinkage of myofibrils,
muscle fibers and entire muscle fascicles, generates drip
channels between fascicles. This mechanism requires that
lateral connections formed by IFs through the costameres
to extracellular structures are sufficiently strong to laterally
displace fluid released by shrinkage of individual myofibrils
to the extracellular spaces (Purslow et al., 1997).
It should be noted, however, that apart from involve-
ment of IF there are many other factors, both genetic
and non-genetic, that can influence waterholding capacity
(Henckel, Karlsson, Oksbjerg, & Petersen, 2000). Newly
discovered polymorph expression products in this respect
are the melanocortin-4 receptor (Kim, Larsen, Short, Pla-
stow, & Rothschild, 2000), the gamma3 subunit of AMP-
activated proteinkinase (Milan et al., 2000), and the ryan-
odine receptor (Henckel et al., 2000). Non-genetic factors
include stress and nutrition status of the animal before
slaughter, creatine phosphate (CP) and glycogen content,
ultimate pH and fibre type composition of meat, stunning
method and cooling regime.
From a perspective of proteome analysis we consider it
of importance, that IF-proteins seem to be implicated in
two primary quality parameters of meat: tenderness and
waterholding capacity.
1.3. Aim of this study
In this study we wanted to analyse the proteome of mus-
cle tissue collected immediately after slaughter, and search
for individual proteins that correlate with drip loss as mea-
sured at 5 days after slaughter. Identification of such (a)
protein(s) could help predict the WHC of meat at a time
that decisions in the slaughterhouse can be made, i.e. dur-
ing the time period between killing of the animal and chill-
ing of the carcass. A more general aim was the assessment
of proteomics as a method for the identification of marker
proteins that can predict meat quality.
2. Materials and methods
2.1. Animals and housing
It was hypothesized, that animals with little room for
exercise may have less trained muscles and therefore
impaired meatquality after slaughter, as compared to ani-
mals kept in a more stimulating housing system. By intro-
ducing two housing systems, ‘‘enriched’’ and ‘‘barren’’, we
though it possible to induce extremes with respect to
meatquality parameters, such as waterholding capacity.
Animals, housing system and experimental set-up were
essentially as described earlier (Klont et al., 2001) except
that the area per animal was 1.83 m2 and 0.80 m2 for the
enriched and barren housing system, respectively, whereas
in the previous study the difference was much smaller
(1.16 m2 and 0.84 m2, respectively).
The experiment was set up in two replicates of 24 pigs.
Twelve litters of a commercial crossbred line, equally
48 D.F.M. van de Wiel, W.L. Zhang / Meat Science 77 (2007) 46–54

divided over the two different housing systems were used mined in triplicate after a 1-h blooming period by
for selection of piglets. From each litter 2 piglets were measuring L*, a*, and b* values with a Minolta CM525i
selected (one male and one female), based on health and Chromometer.
weight. Selection was made at 21 weeks of age, and piglets
were allocated to their original housing system, i.e. either
2.3. Extraction of muscle protein
barren (half concrete area, half metal slats, no straw,
0.8 m2/animal, 10 animals per pen) or enriched pens (daily
Pieces of 1 g of frozen muscle tissue were cut and
fresh straw on concrete lying area, 0.7 m2/animal indoor
weighed, and were immediately homogenized (Ultra
plus 1.13 m2/animal outdoor, 8 animals per pen). All ani-
Turrax homogenizer) in 12.5 mL cold pre-rigor extraction
mals were fed ad lib.
buffer of 50 mM Tris–HCL [Tris(hydroxymethyl)amino-
methane–HCl], 10 mM EDTA (Ethylenediaminetetraacetic
2.2. Slaughter procedure and measurements
acid), 18 mM DTT (Dithiothreitol), pH 8.3] (Note: DTT
was weighed and added immediately before extraction).
Animals were slaughtered at 25 weeks of age, and were
The homogenate was divided into two equal parts for anal-
deprived from food at the day before slaughter. Two repli-
ysis of ‘‘soluble’’ and ‘‘total’’ protein, respectively. For
cates (group 1 and 2) were slaughtered on two different
‘‘soluble’’ protein, the sample was centrifuged (3500 rpm,
days, at a normal commercial slaughterhouse. After trans-
+4 C, 15 min), the supernatant aliquotted and stored at
port (1 h) to the slaughterhouse, pigs were kept in lairage
20 C, and 100 ll used for protein purification (PlusOne
for 2 h. Saliva samples were collected before transport
2D-Clean-up kit, Amersham Biosciences). For ‘‘total’’ pro-
and at the end of the lairage period, for the purpose of cor-
tein, 300 ll homogenate was mixed with 1.2 mL of 10%
tisol determination. Pigs were stunned electrically (300 V,
SDS (Sodium dodecyl sulfate) and heated at 100 C for
3 s) with a pair of stunning tongs, after which they were
5 min. After centrifugation (Eppendorf centrifuge,
shackled by the hind leg and exsanguinated. Carcasses were
13,000 rpm, 10 min, rT), 100 ll of supernatant was used
chilled overnight at +4 oC. Muscle sampling and measure-
for protein purification (PlusOne 2D-Clean-up kit, Amer-
ments were performed on 48 pigs. Temperature (Ebro TFN
sham Biosciences).
1093 SK thermometer) and pH (Schott CG 818 pH meter
Losses of protein due to purification by Clean-up kit
with Xerolyt pH electrode) of the Longissimus dorsi (LD)
appeared to vary between 40% and 60%.
and Biceps Femoris (BF) muscle were measured at 5 min,
Precipitated proteins were finally dissolved in rehydra-
45 min, 4 h and 24 h postmortem. Carcass weight was reg-
tion buffer of 8 M urea, 2% CHAPS [3-[(3-Cholamidopro-
istered at 45 min postmortem. Carcass meat percentage
pyl)-dimethyl-ammonio]-1-propane sulfonate], 0.5% IPG
and backfat thickness were determined with a Hennesy
buffer (ampholite-containing buffer for use with Immobi-
Grading Probe II at 45 min after slaughter.
line DryStrip gels) w/v, 18 mM DTT.
Five minutes after exsanguination, samples were taken
After purification, samples were analysed for protein
with a cork bore from the LD muscle at the height of the
concentration by RC/DC Protein Assay Kit (BioRad).
3rd to 4th lumbar vertebra and from the BF muscle. Mus-
Concentrations of purified protein in the soluble
cle cubes (0.5 cm3) were cut from the LD muscle sample
fractions of the low drip and the high drip composite
and the inside extreme end of the BF muscle cork bore
samples were 690 and 484 lg per 450 lL, respectively.
sample. The BF sample represents the visually determined
However, the samples have to be analysed in triplicate;
red portion of the BF muscle, whereas the LD sample rep-
for this reason and because of limitation of the volumes
resents a typically white muscle tissue. For BF it was con-
available, the amounts of protein per gel were 281 and
sidered that this muscle is especially active during walking,
127 lg for the low drip and high drip group, respectively.
which makes it of interest with respect to the difference
With respect to the total protein fractions, the available
between the barren and enriched housing system.
amounts were large enough for highly sensitive silver-
All samples were rapidly frozen in solid carbon dioxide,
staining (Amersham), but not for MS-compatible silver-
and stored at 80 C. From each cork bore sample, addi-
staining and subsequent MS analysis. In the underlying
tional muscle cubes (0.5 cm3) were cut and rapidly frozen
study, only soluble protein fractions were used for further
in isopentane cooled with liquid nitrogen and stored at
detailed investigation.
80 C, for the purpose of myofiber analysis and measure-
For comparison of ‘‘High Drip’’ and ‘‘Low Drip’’ pro-
ment of capillary density.
teomes, purified samples of four selected animals with
At 24 h postmortem some meat quality characteristics
either high or low % drip loss were combined in such a
were determined on both LD and BF muscle samples.
way, that all animals contributed equal amounts of protein
Water-holding capacity was measured according to the fil-
to the high drip or the low drip mixture, respectively.
ter paper method of Kauffman et al. (Kauffman, Eikelen-
boom, Van der Wal, Merkus, & Zaar, 1986). Drip loss
was determined as percentage of weight loss after 2 and 5 2.4. Two-dimensional electrophoresis
days storage at 4 C according to Lundström and Malmf-
ors (Lundström & Malmfors, 1985). Meat color was deter- Samples were analysed in triplicate.
 D.F.M. van de Wiel, W.L. Zhang / Meat Science 77 (2007) 46–54
First dimension separation of proteins on the basis of
differences in isoelectric point (pI) was performed with
Immobiline DryStrips (24 cm, pH 3–10, Amersham Phar-
macia Biotech) and an IPGphor Isoelectric Focusing Sys-
tem, according to the instructions of the supplier. After
isoelectric focusing and before 2nd dimension, the strips
were equilibrated (15 min, rT) in 14 mL of SDS equilibra-
tion buffer [50 mM Tris–HCl pH 8.8, 6 M urea, 30% glyc-
erol, 2% SDS and a few grains of BFB (Bromophenol
Blue)] plus 140 mg DTT added immediately before use, fol-
lowed by iodoacetamide equilibration (14 mL of SDS
equilibrationbuffer plus 350 mg iodoacetamide added
immediately before use).
Second dimension electrophoresis [MW separation by
PAGE (Polyacrylamide gel electrophoresis)] was per-
formed with an Ettan Dalt II System, and the use of Pre-
cast gels (12.5% polyacrylamide, 255 · 196 · 1 mm,
Amersham Biosciences).
Proteinstaining for spot detection was done with the
highly sensitive Silver Staining Kit of Amersham Biosci-
ences. For preparative purposes, i.e. excision of spots and
protein analysis by mass spectrometry, gels were stained
with the MS-compatible SilverQuest Silver Staining Kit
of Invitrogen Life Technologies, according to a modified
protocol.
Image analysis of 2D gels was performed with the
PDQuest programme of BioRad.
Molecular weights were estimated by application of
marker proteins (Low Molecular Weight SDS Calibration
Kit for SDS electrophoresis, Amersham Pharmacia Bio-
tech), and pI was estimated by assuming linear distribution
of pI over the whole range of the Immobiline DryStrip pH
3–10. Randomly chosen proteins A B, C and D served as
internal controls.
After 2D-elecrophoresis, which was followed by highly
sensitive silverstaining (Amersham) and PDQuest analysis,
approximately 700 spots per gel could be detected and were
used for the alignment.
Statistical analysis was done by student t-test, as pro-
vided by the PDQuest software.
For the purpose of mass-spectrometry analysis, the
experiment was repeated with different aliquots of the same
composite samples but with maximal (i.e. three times more
as compared to the Amersham method) protein concentra-
tions, and proteins were stained with MS-compatible silver
staining method (Invitrogen) After MS-compatible stain-
ing, each spot of a candidate marker protein was identified
on the basis of previous PDQuest analysis, excised by plac-
ing a shortened pipet tip over it, transferred into an Eppen-
dorff tube containing 100 lL of 10% methanol/1% acetic
acid, and stored at 80 C until analysis.
2.5. Chemicals and reagents; digestion of proteins from two-
dimensional gels
Chemicals and reagents were of highest purity grade,
preferably PlusOne-2D (Amersham Biosciences). Trypsin
49
digestion and MS/MS analysis were essentially as described
by Li et al. (2004).
The gel pieces were destained in 60% acetonitrile in
25 mM ammonium bicarbonate buffer, pH 8.5, and then
dehydrated with 100% acetonitrile. The shrunken gel pieces
were reswelled in 25 mM ammonium bicarbonate buffer,
dehydrated again in 100% acetonitrile, and dried in a
speedvac. For gel pieces that were heavily stained the rehy-
dration/dehydration step was repeated once. The gel pieces
were rehydrated in 8 ll of trypsin solution (20 lg/ml) for
1 h, followed by addition of 50 ll of 25 mM ammonium
bicarbonate buffer to completely immerse the gel pieces.
After incubation overnight at room temperature, 0.5 ll of
the incubation buffer was pipetted to the MALDI plate
and mixed with 1 ll of a-cyano-4-hydroxycinnamic acid.
The samples were analyzed by a MALDI TOF/TOF
(Applied Biosystems) mass spectrometer (see below). In
cases where the mass spectrometric signals were weak,
the incubation buffer was loaded into a C18 Ziptip (Milli-
pore) according to the protocol provided by the company.
The bound peptides were eluted from the Ziptip using
1.0 ll of a-cyano-4-hydroxycinnamic acid, which was
directly deposited onto the MALDI plate. The a-cyano-
4-hydroxycinnamic acid matrix concentration was 5 mg/
ml in 50% acetonitrile/50% water containing 0.1% trifluo-
roacetic acid.
2.6. Mass spectrometry of trypsin-digested protein spots
from two-dimensional gels
The mass spectrometer utilized for the protein analysis
was an Applied Biosystems 4700 Proteomics Analyzer with
TOF/TOF Optics. This MALDI tandem mass spectrom-
eter uses a 200 Hz frequency-tripled neodinium YAG laser
operating at a wavelength of 355 nm. For MS/MS, ions
generated by the MALDI process were accelerated at
8 kV through a grid at 6.7 kV into a short, linear, field-free
drift region. In this region the ions passed through a timed-
ion selector (TIS) device that is able to select one peptide
from a mixture of peptides at different m/z for subsequent
fragmentation in the collision cell. After a peptide at a
given m/z was selected by the TIS it passed through a
retarding lens where the ions were decelerated and then
passed into the collision cell, which was operated at 7 kV.
The collision energy was defined by the potential difference
between the source and the collision cell and hence was
1 kV. Inside the collision cell the selected peptide ions col-
lided with air at a pressure of 1 · 10 6 Torr. After passing
through the collision cell the ions (both intact peptide ion,
the precursor, and fragments caused by collision with the
air, the product ions) were reaccelerated in the second
source region at 15 kV, passed through a second, field-free,
linear drift region, into the reflector and finally to the detec-
tor. The detector amplified and converted the signal to
electrical current, which was observed and manipulated
on a PC-based operating system. For reflector mode the
50 D.F.M. van de Wiel, W.L. Zhang / Meat Science 77 (2007) 46–54
operation of the instrument is far simpler. After the
MALDI process generates the peptide ions they are
accelerated at 20 kV through a grid at 14 kV into the first,
short, linear, field-free drift region. After this point the rest
of the instrument can be treated as a continuation of this
field-free, drift region until the ions enter the reflector
and then reach the detector where, as before, the signal
at the detector is amplified and converted to electrical
current.
2.7. Search criteria
The MS resolution for the peptides was generally greater
than 10.000 full-width half-maximum, and the mass accu-
racy better than 10 ppm. For the data base search the mass
tolerance was set at 0.03 Da. The MS/MS resolution was
3000–6000 full-width half-maximum. No internal stan-
dard was used for calibration and the mass tolerance was
set higher at 0.3 Da although mass error was expected to
be less than 0.1 Da. All the mass spectra were searched
against the NCBI data base, using online Mascot software
(Matrix Science), to identify the proteins.
The identification of the proteins is based on several cri-
teria. The most important criterion is the mascot score of
the peptide mass fingerprint (PMF), and in most cases
the scores are significant. Further, the mascot scores of
the daughter ion spectra of the tryptic peptides were also
considered, as well as the matching of the identified pro-
teins to the calculated Mr and pI.
3. Results
3.1. Selection of samples for 2D-analysis
Drip loss was considered one of the most important
meat quality parameters of pork. For this reason, animals
were ranked according to increasing drip loss both for BF
Fig. 1. Ranking of 24 animals according to % drip loss of BF (Biceps
Femoris) muscle. Animals 2, 9, 10 and 15 were selected for high % drip
loss, and animals 1, 3, 8 and 22 were selected for low % drip loss. (drip
5 = % drip loss measured at day 5 postmortem). z = sow, b = barrow,
s = enriched, g = barren.
Fig. 2. Ranking of 24 animals according to % drip loss of LD (longissimus
dorsi) muscle. Numbers and indices as in Fig. 1. The same animals were
selected both for LD and BF (Fig.1).
(Fig. 1) and LD (Fig. 2), and 4 animals were selected either
with high or with low % drip loss in both muscle types.
Selection was made in such a way, that animals were
equally divided over sex and housing system, and no litter-
mates were allocated to the same group (with one excep-
tion: animals #9 and #15 were from the same litter).
It is well known that extreme values of ultimate pH (at
24 h postmortem) can result in abnormal meat quality
characteristics. For this reason, selected animals were
checked for abnormal ultimate pH in both BF and LD.
It appeared that the ranges of ultimate pH values for BF
in the low drip and high drip group were 5.65–5.87 and
5.53–5.73, respectively; for LD the ranges of ultimate pH
in the two groups were 5.49–5.65 and 5.43–5.51, respec-
tively. These values were considered to be normal.
3.2. Candidate marker proteins
Comparison of the high drip and low drip groups
showed that 8 spots were consistently different (Fig. 3).
It appeared that only 3 out of selected 8 spots con-
tained enough protein to allow for identification by MS/
MS analysis. Clear identification by MS was observed
for candidates 2, 7 and 4. Candidate 2 was found to be
creatine kinase M-type; calculation of MW and pI further
confirmed its identity. For candidate 7, confirmation of its
identity as desmin was obtained from data on MW and
pI. However for candidate 4 (SWN/SNF-related matrix-
associated actin-dependent regulator of chromatin a 1 iso-
form b) the calculated pI or MW were different from the
theoretical values as obtained for either rat or human
(Table 1). Nevertheless it was found that candidate 4 is
a protein that is partly homologous to human SNF2L1
transcription activator. The SNF2L1 transcription activa-
tor is functionally linked to ATP (Adenosine triphos-
phate)-binding and actin-dependency, and therefore it
can be hypothetically linked in some (as yet unknown)
way to postmortem intra muscular pH and waterholding
capacity.
 D.F.M. van de Wiel, W.L. Zhang / Meat Science 77 (2007) 46–54 51

Fig. 3. Candidate marker proteins (no’s 1–8) in the Matchset according to PDQuest analysis. Candidates 1–6 are elevated in the high drip sample, and
candidates 7–8 are elevated in the low drip sample. A, B, C and D are randomly chosen spots ranging in density from very strong to very weak, in order to
serve as internal controls. MW = Molecular weight, IEP = Isoelectric point.

Table 1
Protein identification by MS/MS, MW and pI. Sequence coverage (SC, %) and number of matching peptides are also given
cand/spotnr MS/MS MALDI-ToF MW MW pI pI NCBI gi no. SC Number of Reference
mass fingerprint estimated theoretical estimated theoretical (%) matching species
peptides
cand2/1303 Creatine kinase 43,652 43,059 7.20 6.61 gi/17979615 74 9 Porcine
cand7/7607 Desmin 58,884 53,497 5.16 5.21 gi/2959454 29 10 Porcine
cand4/1011 *SWI/SNF-related matrix- 26,915 123,514 7.30 8.67 gi/27709746 16 17 Human
associated actin-dependent rat
regulator of chromatin a1,
isoform a
*id. isoform b SNF2-like 1;

Global transcription activator 26,915 120,231 7.30 7.79 gi/21071046 16 17 Human

homologous sequence

4. Discussion and conclusions
4.1. Candidate marker proteins
Although the observation of creatine kinase as a candi-
date marker protein for waterholding capacity still needs
confirmation, it certainly makes sense in the total picture
of muscle physiology and its relation to meat quality. Cre-
atine kinase is known for enzymatic conversion of creatine
phosphate into creatine and ATP (Fig. 4). The conversion
of muscle to meat is an energy-demanding process and in
the muscle after death, as well as in life, the energy is pro-
vided by splitting of ATP to ADP and inorganic phospho-
rus (Henckel, Karlsson, Jensen, Oksbjerg, & Petersen,
2002). In the muscle after death, the ATP is replenished
by the conversion of ADP to ATP by the transfer of the
higher energy phosphate for creatine phosphate and by deg-
radation of glycogen. The observed decline in pH depends
on this ability by the formation of lactate from the available
glycogen (Bendall, 1973).The basic biochemical reactions
underlying the pH decline postmortem and how this decline
exerts a strong influence on a number of important meat
52 D.F.M. van de Wiel, W.L. Zhang / Meat Science 77 (2007) 46–54
Fig. 4. The reaction is strongly endergonic as written. However, the level of ATP is very high in mitochondria, so the reaction proceeds to the left. Creatine
phosphate then diffuses from mitochondria to the myofibrils, where it provides the energy for muscle contraction.
quality characteristics are well recognised (Bendall, 1973;
Bendall, 1951). In short, creatine phosphate (CP) is firstly
degraded. When this is reduced to approximately 25% of
its resting value, a decrease in glycogen and a concomitant
decrease in ATP are observed. The period, during which CP
is still present in sufficient amounts, is referred to as the
delay phase by analogy to the onset of rigor (Bendall,
1973). The decline in pH thus depends on the initial concen-
tration of creatine phosphate and glycogen (Bendall, 1951),
and in practical situations large variations in the course of
pH decline can be expected (Henckel et al., 2002), which
may relate to variations in drip loss and meat quality.
The enzyme creatine phospho kinase (CPK), which is
responsible for conversion of creatine phosphate into crea-
tine and ATP, is located in the myofibrils particularly in the
M-line (for a review: see Schreurs, 1999). The M-line is the
place in the sarcomere where thick or myosin filaments are
kept in lateral register both transversely and longitudinally.
Myosin together with actin belongs to the socalled contrac-
tile proteins, which are responsible for the mechanical con-
traction of the muscle. Therefore CPK may have an effect
on the rate of muscle contraction postmortem, by its effect
on CP degradation, rate of decline of glycogen and con-
comitant change of pH. Our observation that CPK levels
are higher in muscle with high drip loss may very well fit
in our drip loss hypothesis, in a way that high CPK levels
cause shortening of the delay phase by rapidly degrading
creatine phosphate. This may in turn cause more rapid
pH decline and muscle contraction, and therefore result
in high drip loss.
Another potential marker protein as identified by our
study is desmin. Desmin belongs to the group of cytoskel-
etal proteins, which are responsible for integrity and rigid-
ity of the muscle cell. Desmin is located in the intermediate
filaments; it is responsible for the lateral structure and
integrity of the myofiber because it holds adjacent myofi-
brils together at the Z-line level. It is generally accepted
that the source of drip from pork is intracellular water
which is lost from the muscle fibre postmortem, driven
by a pH and calcium-induced shrinkage of myofibrils dur-
ing rigor development (Honikel, Kim, & Hamm, 1986;
Offer et al., 1989). The rate and quantity of drip formation
in fresh meat is believed to be influenced by the extent of
rigor-shrinkage and the permeability of the cell membrane
to water as well as other factors, such as the extent of pro-
tein denaturation. It has also been recognised that an intact
cytoskeleton within the muscle is necessary to translate
shrinkage of myofibrils into shrinkage of the whole cell
(Offer & Knight, 1989). Adjacent myofibrils are connected
by desmin-rich intermediate filaments and vinculin-rich
sarcomeric structures connect peripheral myofibrils to the
sarcolemma (Stromer, 1998). It is hypothesised that a lat-
eral shrinkage of laterally connected myofibrils results in
a shrinkage of the whole muscle fibre and thereby squeezes
out water (Offer & Knight, 1989). The water accumulates
extra-cellularly between muscle fibre bundles and at a later
stage also between single muscle fibres. From these extra-
cellular compartments water slowly drains to the surface
where it forms drip (Offer & Cousins, 1992).
Proteolytic degradation of desmin in rapidly aging
muscles has been demonstrated during the period 24–72
hrs postmortem (Christensen, Henckel, & Purslow,
2004). The rate of desmin degradation may therefore be
related to the duration and degree of myofibril shrinkage
and therefore to the phenomenon of drip loss. On the
other hand, in a recent 2DE-study on postmortem prote-
olysis no relationship was found between desmin and
driploss (Hwang et al., 2005). This may be partly
explained by the fact that protein identification for aged
muscle (i.e. at 3d) proved extremely complicated, whereas
in our study we analysed samples that were taken imme-
diately after slaughter. Desmin clearly fits in the currently
accepted model for muscle contraction and WHC (Kris-
tensen & Purslow, 2001) and therefore is very likely to
be a marker protein for drip loss. Our observation that
desmin levels are higher in muscle with low drip loss still
has to be reconciled with our mechanistic model for drip
loss.
4.2. Possible practical applications
For a slaughterhouse it is important to have information
at a very early stage, i.e. immediately after slaughter, about
the expected quality of the meat. For this purpose it might
be feasible to develop a protein microarray, that can quan-
tify individual marker proteins or give discriminatory bind-
ing patterns of combined marker proteins. Such a protein
microarray may have the potential advantage over DNA
microarrays, that its measurements are much closer to
the factor of interest, i.e. meat quality.
Development of dipstick tests for rapid on-line testing
could be another option.
 D.F.M. van de Wiel, W.L. Zhang / Meat Science 77 (2007) 46–54
Furthermore, identification of marker proteins for drip
loss can fortify hypothetical models on the mechanism of
drip loss, and can possibly result in preventive measures.
It can also prompt further studies into their corresponding
genes, because the possibility of polymorphisms may allow
for genetic selection against drip loss. This can certainly be
of value for commercial breeding companies, who seek to
improve meat quality characteristics of their breeding
stock.
5. Conclusion
It is concluded that proteomics analysis can be a suc-
cessful method to identify marker proteins for prediction
of meat quality. In our study, candidate marker proteins
have been identified that are highly relevant to the process
of driploss. Continued research almost certainly will lead to
additional marker proteins being identified.
Acknowledgements
The authors wish to thank Mr. Roel van der Schors and
Dr. Ka Wan Li, Department of Molecular and Cellular
Neurobiology, Research Institute Neurosciences, Faculty
of Earth and Life Sciences, Vrije Universiteit, Amsterdam,
for MS/MS analysis. They also wish to thank Dr. Jelle Thole
and Mrs. José Westerveen of Animal Sciences Group of
Wageningen UR for their guidance in 2D-electrophoresis.
One of the authors (Z.W.) gratefully acknowledges a
LNV-IAC fellowship from the Dutch Ministry of Agricul-
ture, Nature and Foodquality.
References
Bandman, E., & Zdanis, D. (1988). An immunological method to assess
protein-degradation in post-mortem muscle. Meat Science, 22, 1–19.
Bendall, J. R. (1951). The shortening of rabbit muscles during rigor
mortis: its relation to the break down of adenosine triphosphate and
creatine phosphate and to muscular contraction. Journal of Physiology,
114, 71–88.
Bendall, J. R. (1973). Post mortem changes in muscle. In G. H. Bourne
(Ed.). Structure and function of muscle (Vol. 3, pp. 227–274). New
York: Academic Press.
Christensen, M., Henckel, P., & Purslow, P. P. (2004). Effect of muscle
type on the rate of post-mortem proteolysis in pigs. Meat Science, 66,
595–601.
Fritz, J. D., & Greaser, M. L. (1991). Changes in titin and nebulin in post-
mortem bovine muscle revealed by gel-electrophoresis, western blot-
ting and immunofluorescence microscopy. Journal of Food Science, 56,
607–610.
Henckel, P., Karlsson, A., Jensen, M. T., Oksbjerg, N., & Petersen, J. S.
(2002). Metabolic conditions in porcine longissimus muscle immedi-
ately pre-slaughter and its influence on peri- and post mortem energy
metabolism. Meat Science, 62, 145–155.
Henckel, P., Karlsson, A., Oksbjerg, N., & Petersen, J. S. (2000). Control
of postmortem pH decrease in pig muscles: experimental design and
testing of animal models. Meat Science, 55, 131–138.
Honikel, K. O., Kim, C. J., & Hamm, R. (1986). Sarcomere shortening of
prerigor muscle and its influence on drip loss. Meat Science, 16,
267–282.
53
Hopkins, D. L., & Thompson, J. N. (2002a). Factors contributing to
proteolysis and disruption of myofibrillar proteins and the impact on
tenderisation in beef and sheep meat. Australian Journal of Agricultural
Research, 53, 149–166.
Hopkins, D. L., & Thompson, J. N. (2002b). The degradation of
myofibrillar proteins in beef and lamb using denaturing electrophore-
sis-An overview. Journal of Muscle Foods, 13, 81–102.
Huff-Lonergan, E., Mitsuhashi, T., Beekman, D. D., Parrish, F. C. J.,
Olson, D. G., & Robson, R. M. (1996). Proteolysis of specific muscle
structural proteins by mu-calpain at low pH and temperature is similar
to degradation in post-mortem bovine muscle. Journal of Animal
Science, 74, 993–1008.
Huff-Lonergan, E., Parrish, F. C., & Robson, R. M. (1995). Effects of
post-mortem aging time, animal age, and sex on degradation of titin
and nebulin in bovine longissimus muscle. Journal of Animal Science,
73, 1064–1073.
Hwang, I. H., Park, B. Y., Kim, J. H., Cho, S. H., & Lee, J. M. (2005).
Assessment of postmortem proteolysis by gel-based proteome analysis
and its relationship to meat quality traits in pig longissimus. Meat
Science, 69, 79–91.
Kauffman, R. G., Eikelenboom, G., Van der Wal, P. G., Merkus, G., &
Zaar, M. (1986). The use of filter paper to estimate driploss of porcine
musculature. Meat Science, 18, 191–200.
Kim, K. S., Larsen, N., Short, T., Plastow, G., & Rothschild, M. F.
(2000). A missense variant of the porcine melanocortin-4 receptor
(MC4R) gene is associated with fatness, growth, and feed intake traits.
Mammalian Genome, 11, 131–135.
Klont, R. E., Hulsegge, B., Hoving-Bolink, A. H., Gerritzen, M. A., Kurt,
E., Winkelman-Goedhart, H. A., et al. (2001). Relationship between
behavioral and meat quality characteristics of pigs raised under barren
and enriched housing conditions. Journal of Animal Science, 79,
2835–2843.
Koohmaraie, M. (1996). Biochemical factors regulating the toughening
and tenderisation processes of meat. Meat Science, 43, 193–201.
Kristensen, K., & Purslow, P. P. (2001). The effect of ageing on the water-
holding capacity of pork: role of cytoskeletal proteins. Meat Science,
58, 17–23.
Lametsch, R., Karlsson, A., Rosenvold, K., Andersen, H. J., Roepstorff,
P., & Bendixen, E. (2003). Postmortem proteome changes of porcine
muscle related to tenderness. Journal of Agricultural and Food
Chemistry, 51, 6992–6997.
Li, K. W., Hornshaw, M. P., Van der Schors, R. C., Watson, R., Tate, S.,
Casetta, B., et al. (2004). Proteomics analysis of rat brain postsynaptic
density. Journal of Biological Chemistry, 279, 987–1002.
Lundström, K., & Malmfors, G. (1985). Variation in light scattering and
water-holding capacity along the porcine longissimus dorsi muscle.
Meat Science, 15, 203–214.
Milan, D., Jeon, J.-T., Looft, C., Amarger, V., Robic, A., Thelander, M.,
et al. (2000). A mutation in PRKAG3 associated with excess glycogen
content in pig skeletal muscle. Science, 288, 1248–1251.
Offer, G., & Cousins, T. (1992). The mechanism of drip production:
formation of two compartments of extracellular space in muscle
post-mortem. Journal of the Science of Food and Agriculture, 58,
107–116.
Offer, G., & Knight, P. (1989). The structural basis of water-holding in
meat. Part 2. Drip losses. In R. Lawrie (Ed.). Developments in meat
science (Vol. 4, pp. 173–243). London: Elsevier Applied Science.
Offer, G., Knight, P., Jeacocke, R., Almond, R., Cousins, T., Elsey, J.,
et al. (1989). The structural basis of the water-holding appearance and
toughness of meat and meat products. Food Microstructure, 8,
151–170.
Purslow, P., Morrison, H., Mielche, M. (1997). Are there intermediate
filament proteins involved in determining meat quality? In Proceedings
43rd international congress of meat science and technology, (p. 598)
Auckland, New Zealand, G1-25.
Schreurs, F. J. G. (1999). Post-mortem changes in chicken muscle. Some
key biochemical processes involved in the conversion of muscle to meat.
PhD thesis. Agricultural University of Wageningen.
54 D.F.M. van de Wiel, W.L. Zhang / Meat Science 77 (2007) 46–54

Sosnicki, A. A., Pommier, S., Klont, R. A., Newman, S., Plastow, G.
(2003). Best-cost production of high quality pork: bridging the gap
between pig genetics, muscle biology/meat science and consumer
trends. In Proceedings Manitoba Pork Seminar, January 9–31.
Steen, D., Claeys, E., Uytterhaegen, L., DeSmet, D., & Demeyer, D.
(1997). Early post-mortem conditions and the calpain/calpastatin
system in relation to tenderness of double-muscled beef. Meat Science,
45, 307–319.
Stromer, M. H. (1998). The cytoskeleton in skeletal, cardiac and smooth
muscle cells. Histology and Histopathology, 13, 283–291.
Taylor, R. G., Geesink, G. H., Thompson, V. E., Koohmaraie, M., &
Goll, D. E. (1995). Is Z-disk degradation responsible for post-mortem
tenderisation? Journal of Animal Science, 73, 1351–1367.
Wheeler, T. L., & Koohmaraie, M. (1999). The extent of proteolysis is
independent of sarcomere length in lamb longissimus and psoas major.
Journal of Animal Science, 77, 2444–2451.
 MEAT
SCIENCE
Meat Science 77 (2007) 55–62
www.elsevier.com/locate/meatsci

Microbial ecosystems of traditional fermented meat products:

The importance of indigenous starters
R. Talon *, S. Leroy, I. Lebert
Institut National de la Recherche Agronomique (INRA), Centre de Clermont-Ferrand Theix, Unité de Recherche Microbiologie,
63122 Saint-Genès Champanelle, France

Received 8 March 2007; received in revised form 20 April 2007; accepted 20 April 2007

Abstract

This paper reviews the diversity of microbiota, both in the environment and in traditional fermented European sausages. The envi-
ronments of processing units were colonised at variable levels by resident spoilage and technological microbiota, with sporadic contam-
ination by pathogenic microbiota. Several critical points were identified such as the machines, the tables and the knives – knowledge
crucial for the improvement of cleaning and disinfecting practices. Traditionally fermented sausages generally did not present a sanitary
risk. The great diversity of lactic acid bacteria and staphylococci was linked to manufacturing practices. Development of indigenous
starters is very promising because it enables sausages to be produced with both high sanitary and sensory qualities. Our increasing knowl-
edge of the genomes of technological bacteria will allow a better understanding of their physiology in sausages.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Microbial ecosystem; Traditional fermented sausage; Indigenous starter; Lactic acid bacteria; Coagulase-negative cocci

1. Introduction industrial development has led to the use of starter cultures

In Europe, naturally fermented sausages have a long tra-
dition originating from Mediterranean countries during
Roman times (Comi et al., 2005). Production then spread
to Germany, Hungary and others countries including the
United States, Argentina and Australia (Demeyer, 2004).
Europe is still currently the major producer and consumer
of dry fermented sausages (Talon, Leroy-Sétrin, & Fadda,
2004).
There is a wide variety of dry fermented products on the
European market as a consequence of variations in the raw
materials, formulations and manufacturing processes,
which come from the habits and customs of the different
countries and regions. However, Northern products have
a pH below 5, while Mediterranean products have a pH
of 5.3–6.2 and are highly desiccated. In both categories,
*
Corresponding author. Tel.: +33 (0)4 73 62 41 70; fax: +33 (0)4 73 62
45 81.
E-mail address: talon@clermont.inra.fr (R. Talon).
to standardise and control sausage manufacturing. How-
ever, artisanal slightly fermented sausages form a group
of traditional Mediterranean products with a great regional
diversity, both between and within countries. These tradi-
tional sausages can be defined as a meat product made of
a mixture of meat (often pork), pork fat in variable ratio
and salt including eventually sugar, nitrate and/or nitrite
(Fontana, Cocconcelli, & Vignolo, 2005; Lebert et al.,
2007). The fermentation and ripening/drying do not always
constitute two separated steps and they can be carried out
in a natural dryer depending on local climatic conditions
(Corbière Morot-Bizot, Leroy, & Talon, 2006; Lebert
et al., 2007). Traditional dry sausages rely on natural con-
tamination by environmental microflora. This contamina-
tion occurs during slaughtering and increases during
manufacturing.
The objective of this paper is to review the diversity of
microbiota both in the environment and in traditional fer-
mented sausages. The diversity of the technological micro-
biota (lactic acid bacteria or LAB, and coagulase-negative

0309-1740/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.023
56 R. Talon et al. / Meat Science 77 (2007) 55–62

cocci or CNC) is also highlighted. Finally, the importance
of developing indigenous starters and of increasing knowl-
edge on starters is underlined.
2. Microbiota in the environment of small-scale processing
units
Many authors support the belief that the microorgan-
isms present in traditional sausages are derived from the
raw materials or from the environment of manufacturing
(Mauriello, Casaburi, Blaiotta, & Villani, 2004; Rantsiou,
Urso et al., 2005). This microbiota is usually referred to
as ‘‘house flora’’ (Garcia-Varona, Santos, Jaime, & Rovira,
2000). If the microbiota isolated from traditional sausages
is well described (see paragraph below), the resident micro-
biota in the environment of the processing units is still
poorly known.
We have shown by studying a French small-scale pro-
cessing unit manufacturing sausages that all the process-
ing surfaces and the equipments were colonised by CNC
and yeasts/moulds (Chevallier et al., 2006; Corbière
Morot-Bizot et al., 2006). Their level and particularly
for CNC could reach up to 4.7 log cfu/cm2 (cfu = colony
forming unit). Sporadic cases of contamination was
recorded for Staphylococcus aureus and Listeria monocyt-
ogenes. By extending our research to 10 traditional
French processing units (PUs) during the European pro-
ject Tradisausage (QLK1-CT-2002-02240, http://
www2.clermont.inra.fr/tradisausage), we have shown that
regardless of the microorganisms, cold rooms and mixing
machines had the lowest levels of contamination with
median levels of 2.2 log cfu/100 cm2, while knives and
tables had the highest ones (from 2.1 to 4.4 log cfu/
100 cm2) (Table 1) (Lebert et al., 2007). CNC and
yeasts/moulds but also Pseudomonas colonised all the sur-
faces and equipments of these processing units (Table 1).
Considering the pathogenic bacteria, Salmonella and S.
aureus were not detected in the environment while L.
monocytogenes was enumerated in the table and the knife
of one processing unit (Lebert et al., 2007). In the Euro-
pean project Tradisausage, the residual microbiota was
also assessed in the environment of 43 traditional process-
ing units of Spain, Portugal, Italy, Greece and Slovakia
(Talon et al., 2007). All the PUs’ environments were col-
onised at variable levels by spoilage and technological
microbiota with some PUs having too high contamina-
tion. Salmonella and L. monocytogenes were detected in
4.8% and 7.6% of the samples and S. aureus was enumer-
ated in 6.1%. Several critical points were identified such as
the machines for S. aureus and the tables and the knives
for L. monocytogenes; their knowledge is crucial for the
improvement of the control systems. The variability of
the residual contamination highlighted the different clean-
ing, disinfecting and manufacturing practices of the small-
scale processing units (Talon et al., 2007). In fact,
unclean, insufficiently or inadequately cleaned pieces of
equipment have often been identified as the source of
pathogens (Reij & Den Aantrekker, 2004). Many studies
investigated the pathogen flora of food processing envi-
ronments and food processing lines such as L. monocytog-
enes in pork and poultry processing plants and products
(Chasseignaux et al., 2002) and Salmonella species in pork
slaughter and cutting plants (Giovannacci et al., 2001).
3. Microbiota of products
Fermentation of traditional dry sausages relies on the
indigenous microbiota. We have shown that this microbi-
ota was enumerated in the batters manufactured in French
traditional processing units at average level of 4.0 log cfu/g
(Fig. 1) (Lebert et al., 2007). It was composed of useful
microorganisms (CNC, LAB, yeasts/moulds) for the fer-
mentation and flavour of sausages, but also spoilage micro-
biota (Pseudomonas, enterobacteria) and enterococci. This
indigenous microbiota could arise from the raw materials
or from the environment of manufacturing (microbiota
described in Table 1) as mentioned by several authors.
However, in this study, we found no evidence of cross-con-
tamination between environment and meat. We assumed
that the contamination of the batter was mainly due to
the microbiota of the raw materials and sometimes of the
casings. We have shown that lean meat and fat used to pre-
pare the batter were sometimes contaminated with high
levels of CNC, yeasts/moulds, coliforms and Pseudomonas
(Chevallier et al., 2006). Comi et al. (2005) also showed
that raw meats had total aerobic counts of between 4.3
and 5.8 log cfu/g and casings of between 3.8 and
6.1 log cfu/g.

Table 1
Microbiota of the environments of 10 French processing units manufacturing traditional fermented sausages
Tables Knives Cold room Mincing machines Mixing machines Stuffing machines
Median Min Max Median Min Max Median Min Max Median Min Max Median Min Max Median Min Max
Y/M 4.0 <0.7 6.8 3.7 2.7 6.7 1.1 <0.7 6.7 2.6 1.5 4.7 1.8 <0.7 5.4 2.9 <0.7 6.7
LAB 4.0 <1.7 7.1 3.3 1.7 5.6 <1.7 <1.7 5.0 2.6 <1.7 5.5 <1.7 <1.7 5.8 2.2 <1.7 5.4
CNC 4.4 <1.7 6.9 4.0 <1.7 5.3 0.9 <1.7 4.8 3.4 1.7 5.5 2.3 <1.7 5.3 2.7 <1.7 5.7
ENC 2.2 <0.7 3.1 2.6 <0.7 5.0 0.2 <0.7 4.7 2.2 <0.7 3.4 <0.7 <0.7 4.0 1.7 <0.7 3.9
ENB 3.2 <0.7 6.4 2.7 <0.7 6.0 < 0.7 <0.7 1.7 <0.7 <0.7 3.1 <0.7 <0.7 2.1 1.7 <0.7 6.1
PSE 4.3 <0.7 7.1 4.2 2.7 7.3 1.5 <0.7 3.9 2.2 <0.7 5.3 1.5 <0.7 3.5 2.7 <0.7 7.1
Data expressed in log cfu/100 cm2. Y/M, yeasts and moulds; LAB, lactic acid bacteria; CNC, coagulase-negative cocci; ENC, enterococci; ENB,
enterobacteria; PSE, Pseudomonas.
 R. Talon et al. / Meat Science 77 (2007) 55–62
8.5 8.5
7.5 7.5
6.5 6.5
log(CFU/g)
5.5 5.5
4.5 4.5
3.5 3.5
2.5 2.5
1.5 1.5
Z M F Z
Fig. 1. Evolution of the microbiota during the processing of 10 traditional French fermented sausages: batter (Z), after fermentation (M) and final product
(F). (a) (— j —) Coagulase-negative cocci; (- - r - -) lactic acid bacteria. (b) (— j —) Pseudomonas; (- - r - -) Enterobacteriaceae. (c) (— j —) Enterococcus;
(- - r - -), yeasts and moulds. Data were calculated from the average of the 10 processing units (log cfu/g). Vertical lines: standard deviation.
The type of microbiota that develops in traditional sau-
sages is related to the diversity in formulation, and to the
fermentation and ripening practices. These practises could
be very different in terms of temperature, duration and rel-
ative humidity (Lebert, Leroy, & Talon, 2007). In Europe,
fermentation can be carried out at high temperature (18–
24 °C) between 1 and 2 days (Cocolin, Manzano, Cantoni,
& Comi, 2001; Comi et al., 2005; Mauriello et al., 2004) or
at low temperature (10–12 °C) during 1 week (Lebert et al.,
2007; Mauriello et al., 2004). Similarly, temperatures of
drying/ripening ranged mainly between 10 and 14 °C for
French (Lebert et al., 2007), Italian (Cocolin et al., 2001;
Cocolin, Manzano, Aggio, Cantoni, & Comi, 2001; Comi
et al., 2005; Coppola, Mauriello, Aponte, Moschetti, & Vil-
lani, 2000; Rantsiou, Drosinos et al., 2005) and Greek sau-
sages (Papamanoli, Tzanetakis, Litopoulou-Tzanetaki, &
Kotzekidou, 2003). The duration of this step varied from
4 to 12 weeks. The diversity in the relative humidity leads
to variable water content of traditional sausages at the
end of drying ranging from 0.83 to 0.93 in French, Spanish,
Portuguese and Italian sausages (Lebert, Leroy et al.,
2007).
Despite the different practises within the countries, or
the regions, the microbial populations showed similar evo-
lutions as those we have shown for French sausages
(Fig. 1). In traditional sausages, LAB constituted the major
microbiota at the end of the ripening stage (Fig. 1a). Even
if their initial levels varied their final levels were close to the
one of industrial products manufactured with starter cul-
tures. LAB usually increased the very first days of fermen-
tation and they remained constant during ripening at 7–
9 log cfu/g (Cocolin et al., 2001; Comi et al., 2005) or they
can increased during all the process as shown for French
sausages in Fig. 1a and reached similar final value. The ini-
tial population can be low, as observed in Salame Milano
in Italy (Rebecchi, Crivori, Sarra, & Cocconcelli, 1998)
and in a French sausage manufactured at low temperature
(Chevallier et al., 2006) but it was generally comprised
57
8.5
7.5
6.5
5.5
4.5
3.5
2.5
1.5
M F Z M F
between 3.2 and 5.3 log cfu/g. LAB growth was often cor-
related with the decrease in pH in the first stage of matura-
tion (Cocolin et al., 2001; Cocolin et al., 2001).
CNC constituted the second microbiota at the end of
ripening with a population of 6–8 log cfu/g, population
generally inferior to that of the LAB and closed to the
one reached in industrial sausages (Fig. 1a). Their initial
level varied from 3.1 to 4.4 log cfu/g (Lebert et al., 2007).
CNC sometimes grew during the fermentation period to
106–108 cfu/g or they can grow during ripening (Comi
et al., 2005) or during all the process (Fig. 1a).
Spoilage bacteria such as Pseudomonas and enterobacte-
ria had far different initial levels according to the type of
sausages. They ranged from 1.7 to 4.4 log cfu/g for entero-
bacteria and from 1.5 to 5.2 log cfu/g for Pseudomonas
(Lebert et al., 2007). In French sausages, they remained
constant during the fermentation and decreased during
the ripening (Fig. 1b). In Greek sausages, they were pro-
gressively eliminated regardless of their initial population
(Drosinos et al., 2005; Samelis, Metaxopoulos, Vlassi, &
Pappa, 1998). Other authors found that enterobacteria
and Pseudomonas increased during the fermentation (Comi
et al., 2005). Then enterobacteria remained constant until
the end while Pseudomonas remained constant or disap-
peared (Chevallier et al., 2006; Comi et al., 2005).
Yeasts and moulds were usually detected in all sausages
in the batter at levels varying from 2.0 to 4.5 log cfu/g
(Lebert et al., 2007). Some authors observed growth during
the fermentation period with levels not increasing above
5 log cfu/g (Fig. 1c), then a stability or decrease in the pop-
ulation (Comi et al., 2005; Drosinos et al., 2005). Yeasts
and moulds were not detected in Italian dry sausages Sal-
ame Milano at the end of ripening (Rebecchi et al., 1998).
Enteroccocci had an initial level between 2 and
4 log cfu/g (Fig. 1c). Enterococci usually grew during early
fermentation and remained constant at a level of 4–6 log
until the end of the whole process (Fig. 1c) (Comi et al.,
2005; Drosinos et al., 2005; Rebecchi et al., 1998). In few
58 R. Talon et al. / Meat Science 77 (2007) 55–62
cases their counts declined. Enterococci are poor acidifiers
and in traditional sausages of high pH they find good con-
ditions for survival and growth (Hugas, Garriga, & Ayme-
rich, 2003). There is still controversy over considering them
as Generally Recognised as Safe (GRAS) microorganisms
(Giraffa, 2002). However, studies point out that meat
enterococci, especially Enterococcus faecium have a much
lower pathogenicity potential than clinical strains and
some strains of E. faecium are already used as starter cul-
ture or probiotic (Hugas et al., 2003; Martı́n, Garriga,
Hugas, & Aymerich, 2005).
Considering the pathogenic bacteria, we have shown in
the European project Tradisausage, that Salmonella was
detected in 5.6% of 54 ripened sausages, S. aureus was enu-
merated in 7.4% of the sausages at a level in excess of the
limit of 2.7 log cfu/g (Commission Regulation (EC) No.
2073/2005, 2005) and L. monocytogenes was numerated
at a level in excess of the limit of 2.0 log cfu/g (Commission
Regulation (EC) No. 2073/2005, 2005) in only one sample.
In Salame Milano, S. aureus was observed at the beginning
of the production process, decreased during ripening until
it was undetectable at the end of the process (Rebecchi
et al., 1998). While, it was still enumerated in sausages pro-
duced in Italian artisanal plants from level ranging from 2
to 4 log cfu/g (Blaiotta et al., 2004). L. monocytogenes was
sometimes present in initial samples, but diminished or was
not detected by the end of fermentation in Greek sausages
(Drosinos et al., 2005; Samelis et al., 1998).
4. Diversity of technological microbiota
Lactic acid bacteria and Staphylococcus or Kocuria
belonging to the CNC group are considered as technolog-
ical microbiota because they are involved in the develop-
ment of hygienic and sensory qualities of the final
product. Lactic acid bacteria are involved mainly through
their acidification. Staphylococcus and Kocuria contribute
to the development of colour and flavour in fermented
meat products mainly by degrading free amino acids and
inhibiting the oxidation of unsaturated free fatty acids
(Talon & Leroy, 2006; Talon et al., 2004).
4.1. Lactic acid bacteria (LAB)
Phenotypical methods have been widely used to identify
LAB, however, methods relying on these methods are
ambiguous and time consuming. Molecular methods such
as species-specific PCR (Aymerich et al., 2006), PCR-dena-
turing gel electrophoresis (Rantsiou, Drosinos et al., 2005)
and real time PCR (Furet, Quenee, & Tailliez, 2004) have
been developed.
By using these molecular methods, it has been shown
that the LAB species the most commonly identified in tra-
ditional fermented sausages were Lactobacillus sakei, Lac-
tobacillus curvatus and Lactobacillus plantarum (Lebert,
Leroy et al., 2007). L. sakei is often the dominant one
and can represent more than 42% of the isolates (Comi
et al., 2005; Coppola et al., 2000; Greco, Mazette, De San-
tis, Corona, & Cosseddu, 2005; Papamanoli et al., 2003;
Urso, Comi, & Cocolin, 2006). Aymerich et al. (2006)
showed that L. sakei was identified in all of the Spanish
sausages and represented 89% in chorizo and 76% in a tra-
ditional Spanish sausage ‘‘fuet’’. In a French sausage, L.
sakei represented 100% of the isolates on the final products
although it was minor in the raw materials (Ammor et al.,
2005). L. curvatus is the second species identified; it is dom-
inant in some Greek or Italian sausages (Comi et al., 2005;
Rantsiou, Drosinos et al., 2005). L. plantarum is the third
one; it dominates the LAB flora in a Greek sausage (Dro-
sinos et al., 2005). Many other LAB are identified but rep-
resent a minor population (L. alimentarius, L. casei, L.
delbrueckii, L. farciminis, L. paraplantarum, L. pentosus
and L. sharpeae).
The diversity at strain level inside the dominant species
is important. The combination of results obtained with
plasmid profiling and randomly amplified polymorphic
DNA (RAPD)-PCR, allowed to distinguish 112 different
strains out of 185 isolates of L. sakei and 23 profiles for
53 isolates of L. curvatus (Aymerich et al., 2006). The
analysis by RAPD-PCR of 100 strains of L. curvatus iso-
lated from Greek, Hungarian and Italian naturally fer-
mented sausages revealed that nine profiles were
obtained, while 168 strains of L. sakei from the same sam-
ples gave 19 major clusters (Rantsiou, Drosinos et al.,
2005). Urso et al. (2006) have also used RAPD to deter-
mine the diversity and the distribution of 353 strains of L.
sakei and 67 strains of L. curvatus during a natural fer-
mentation of three Italian sausages. Clusters containing
strains isolated from different plants but also clusters
formed by strains isolated from a specific fermentation
were observed.
4.2. Coagulase-negative cocci (CNC)
The identification of staphylococci to the species level by
phenotypical methods has limitations and may have
resulted in misidentifications (Giammarinaro, Leroy, Cha-
cornac, Delmas, & Talon, 2005). To provide increasingly
reliable identifications, several molecular methods have
been developed, including PCR-based methods such as
PCR-DGGE (Cocolin et al., 2001), species-specific PCR
(Aymerich, Martı́n, Garriga, & Hugas, 2003; Morot-Bizot,
Talon, & Leroy-Sétrin, 2003), multiplex PCR (Corbière
Morot-Bizot et al., 2006) and oligonucleotide array target-
ing sodA gene (Giammarinaro et al., 2005). This last
method developed in our laboratory identifies the 36 vali-
dated species and constitutes presently a powerful tool
for the identification of staphylococci (Giammarinaro
et al., 2005).
Staphylococcus xylosus is the most common species in
Greek, Italian and Spanish traditional products at the
end of ripening (Blaiotta et al., 2004; Cocolin et al., 2001;
Garcia-Varona et al., 2000; Iacumin, Comi, Cantoni, &
Cocolin, 2006a; Martı́n et al., 2006; Mauriello et al.,
 R. Talon et al. / Meat Science 77 (2007) 55–62
2004; Papamanoli et al., 2003; Rebecchi et al., 1998). It rep-
resents from 17% to 100% of the isolates according to the
type of sausages. Even if it was not the dominant species
in the batter, S. xylosus became rapidly dominant in Sal-
ame Milano (Rebecchi et al., 1998). S. saprophyticus is
the second dominant species identified. It is dominant in
some Greek and Italian sausages (Drosinos et al., 2005;
Mauriello et al., 2004; Papamanoli et al., 2003). In the Ital-
ian products, Staphylococcus equorum and Staphylococcus
succinus were also isolated (Mauriello et al., 2004). In three
French small producers, we have shown that the staphylo-
coccal microflora of the product and the environment was
dominated by S. equorum (49%, 56% and 71% of the iso-
lates), the second species was S. succinus for two producers
(33% and 12%) while it was S. xylosus (19%) and S. sap-
rophyticus (19%) for the third producer (Corbière Morot-
Bizot et al., 2006; Leroy, Chevallier, Lebert, Chacornac,
& Talon, 2006). Many other minor species were identified
in the different traditional sausages (S. aureus, S. auricu-
laris, S. carnosus, S. cohnii, S. epidermidis, S. haemolyticus,
S. hominis, S. intermedius, S. lentus, S. pasteuri, S. vitulus
and S. warneri).
The diversity of S. xylosus strains have been character-
ised by different typing methods. The combination of plas-
mid profiling and RAPD-PCR allowed the discrimination
of 169 different profiles out of 194 S. xylosus isolates
(Martı´n et al., 2006). By genotypic clustering based on
comparison of RAPD-PCR profiles (Rossi, Tofalo, Torri-
ani, & Suzzi, 2001) distinguished 22 clusters at 70% of
homology. A total of 249 strains of S. xylosus isolated from
three plants manufacturing sausages were genetically char-
acterised using RAPD, Rep-PCR and Sau-PCR techniques
(Iacumin, Comi, Cantoni, & Cocolin, 2006b). The results
obtained allowed the discrimination of the strains coming
from different plants.
5. Importance of the indigenous starter
The manufacture of traditional sausages is more an art
depending on the skill and experience of the meat manufac-
turer rather than a process fully based on scientific and
technological means. This is because meat fermentation is
a complex biological phenomenon accelerated by the desir-
able action of certain microbes in the presence of a great
variety of competing or synergistically acting species. These
traditional practises lead to a great variability in the quality
of the products. Few sporadic studies conducted on tradi-
tional products have shown that hygienic shortcomings
can lead up to 25% of product loss with high economic
consequences and may undermine consumer confidence
for traditional products.
The development of starters from the natural fermenta-
tive communities of traditional fermented products may
avoid or limit this variability in the production. Moreover,
the development of indigenous starters may diversify the
market that will be able to produce many typical regional
fermented sausages with specific flavours.
59
The development of indigenous starters improving
hygienic quality and keeping the sensorial one is a real
challenge. This challenge was considered in the European
project Tradisausage. We have developed a starter com-
posed of a mixture of the indigenous bacteria isolated from
a traditional sausages manufactured by a French producer,
it consisted of L. sakei, S. equorum and S. succinus (Lebert,
Leroy, Lebecque, Chacornac, & Talon, 2006; Talon, 2006).
This starter was added to the batter and positive results
were observed. Its addition resulted in sausages with higher
sanitary qualities compared to the control: reduction of the
level of L. monocytogenes (1.1 log cfu/g versus 2.7 for the
control) and enterococci (4.2 log cfu/g versus 6.2 for the
control), reduction of 1.7-fold the level of biogenic amines
(233.8 mg/kg DM versus 404.4 mg/kg DM), reduction of
lipid oxidation (0.61 mg MDA/kg versus 0.71 mg MDA/
kg) and total cholesterol oxides (0.95 mg/kg versus
4.75 mg/kg). Finally this starter did not modify the flavour
of the sausages as evaluated by a panel of jury but
improved slightly the texture. Similar experiences have
been carried out in different countries in the EU project
Tradisausage and have lead to close conclusion.
6. Towards genomics and post-genomics to characterise
starters
Up to recently, classical methods based on biochemical
and physiological traits have been used to select the most
performing strains for technological use. They were mainly
based on acidification and antimicrobial properties for lac-
tic acid bacteria and colour and flavour developments for
staphylococci (Talon & Leroy, 2006). During the last dec-
ade, genetic studies have provided basic knowledge on tar-
geted metabolic activities. Now global approaches with the
sequencing of whole genome of bacteria are developed and
will allow a better understanding of their physiology in
meat ecosystem (Champomier-Vergès et al., 2007; Talon
& Leroy, 2006).
Most of the genetic information concerned L. sakei 23K,
initially isolated from a fermented dry sausage (Champo-
mier-Vergès, Chaillou, Cornet, & Zagorec, 2002). Its chro-
mosome map was obtained by the use of 47 genetic loci
(Dudez et al., 2002). In parallel, a proteomic approach
was developed to study the genes involved in adaptation
to its environment (Marceau, Mera, Zagorec, & Champo-
mier-Vergès, 2001). In 2005, the complete genome sequence
of L. sakei 23K was published (Chaillou et al., 2005). From
the known involvement of L. sakei as sausage starter, the
genome analysis confirmed that the main role of L. sakei
is to ferment sugars into lactic acid, and that it lacks main
aroma production pathways. L. sakei 23K also lacks genes
responsible for biogenic amine production. A battery of
functions was identified which might explain adaptation
or resistance of this species to stressing environmental con-
ditions used during sausage manufacturing (presence of
curing agents, spices, smoke, low temperature) (Chaillou
et al., 2005).
60 R. Talon et al. / Meat Science 77 (2007) 55–62
The physical and genetic map of S. carnosus TM300 was
established and the size of the chromosome was estimated
to be 2590 kb (Wagner, Doskar, & Götz, 1998). Eighteen
genetic markers were located within them genes related to
sugar metabolism and to nitrate and nitrite reduction
which are important traits for meat fermentation and col-
our development. Other genes of S. carnosus involved in
technological properties such as the branched-chain amino
acid aminotransferase producing flavour compounds
(Madsen et al., 2002), the superoxide dismutase and the
catalase contributing to the control of lipid oxidation (Bar-
rière, Brückner, & Talon, 2001) have been also identified,
sequenced and characterised. The annotation of the com-
pleted sequence of the chromosome of S. carnosus
TM300 is under way (Rosenstein, Nerz, Resch, & Götz,
2005). The first comparison of the gene products revealed
that about 20% were specific of this species. Their identifi-
cation will reveal the originality of this species.
We have established the physical and genetic map of S.
xylosus C2a by locating 47 restriction fragments and 33
genetic markers (Dordet-Frisoni, Talon, & Leroy, 2007).
Twenty-three previously identified loci mainly concerned
carbohydrate utilization and carbon catabolite repression
(Bassias & Brückner, 1998; Brückner, Wagner, & Götz,
1993; Egeter & Brückner, 1996; Fiegler, Bassias, Jankovic,
& Brückner, 1999) and antioxidant capacities (Barrière
et al., 2001; Barrière, Leroy-Sétrin, & Talon, 2001) and
10 were newly identified (Dordet-Frisoni et al., 2007).
The S. xylosus C2a chromosome contains six rrn operons,
this number varies among strains from 5 to 6 as already
observed in other staphylococcal species. The sequencing
of the complete genome of S. xylosus in collaboration with
the genoscope (http://www.genoscope.cns.fr) is just
achieved. In parallel we have developed a proteomic
approach to study the physiology of S. xylosus C2a strain
in biofilm to allow a better understanding of its survival in
food processing plants (Planchon, 2006).
7. Conclusion
This review outlined the diversity in the microbiota of
naturally fermented sausages but also the strong impact
of the process on its growth. Traditional fermented sau-
sages generally did not present a sanitary risk. The great
diversity in the LAB and staphylococci species and also
within the species was certainly linked to the formulation,
fermentation and ripening practices. The most promising
strains for starter cultures are those which are isolated from
the indigenous microbiota of traditional products. These
strains are well adapted to the meat environment and to
the specific manufacturing process and thus are capable
of dominating the microbiota of products. Above all,
development of indigenous starters leads to sausages with
high sanitary quality and typical sensorial quality. The
exploitation of the data of the genomes of species of tech-
nological interest will offer new research opportunities by
revealing some properties that could explain their adapta-
tion to the meat environment. Global approaches based
on proteomics and transcriptomics are in progress and will
allow a better understanding of their interactions with the
ecosystem and the meat substrate.
Acknowledgement
Parts of the data of this review come from EU Program
QLK1-CT2002-02240.
References
Ammor, S., Rachman, C., Chaillou, S., Prévost, H., Dousset, X., Zagorec,
M., et al. (2005). Phenotypic and genotypic identification of lactic acid
bacteria isolated from a small-scale facility producing traditional dry
sausages. Food Microbiology, 22, 373–382.
Aymerich, T., Martı́n, B., Garriga, M., & Hugas, M. (2003). Microbial
quality and direct PCR identification of lactic acid bacteria and
nonpathogenic staphylococci from artisanal low-acid sausages. Applied
and Environmental Microbiology, 69(8), 4583–4594.
Aymerich, T., Martı́n, B., Garriga, M., Vidal-Carou, M. C., Bover-Cid,
S., & Hugas, M. (2006). Safety properties and molecular strain typing
of lactic acid bacteria from slightly fermented sausages. Journal of
Applied Microbiology, 100, 40–49.
Barrière, C., Brückner, R., & Talon, R. (2001). Characterization of the
single superoxide dismutase of Staphylococcus xylosus. Applied and
Environmental Microbiology, 67(9), 4096–4104.
Barrière, C., Leroy-Sétrin, S., & Talon, R. (2001). Characterization of
catalase and superoxide dismutase in Staphylococcus carnosus 833
strain. Journal of Applied Microbiology, 91(3), 514–519.
Bassias, J., & Brückner, R. (1998). Regulation of lactose utilization genes
in Staphylococcus xylosus. Journal of Bacteriology, 180(9), 2273–2279.
Blaiotta, G., Pennacchia, C., Villani, F., Ricciardi, A., Tofalo, R., &
Parente, E. (2004). Diversity and dynamics of communities of
coagulase-negative staphylococci in traditional fermented sausages.
Journal of Applied Microbiology, 97, 271–284.
Brückner, R., Wagner, E., & Götz, F. (1993). Characterization of a
sucrase gene from Staphylococcus xylosus. Journal of Bacteriology,
175(3), 851–857.
Chaillou, S., Champomier-Vergès, M. C., Cornet, M., Crutz-Le Coq, A.
M., Dudez, A. M., Martı́n, V., et al. (2005). The complete genome
sequence of the meat-borne lactic acid bacterium Lactobacillus sakei
23K. Nature Biotechnology, 23(12), 1527–1533.
Champomier-Vergès, M. C., Chaillou, S., Cornet, M., & Zagorec, M.
(2002). Lactobacillus sakei: Recent developments and future prospects.
Research in Microbiology, 153(2), 115–123.
Champomier-Vergès, M. C., Crutz-Le Coq, A. M., Zagorec, M., Leroy,
S., Dordet-Frisoni, E., Planchon, S., et al. (2007). Genetics of
microbial starters – Chapter 17. In F. Toldra, Y. H. Hui, I. Astiasaran,
W. K. Nip, J. G. Sebranek, E. T. F. Silveira, L. H. Stahnke, & R.
Talon (Eds.), Handbook of fermented meat and poultry (pp. 177–186).
Blackwell publishing.
Chasseignaux, E., Gerault, P., Toquin, M. T., Salvat, G., Colin, P., &
Ermel, G. (2002). Ecology of Listeria monocytogenes in the environ-
ment of raw poultry meat and raw pork meat processing plants. FEMS
Microbiology Letters, 210(2), 271–275.
Chevallier, I., Ammor, S., Laguet, A., Labayle, S., Castanet, V., Dufour,
E., et al. (2006). Microbial ecology of a small-scale facility producing
traditional dry sausage. Food Control, 17(6), 446–453.
Cocolin, L., Manzano, M., Cantoni, C., & Comi, G. (2001). Denaturing
gradient gel electrophoresis analysis of the 16S rRNA gene V1 region
to monitor dynamic changes in the bacterial population during
fermentation of Italian sausages. Applied and Environmental Microbi-
ology, 67(11), 5113–5121.
Cocolin, L., Manzano, M., Aggio, D., Cantoni, C., & Comi, G. (2001). A
novel polymerase chain reaction (PCR) -denaturing gradient gel
 R. Talon et al. / Meat Science 77 (2007) 55–62
electrophoresis (DGGE) for the identification of Micrococcaceae
strains involved in meat fermentations. Its application to naturally
fermented Italian sausages. Meat Science, 58, 59–64.
Comi, G., Urso, R., Iacumin, L., Rantsiou, K., Cattaneo, P., Cantoni, C.,
et al. (2005). Characterisation of naturally fermented sausages pro-
duced in the North East of Italy. Meat Science, 69, 381–392.
Commission Regulation (EC) No. 2073/2005 (2005). Commission Regu-
lation (EC) No. 2073/2005 of 15 November 2005 on microbiological
criteria for foodstuffs. Official Journal of the European Union, L338,
1–26.
Coppola, S., Mauriello, G., Aponte, M., Moschetti, G., & Villani,
F. (2000). Microbial succession during ripening of Naples-type
salami, a southern Italian fermented sausage. Meat Science, 56,
321–329.
Corbière Morot-Bizot, S., Leroy, S., & Talon, R. (2006). Staphylococcal
community of a small unit manufacturing traditional dry fermented
sausages. International Journal of Food Microbiology, 108(2), 210–217.
Demeyer, D. (2004). Meat fermentation: principles and applications –
Chapter 20. In Y. H. Hui, L. Meunier-Goddik, A. S. Hansen, J.
Josephsen, W.-K. Nip, & P. S. Stanfield, et al. (Eds.), Handbook of
food and beverage fermentation technology (pp. 353–368). New York:
Marcel Decker Inc..
Dordet-Frisoni, E., Talon, R., & Leroy, S. (2007). Physical and genetic
map of the Staphylococcus xylosus C2a chromosome. FEMS Micro-
biology Letters, 266(2), 184–193.
Drosinos, E. H., Mataragas, M., Xiraphi, N., Moschonas, G., Gaitis, F.,
& Metaxopoulos, J. (2005). Characterization of the microbial flora
from a traditional Greek fermented sausage. Meat Science, 69,
307–317.
Dudez, A. M., Chaillou, S., Hissler, L., Stentz, R., Champomier-Vergès,
M., Alpert, C. A., et al. (2002). Physical and genetic map of the
Lactobacillus sakei 23K chromosome. Microbiology, 148(Pt. 2),
421–431.
Egeter, O., & Brückner, R. (1996). Catabolite repression mediated by the
catabolite control protein CcpA in Staphylococcus xylosus. Molecular
Microbiology, 21(4), 739–749.
Fiegler, H., Bassias, J., Jankovic, I., & Brückner, R. (1999). Identification
of a gene in Staphylococcus xylosus encoding a novel glucose uptake
protein. Journal of Bacteriology, 181(16), 4929–4936.
Fontana, C., Cocconcelli, P. S., & Vignolo, G. (2005). Monitoring the
bacterial population dynamics during fermentation of artisanal
Argentinean sausages. International Journal of Food Microbiology,
103(2), 131–142.
Furet, J. P., Quenee, P., & Tailliez, P. (2004). Molecular quantification of
lactic acid bacteria in fermented milk products using real-time
quantitative PCR. International Journal of Food Microbiology, 97(2),
197–207.
Garcia-Varona, M., Santos, E. M., Jaime, I., & Rovira, J. (2000).
Characterisation of micrococcaceae isolated from different varieties of
chorizo. International Journal of Food Microbiology, 54(3), 189–195.
Giammarinaro, P., Leroy, S., Chacornac, J.-P., Delmas, J., & Talon, R.
(2005). Development of a new oligonucleotide array to identify
staphylococcal strains at species level. Journal of Clinical Microbiology,
43(8), 3673–3680.
Giovannacci, I., Queguiner, S., Ragimbeau, C., Salvat, G., Vendeuvre, J.
L., Carlier, V., et al. (2001). Tracing of Salmonella spp. in two pork
slaughter and cutting plants using serotyping and macrorestriction
genotyping. Journal of Applied Bacteriology, 90(1), 131–147.
Giraffa, G. (2002). Enterococci from foods. FEMS Microbiology Reviews,
26(2), 163–171.
Greco, M., Mazette, R., De Santis, E. P. L., Corona, A., & Cosseddu, A.
M. (2005). Evolution and identification of lactic acid bacteria isolated
during the ripening of Sardinian sausages. Meat Science, 69, 733–739.
Hugas, M., Garriga, M., & Aymerich, M. T. (2003). Functionality of
enterococci in meat products. International Journal of Food Microbi-
ology, 88(2–3), 223–233.
Iacumin, L., Comi, G., Cantoni, C., & Cocolin, L. (2006a). Ecology and
dynamics of coagulase-negative cocci isolated from naturally fer-
61
mented Italian sausages. Systematic and Applied Microbiology, 29,
480–486.
Iacumin, L., Comi, G., Cantoni, C., & Cocolin, L. (2006b). Molecular and
technological characterisation of Staphylococcus xylosus isolated from
naturally fermented Italian sausages by RAPD, Rep-PCR and Sau-
PCR analysis. Meat Science, 74, 281–288.
Lebert, I., Leroy, S., Lebecque, A., Chacornac, J. P., & Talon, R. (2006).
Saucissons secs fermiers du Massif central – Développement d’un
ferment indigène pour améliorer la sécurité et maintenir la typicité des
saucissons. Viandes et Produits Carnés, 25(5), 177–180.
Lebert, I., Leroy, S., Giammarinaro, P., Lebert, A., Chacornac, J. P.,
Bover-Cid, S., et al. (2007). Diversity of micro-organisms in environ-
ments and dry fermented sausages of French traditional small units.
Meat Science, 76, 112–122.
Lebert, I., Leroy, S., & Talon, R. (2007). Microorganisms in traditional
fermented meats – Chapter 11. In F. Toldra, Y. H. Hui, I. Astiasaran,
W. K. Nip, J. G. Sebranek, E. T. F. Silveira, L. H. Stahnke, & R.
Talon (Eds.), Handbook of fermented meat and poultry (pp. 113–1124).
Blackwell publishing.
Leroy, S., Chevallier, I., Lebert, I., Chacornac, J. P., & Talon, R. (2006).
Saucissons secs fermiers du Massif central – Identification et caractéri-
sation de la flore d’intére˛t technologique : bactéries lactiques et
staphylocoques à coagulase négative. Viandes et Produits Carnés,
25(5), 171–175.
Madsen, S. M., Beck, H. C., Ravn, P., Vrang, A., Hansen, A. M., &
Israelsen, H. (2002). Cloning and inactivation of a branched-chain-
amino-acid aminotransferase gene from Staphylococcus carnosus and
characterization of the enzyme. Applied and Environmental Microbiol-
ogy, 68(8), 4007–4014.
Marceau, A., Mera, T., Zagorec, M., & Champomier-Vergès, M. C.
(2001). Protein expression under uracil privation in Lactobacillus sakei.
FEMS Microbiology Letters, 200(1), 49–52.
Martı́n, B., Garriga, M., Hugas, M., & Aymerich, T. (2005). Genetic
diversity and safety aspects of enterococci from slightly fermented
sausages. Journal of Applied Microbiology, 98(5), 1177–1190.
Martı́n, B., Garriga, M., Hugas, M., Bover-Cid, S., Veciana-Nogues, M.
T., & Aymerich, T. (2006). Molecular, technological and safety
characterization of Gram-positive catalase-positive cocci from slightly
fermented sausages. International Journal of Food Microbiology,
107(2), 148–158.
Mauriello, G., Casaburi, A., Blaiotta, G., & Villani, F. (2004). Isolation
and technological properties of coagulase negative staphylococci from
fermented sausages of southern Italy. Meat Science, 67(1), 149–158.
Morot-Bizot, S., Talon, R., & Leroy-Sétrin, S. (2003). Development of
specific PCR primers for a rapid and accurate identification of
Staphylococcus xylosus, a species used in food fermentation. Journal of
Microbiological Methods, 55(1), 279–286.
Papamanoli, E., Tzanetakis, N., Litopoulou-Tzanetaki, E., & Kotzekidou,
P. (2003). Characterization of lactic acid bacteria isolated from a greek
dry-fermented sausage in respect of their technological and probiotic
properties. Meat Science, 65(2), 859–867.
Planchon, S. (2006). Aptitude de Staphylococcus carnosus et Staphylococ-
cus xylosus à former des biofilms. Etude d’une souche ‘‘biofilm positif’’
par une approche protéomique. Thèse de Docteur d’Université –
Spécialité Sciences des Aliments, Option Microbiologie. Université
Blaise Pascal, Clermont-Ferrand II, France.
Rantsiou, K., Drosinos, E. H., Gialitaki, M., Urso, R., Krommer, J.,
Gasparik-Reichardt, J., et al. (2005). Molecular characterization of
Lactobacillus species isolated from naturally fermented sausages
produced in Greece, Hungary and Italy. Food Microbiology, 22, 19–28.
Rantsiou, K., Urso, R., Iacumin, L., Cantoni, C., Cattaneo, P., Comi, G.,
et al. (2005). Culture-dependent and -independent methods to inves-
tigate the microbial ecology of Italian fermented sausages. Applied and
Environmental Microbiology, 71(4), 1977–1986.
Rebecchi, A., Crivori, S., Sarra, P. G., & Cocconcelli, P. S. (1998).
Physiological and molecular techniques for study of bacterial commu-
nity development in sausage fermentation. Journal of Applied Micro-
biology, 84, 1043–1049.
62 R. Talon et al. / Meat Science 77 (2007) 55–62
Reij, M. W., & Den Aantrekker, E. D. (2004). Recontamination as a
source of pathogens in processed foods. International Journal of Food
Microbiology, 91(1), 1–11.
Rosenstein, R., Nerz, C., Resch, A., & Götz, F. (2005). Comparative
genome analyses of staphylococcal species. In Proceedings of the
second european conference on prokaryotic genomics.
Rossi, F., Tofalo, R., Torriani, S., & Suzzi, G. (2001). Identification by
16S-23S rDNA intergenic region amplification, genotypic and pheno-
typic clustering of Staphylococcus xylosus strains from dry sausages.
Journal of Applied Microbiology, 90, 365–371.
Samelis, J., Metaxopoulos, J., Vlassi, M., & Pappa, A. (1998). Stability
and safety of traditional Greek salami – A microbiological ecology
study. International Journal of Food Microbiology, 44(1–2), 69–82.
Talon, R. (2006). Saucissons secs fermiers du Massif central – Principaux
résultats du projet européen. Viandes et Produits Carnés, 25(5),
187–188.
Talon, R., Lebert, I., Lebert, A., Leroy, S., Garriga, M., Aymerich, T.,
Drosinos, E. H., Zanardi, E., Ianeri, A., Fraqueza, M. J., Patarata, L.,
& Laukova, A. (2007). Traditional dry fermented sausages produced in
small-scale processing units in Mediterranean countries and Slovakia.
1. Microbial ecosystems of processing environments. Meat Sciences,
doi:10.1016/j.meatsci.2007.05.006.
Talon, R., & Leroy, S. (2006). Latest developments in meat bacterial
starters – Chapter 16. In L. M. L. Nollet & F. Toldra (Eds.), Advanced
technologies for meat processing (pp. 401–418). New York: CRC Press/
Taylor and Francis Group.
Talon, R., Leroy-Sétrin, S., & Fadda, S. (2004). Dry fermented sausages –
Chapter 23. In Y. H. Hui, L. Meunier-Goddik, A. S. Hansen, J.
Josephsen, W.-K. Nip, & P. S. Stanfield, et al. (Eds.), Handbook of
food and beverage fermentation technology (pp. 397–416). New York:
Marcel Dekker Inc..
Urso, R., Comi, G., & Cocolin, L. (2006). Ecology of lactic acid
bacteria in Italian fermented sausages: Isolation, identification and
molecular characterization. Systematic and Applied Microbiology,
29, 671–680.
Wagner, E., Doskar, J., & Götz, F. (1998). Physical and genetic map of the
genome of Staphylococcus carnosus TM300. Microbiology, 144(Pt. 2),
509–517.
 MEAT
SCIENCE
Meat Science 77 (2007) 63–80
www.elsevier.com/locate/meatsci

A fresh look at meat flavor

C.R. Calkins *, J.M. Hodgen
Department of Animal Science, University of Nebraska, A213 Animal Science, Lincoln, NE 68583-0908, United States

Received 13 March 2007; received in revised form 12 April 2007; accepted 12 April 2007

Abstract

Hundreds of compounds contribute to the flavor and aroma of meat. Complex interactions between various compounds influence the
perception of meat flavor. Inherent flavor of a meat product can be influenced by oxidation, lipid content, feeding/diet, myoglobin, and
pH. Diet plays an important role in both ruminants and nonruminants. New research reveals important relationships in flavor among
multiple muscles within a single animal carcass. This animal effect includes the presence of off-flavors. Diets high in polyunsaturated fatty
acids may be contributing to the appearance of off-flavors in beef. Compounds associated with liver-like off-flavor notes in beef have been
identified in raw tissue.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Flavor; Meat; Beef; Off-flavor; Liver flavor

1. Importance of flavor to eating satisfaction
Flavor is a very complex attribute of meat palatability.
The Beef Customer Satisfaction Survey found many factors
that influenced overall like ratings of top round, top sirloin,
and top loin steaks with flavor and tenderness contributing
equally to overall like ratings (Lorenzen et al., 1999; Neely
et al., 1998; Neely et al., 1999; Savell et al., 1999). For clod
steaks, flavor like had the largest simple correlation (0.86)
to overall like ratings and was the first variable to enter
the stepwise regression for predicting overall like rating
(Goodson et al., 2002).
While it has been known for many years that tenderness
plays a large role in acceptability of meat by a consumer, it
has become increasingly apparent that flavor also needs to
be addressed. In a large, multiple-city study, flavor was
found to be the most important factor affecting consumers’
meat buying habits and preferences when tenderness was
held constant (Sitz, Calkins, Feuz, Umberger, & Eskridge,
2005). Huffman et al. (1996) found flavor had a stronger
relationship (R2 = 0.67) to overall steak palatability ratings
*
Corresponding author. Tel.: +1 402 472 2907; fax: +1 402 472 6362.
E-mail address: ccalkins1@unl.edu (C.R. Calkins).
than any other factor when consumers prepared meat at
home. Platter et al. (2003) determined that small changes
in sensory ratings for flavor greatly influenced the overall
acceptability of steaks.
Because of the relationship of flavor to meat palatability
it is important to gain a better understanding of factors
that influence flavor in order to produce the most flavorful
and consistent product possible.
2. Basic flavor compounds formed upon cooking
There are literally hundreds of compounds in meat that
contribute to flavor and aroma. Many of them are altered
through storage and cooking, making meat flavor an
incredibly complex topic. Entire books have been written
on individual aspects of meat flavor. The intent of this
paper is to highlight some of the major flavor compounds
and to discuss a unique flavor problem with global implica-
tions currently facing beef producers in the US. The pri-
mary focus of this paper is the inherent flavor of meat.
That is, little attention will be given to post-cooking flavor
changes (like the development of warmed-over flavor).
Flavor and aroma compounds found in meat (Table 1)
include a broad array of compounds, including hydrocar-

0309-1740/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.016
64 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80

Table 1
Compounds identified in beef and characteristic flavors and/or aromas associated with those compounds
Compound name Characteristic flavors/aromas
Benzaldehyde Volatile almond oil, bitter almond, burning aromatic taste
Benzene Pleasant, distinct
sec-Butanamine Seafood, green, onion
Butenal Malty, green, roast
n-Caprioc acid Goaty
3-Carene Sweet and pungent odor but more agreeable than turpentine, orange peel, lemon, resin
Cyclobutanol Roasted
2,2,6-Trimethylcyclohexanone Mint, acetone
2,4-Decadienal Deep fat flavor, chicken flavor at 10 ppm, citrus/orange/grapefruit flavor at lower dilutions
Decanal Powerful, waxy, aldehydic, orange, citrus peel
2-Decenal Tallow, orange
1,3-Bis(1,1-dimethylethyl)benzene Cooked beef
2,3-Dimethyloxirane No reference
N,N 0 -Dimethyl 1,2-ethanediamine Ammonia
5-Ethylcyclopent-1-enecarboxaldehyde Fragrant, perfume
2-Pentylfuran Green bean, butter
2,4-Heptadienal Nut, fat
Heptanal Oily, fatty, rancid, unpleasant, penetrating fruity odor in liquid
5-Methyl 2-heptanamine No reference
1-Heptanol Fragrant, woody, oily, green, fatty, winey, sap, herb
2-Heptanone Fruity, spicy, cinnamon, penetrating fruity odor in liquid
6-Methyl 2-heptanone Cloves, menthol, eugenol
2-Heptenal Soapy, fatty, almond, fishy, unpleasant
3-Ethyl-2-methyl 1,3-hexadiene No reference
Hexanal Fatty-green, grassy, strong green, tallow, fat, unripe fruit when dilute
Hexane Faint peculiar odor
Hexanol Woody, cut grass, chemical-winey, fatty, fruity, weak metallic
2-Ethyl 1-hexanol Resin, flower, green
2-Hexen-1-ol Green, sharp, leafy, fruity, unripe banana
Hydroxymandelic acid No reference; catecholamine metabolite
Limonene Pleasant lemon-like, turpentine, citrus, fruity, fresh, light
Bis(2-trimethylsilyethyl ester) malonic acid No reference
3-Methylbutanal Pungent apple-like odor, malt
Methyl salicylate Cooling sensation, wintergreen, gaultheria
2,4-Nonadienal Fat, wax, green, watermelon, geranium, pungent
Nonanal Floral, citrus, fatty, grassy, waxy, green
2-Nonanone Hot milk, soap, green, fruity, floral
2-Nonenal Cardboardy, orris, fat, cucumber, paper
Octadecanal Oil
Octanal Harsh, fatty, orange peel, soapy, lemon, green, honey
1-Octanol Penetrating aromatic odor, fatty, waxy, citrus, oily, walnut, moss, chemical, metal, burnt
2-Methyl 3-octanone Herb, butter, resin, gasoline
2-Octenal Green, nut, fat
(Z)-3-Octene Fruity, old apples
1-Octen-3-ol Mushrooms, compound excreted by many insects
2-Octen-1-ol Green citrus
3-Octen-2-one Nut, crushed bug, earthy, spicy, herbal, sweet, mushroom, hay, blueberry
Hexadecyl oxirane No reference
Pentanal Almond, malt, pungent, acrid
Pentane Very slight warmed-over flavor, oxidized
1-Pentanol Mild odor, fusel oil, fruit, balsamic
5-Amino 1-pentanol Mild
a-Pinene Piney, fruity, citrus, turpentine
b-Pinene Pine, citrus, fruity, resin, turpentine
Piperazine Salty
N-Methyl 1,3-propanediamine No reference
Propanol Alcoholic
Styrene Penetrating odor, sweet smell
Tetradecane Alkane
Tridecane Alkane
2-Tridecenal Sweet, strong, spicy
Undec-4-enal No reference; animal pheromone
 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
bons, aldehydes, ketones, alcohols, furans, thriphenes,
pyrrols, pyridines, pyrazines, oxazols, thiazols, sulfurous
compounds, and many others (MacLeod, 1994; Ho, Oh,
& Bae-Lee, 1994). A few of the key reactions will be
discussed below.
2.1. Maillard reaction
The Maillard reaction, or nonenzymatic browning,
helps explain amine and carbonyl reactions in food. In gen-
eral, amino compounds condense with the carbonyl group
of a reducing sugar in the presence of heat. This produces
gylcosylamine which is rearranged and dehydrated to form
furfural, furanone derivatives, hydroxyketones, and dicar-
bonyl compounds. All of these compounds contribute to
flavor. As the reaction progresses, the intermediates can
react with other amines, amino acids, aldehydes, hydrogen
sulfide, and ammonia through the Amadori rearrangement,
Strecker degradation, and Schiff bases pathways.
Once the reaction has progressed through the Schiff
base, Strecker degradation, or other pathways, the reac-
tions can lead to melanoidins (brown, high molecular
weight polymers from the condensation of cyclic com-
pounds) (Fay & Brevard, 2005). These products can be
pleasing or unacceptable flavors and aromas (Manley &
Choudhury, 1999).
Different sugars and amino groups can produce different
end products. Cysteine and glucose produce mainly sulfur
compounds whereas cysteine and glucose under oxidized
conditions produce more pyrazines and furans (Tai &
Ho, 1997). Maillard volatile compounds from glutathione
and glucose produce sulfur-containing compounds – thi-
ophenes, thiazoles, and cyclic polysulfides – at pH 6 and
8, but furans are more often the end products at more
acidic pH. When glutathione is oxidized it becomes gluta-
thionesulfonic acid, which produces furans, carbonyls, pyr-
roles, and pyrazines with glucose. Sulfur-containing
compounds are not formed when glutathione is oxidized
(Tai & Ho, 1998).
Cysteine and ribose, through the Maillard reaction, as
well as thiamin have been shown to create compounds such
as 2-methyl-3-furanthiol (Mottram & Whitfield, 1994).
From these compounds thiols and disulfides can originate.
An exhaustive review of the nine most common aro-
matic compound classes from precursors of the Maillard
reaction can be found in Manley and Choudhury (1999).
However, more extensive research is needed to determine
the role of each amino acid and sugar in meat to create
the browned, cooked meat flavor. It would be useful to
know if there are differences among muscles, cooking meth-
ods, and degrees of doneness, and how these differences
affect the end flavor profile.
2.2. Flavor compounds
Sulfurous and carbonyl compounds seem to be the pre-
dominant contributor to meat flavor (Mottram & Mad-
65
ruga, 1994; Shahidi, 1989). Gasser and Grosch (1988)
established that 2-methyl-3-furanthiol and bis(2-methyl-3-
furyl) disulfide contribute to the desirable aroma of beef
when isolated from other possible compounds contributing
to meat aroma. Bis(2-methyl-3-furyl) has an odor threshold
of 0.02 ng/g in water and methylating 2-methyl-3-furanth-
iol increases the odor threshold. This suggests 2-methyl-3-
(methylthio)furan is only a minor contributor to the flavor
of meat (MacLeod, 1986). Gasser and Grosch (1988) also
established flavor dilution factors (FD; aroma extracts
are stepwise diluted until sniffers cannot detect odorants
with the highest level reported as flavor dilution) to evalu-
ate the contribution of a compound to overall beef flavor.
Forty compounds were determined in cooked beef to have
FD factors larger than four, and 17 compounds, which
where stated to be major contributors to cooked flavor,
had FD factors larger than 64. Farmer and Patterson
(1991) isolated five compounds that were found to be desir-
able to cooked beef flavor: bis(2-methyl-3-furyl) disulfide,
2-furfuryl 2-methyl-3-furyl disulfide, bis(2-furfuryl) disul-
fide, dimethylfuryl 2-methyl-3-furyl disulfide, 2-methyl-3-
furyl 2-methyl-3-thienyl disulfide.
A Strecker aldehyde, methional, is a low threshold
(0.2 lg/g in water) sulfur compound that is described as
‘pleasant, warm-meat, or soup-like’ (Gasser & Grosch,
1988). The roasty note in beef is in part produced by
2-acetyl-1-pyrroline and 2-acetylthiazole.
Many compounds that contribute to meat smell and
flavor are lipid breakdown products. Fatty acids such as
linoleic and arachidonic acid start to autoxidize to 9-hydro-
peroxide and 11-hydroperoxide, respectively, which can
form 2,4-decadienal, 2-nonenal, 1-octen-3-one, 2,4-nonadi-
enal, and 2-octenal through b-scission with 2-nonenal and
2,4-decadienal having as high of FD values as the sulfur
compounds contributing to meaty flavor.
Through oxidation of b-carotene, a very intense aro-
matic b-ionone is formed (Sanderson, Co, & Gonzalez,
1971) which likely comes from animal feed (Gasser &
Grosch, 1988). One compound responsible for the tallowy
and/or beef-like smell is 12-methyltridecanal (Guth &
Grosch, 1993), a compound derived from plasmalogens,
glycerophospholipids with a long chain fatty aldehyde
linked to the sn-1 position by a vinyl ester bond (Dannen-
berger et al., 2006; Guth & Grosch, 1993). The 12-methylt-
ridecanal’s odor threshold is 0.1 lg/kg in water and the
compound is found in higher amounts in lean from beef
than other species (Guth & Grosch, 1993). Increasing the
amount of forage in the diet of cattle increases 12-methylt-
ridecanal in the phospholipids of muscles as well as
n-octadecanal, another plasmalogen. Octadecenal, a mono-
unsaturated plasmalogen with a low odor threshold, was
reduced when forages were fed instead of concentrates,
but n-hexadecanal concentration was not affected by diet
(Dannenberger et al., 2006).
Hexanal and 2,4-decadienal contribute positively to beef
flavor, but may produce undesirable flavors at higher con-
centrations (Melton, 1983). This is probably because these
66 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
two compounds are produced in the greatest amounts dur-
ing oxidation of 18:2 during heating as well as overshadow-
ing of compounds that also help produce typical beef
flavors. Others have found that hexanal is the most prom-
inent volatile compound in cooked meat with the amount
being directly proportional to thiobarbituric acid reactive
substances (TBARS), a measure of oxidation, and inversely
proportional to flavor acceptability (Shahidi & Pegg, 1994;
Ullrich & Grosch, 1987).
2.3. Species differences in fatty acid deposition
Because of differences in digestive systems between
ruminants and nonruminants, deposition of fatty acids is
different. Poultry and pork muscle typically have higher
levels of polyunsaturated fatty acids than lamb or beef.
Pork muscle has more linoleic acid than beef or lamb which
contributes to the higher polyunsaturated:saturated fatty
acid ratio. However, beef and lamb commonly have more
favorable n6:n3 fatty acids ratios than pork (Wood &
Enser, 1997). Through diet manipulation some change
can in the fatty acid composition can be achieved. Since
dietary polyunsaturated fatty acids undergo little change
during digestion in pork and poultry, feeding products with
higher levels of linolenic, eicosapentaenoic, and docosa-
hexaenoic has been studied to modify the n-6 concentration
of fatty acids in those species.
Modifying the levels of unsaturated fatty acids in pork
and poultry has lead to conflicting reports on the effect
on flavor. Rhee, Davidson, Cross, and Ziprin (1990) found
all taste panel palatability scores for pork improved as oleic
acid concentration increased. Myer et al. (1992) found
feeding high oleic acid peanuts to pigs did not affect taste
panel scores for flavor. However, they did see an increase
in off-flavor scores when canola oil was fed at levels where
the content of linolenic was raised fivefold in subcutaneous
fat. Shackelford, Reagan, Haydon, and Miller (1990) also
found off-flavors increased with feeding canola oil to pigs,
compared with other high oleic acid dietary products,
which they attributed to the increase in 2-pentenal and
2,4-heptandienal, derivatives of linolenic acid. Based on
these studies it appears raising monounsaturated fatty
acids does not lead to the negative flavor profile that is seen
by increases in polyunsaturated fatty acids in pork.
Wood and Enser (1997) reviewed factors influencing
fatty acids in meat and their effect on quality and con-
cluded that poultry and pork respond similarly when trying
to manipulate the n6:n3 fatty acid ratio. Increases in n3
fatty acids in bacon, like those found when feeding fish
oil (eicosapentaenoic acid and docosahexaenoic acid), lead
to increases in perception of off-flavors by taste panelists.
This trend was not seen for the pork loin (Romans, John-
son, Wulf, Libal, & Costello, 1995; Romans, Wulf, John-
son, Libal, & Costello, 1995).
Camfield, Brown, Lewis, Rakes, and Johnson (1997)
presented simple correlations of medium and long chain
fatty acids to specific flavors and concluded unsaturated
fatty acids were positively correlated to cowy, cardboard,
painty, and livery flavors that are unsatisfactory for beef
palatability.
Differences in production systems create contradictory
findings in the relationship of fatty acid type and concen-
tration to flavor. Some studies have shown flavor differ-
ences when beef cattle are fed forage- and cereal-based
diets (Mandell, Buchanan-Smith, & Campbell, 1998; Mel-
ton, 1990; Raes et al., 2003; Young & Baumeister, 1999).
In one review, it was concluded flavor differences could
not be reliably determined (Muir, Deaker, & Brown,
2003). There are many reasons for the discrepancies in
these studies. Consumer preferences and animal manage-
ment systems vary greatly, as do the amount of grain/con-
centrate and type of forage used for comparison. Several
studies have investigated breed effects, age of the animal,
estimated fat cover end point, or supplements, all of which
could impact the fatty acid profile of fat deposits. Studies
with USA consumers demonstrate they prefer meat from
grain-finished beef (Killinger, Calkins, Umberger, Feuz,
& Eskridge, 2004; Sitz et al., 2005). However, those same
studies revealed a small sector of the US population that
prefers the flavor and palatability of steaks from grass-fin-
ished animals. This split preference for meat from different
production systems is probably true in many countries. In
general, grass-fed and grain-fed ruminant animals have
been shown to have different fatty acid profiles with
grass-fed animals having higher levels of polyunsaturated
fatty acids, especially linolenic acid and other n-3 fatty
acids, while concentrate-fed animals contain higher levels
of n-6 fatty acids and a lower proportion of n-3 fatty acids.
3. Effect of fat, lipid content and oxidation on flavor
3.1. Fat and lipid content
It is commonly known that the degree of lipid oxidation
in meat is dependent on the composition of the phospholip-
ids, amount of polyunsaturated fatty acids, and the concen-
trations of metal ions, oxygen, salt and other pro-oxidants.
Peroxides are formed by the free radical chain mechanism
between a polyunsaturated fatty acid and oxygen. This oxi-
dation forms products like aldehydes, lactones, hydrocar-
bons, furans, and ketones that create undesirable, rancid
off-flavors due to the oxidation of meat (Ladikos & Lougo-
vois, 1990).
Oxidation can be controlled by the amount of antioxi-
dant compounds found in the muscle tissue. Grass-fed beef
may not be much more prone to lipid oxidation than grain-
fed beef because of the increased levels of vitamins A, C,
and E, cartonenoids, and flavonoids found in forages
(Wood & Enser, 1997). Grain-fed animals also consume
polyphenols or may be fed vitamin E supplements during
the finishing period which act as antioxidants.
Lipids serve several roles in flavor development. Lipids
in meat act as a solvent for the volatile compounds that
develop during production, handling, and thermal process-
 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
ing (Moody, 1983). They undergo thermal oxidative
change to produce compounds that influence beef flavor
and react with components of lean tissue to give distinct
flavor compounds (Mottram & Edwards, 1983).
Correlations between 14:1, 16:1, 18:0, 18:1, 18:2, 18:3,
and desirable beef flavor have been reported (Melton,
Black, Davis, & Backus, 1982). Although species flavor
depends primarily on unsaturated aldehydes, fatty acids,
ketones, and saturated aldehydes all play a role in beef fla-
vor – especially since many of the aldehydes are derived
from pathways with fatty acids (Melton et al., 1982).
Fat content has been shown to affect palatability,
including flavor. As intramuscular fat increases, the fat fla-
vor increases which is preferred by most US consumers
(Miller, Moeller, Goodwin, Lorenzen, & Savell, 2000).
The minimal level of intramuscular fat in beef for US con-
sumer acceptance and preference, described as slightly
intense fat flavor, is approximately 3% (Miller, 2001). Lev-
els of fat above 7.3% in meat may have a negative effect on
perception of flavor and acceptability (Miller, 2001). Fran-
cis, Romans, and Norton (1977) also found a bell-shaped
curve for the effect of marbling on flavor. Conversely, loin
steaks had a linear decrease in flavor desirability as quality
grade went down from USDA Prime through USDA Cut-
ter (Smith, Savell, Cross, & Carpenter, 1983). Flavor desir-
ability ratings in top round steak were less affected by
grade and marbling score.
3.2. Oxidation and volatiles from fatty acids
To determine the volatile compounds that are derived
from linoleic acid and methyl linoleate, Ullrich and Grosch
(1987) used gas chromatography to derive a D-value (the
highest dilution at which the aroma of a substance is still
detected) to reveal the most intense flavor compounds.
With both lipids, hexanal, 2-octenal, and 2-nonenal had
the highest D-values. The fourth highest compound for lin-
oleic acid was 1-octen-3-ol while 1-octen-3-one was for
methyl linoleate. After 24 h of linoleic autoxidation, 2-non-
enal was the most potent volatile, with hexanal the most
potent volatile after 48 h, and hexanal and 2-octenal the
most potent volatiles after 72 h. The authors also found
pentane to be a better indicator of lipid peroxidation
because it has a shorter induction period than other volatile
compounds even though its D-value is not as high.
Arachidonic acid, a long chain omega-6 polyunsatu-
rated fatty acid, was originally found to be autoxidized in
meat into hexanal, methyl 5-oxopentanoate, pentane, and
2,4-decadienal volatile compounds (Artz, Perkins, & Salva-
dor-Henson, 1993). However, the most intense aroma
compound from the oxidation of arachidonic acid is
trans-4,5-epoxy-(E)-2-decenal followed by 1-octen-3-one,
2,4-decadienal, 2,4,7-tridecatrienal, and hexanal (Blank,
Lin, Vera, Welti, & Fay, 2001).
For human health reasons, there is a desire to increase
the PUFA in the lean portion of meat. However, when lev-
67
els become too high, off-flavors can develop, especially dur-
ing cooking (Elmore, Campo, Enser, & Mottram, 2002).
4. Effect of diet on flavor
4.1. Flavor intensity
The primary focus on the effect of diet and flavor accept-
ability has been comparing pasture-fed cattle to grain-fed
cattle. A wide range of results have been reported: some
papers suggest there are no differences between forage-fed
and grain-fed animals and others stating there are large dif-
ferences. Most of the differences can probably be explained
by the different production systems which affect the level of
energy intake, days on feed, growth rate, age of the animal,
fat deposition, fat composition, and carcass weight. Brown,
Melton, Riemann, and Backus (1979) stated sensory panels
do not find a lack of flavor in grass-fed beef, ‘but rather the
presence of an off-flavor.’
4.2. Lipid deposition and fatty acid oxidation
Compared to same-age steers fed corn silage-, pasture-,
and Bermuda pellets, steers finished 90 d on high energy,
corn-based diets had more desirable or intense beef flavor
(Melton, 1983). When feeding to a constant fat thickness
in different production management systems, flavor differ-
ences existed (Bowling et al., 1978). Even when comparing
corn diets to corn silage diets, significant differences were
seen in flavor profiles of beef, although not to the extent
of grass or alfalfa finished steers (Berry, Leddy, Bond,
Rumsey, & Hammond, 1988). Conversely, no difference
in flavor was seen between the grass- and grain-fed animals
when animals were blocked by growth rate (French et al.,
2001). The high growth rate animals fed on grass had little
difference in meat quality to concentrate-fed cattle which
the authors attributed to high protein turnover (French
et al., 2001). Global differences in flavor preference may
partially explain the latter result.
Several grasses in ruminant diets have been demon-
strated to cause less desirable meat flavor (Melton, 1990).
In contrast, Bidner, Montgomery, Bagley, and McMillin
(1985) found no difference in flavor intensity in meat from
animals fed high quality Bermuda grass pasture compared
to corn-based diets. French, O’Riordan, et al. (2000)
reported no differences in flavor after aging the meat 2 d
when steers were finished on autumn grass, grass silage,
or concentrate diets with low levels of supplements to
maintain constant carcass growth rate between treatments.
Melton (1983) suggested differences in results could be due
to differences in sensory panels or quality of the grasses.
Hay diets were also found to produce meat less desirable
in flavor than corn silage diets with no direct link to intra-
muscular fat (Dube et al., 1971), while another study
showed the opposite effect (meat from animals on a 91%
corn diet were less desirable in flavor than meat from ani-
mals fed alfalfa or Tmothy hay) when using hay as the
68 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
energy source (Oltjen, Rumsey, & Putnam, 1971). Further-
more, hay versus grass silage diets fed at the same net
energy do not affect flavor (Listrat et al., 1999). Melton
(1983) concluded corn could be replaced partially or totally
with high quality alfalfa or in combination with Timothy
hay without a significant change in flavor.
Corn is the staple grain used in grain-fed cattle in the US
while Canada and Japan predominately use barley. Sitz
et al. (2005) and Jeremiah et al. (1998) found US consum-
ers preferred the flavor of domestic beef over Canadian
barley-fed beef. In contrast, no differences in flavor inten-
sity, 12 aromatics, two mouthfeels (astringent and metal-
lic), and three tastes – as determined by a trained flavor
and descriptive panel – were found when comparing corn,
barley, and 50–50 corn/barley diets in the meat from young
animals (Miller, Rockwell, Lunt, & Carstens, 1996).
The majority of the flavor effect due to feeding of for-
ages is hypothesized to be due to changes in lipid deposi-
tion and fatty acid composition. Using sheep as a
ruminant model, Lee et al. (2004) hypothesized red clover
fed to grass-finished steers would increase both n-6 and
n-3 polyunsaturated fatty acids (PUFA) due to reductions
in ruminal biohydrogenation of PUFA. French, Stanton,
et al. (2000) found meat from cattle that were grass supple-
mented to maintain constant growth rate with concentrate-
fed animals had a linear decrease in saturated fats and n-
6:n-3 PUFA ratios and increases in unsaturated fats and
conjugated linoleic acid (CLA) when concentrate percent-
age went down, without affecting flavor scores (French,
O’Riordan, et al., 2000). Fishy off-flavor was significantly
higher and overall flavor liking scores were significantly
lower in meat from grass-finished cattle with increased
18:1trans isomers and, notably, CLAcis-9, trans-11 (Nuern-
berg et al., 2005). Animals backgrounded on grass and then
finished approximately 190 d on a high energy diet of
silage, hay, and barley had meat with higher levels of n-3
fatty acids than animals fed concentrate after weaning,
but no difference in CLAcis-9, trans-11 in the lipids of
the longissmus muscle (Dannenberger et al., 2004) and sub-
cutaneous fat (Dannenberger et al., 2005). However, CLA-
trans-7, cis-9 was the second most abundant CLA isomer in
meat from concentrate-fed animals whereas CLAtrans-11,
cis-13 was the second most abundant in grass-fed. Total
CLA isomers were increased in the longissmus, subcutane-
ous fat, heart, and liver, but not in the semitendinosus
(Dannenberger et al., 2005) in grass-fed animals. Most
importantly, this study showed D9-desaturase activity was
decreased due to pasture feeding. This elongase, in con-
junction with trans vaccenic acid, is responsible for the syn-
thesis of CLAcis-9, trans-11. By disrupting the elongase
activity, flavor changes might occur because of the unused
trans vaccenic acid, a fatty acid implicated in off-flavors
(Camfield et al., 1997) as well as less CLAcis-9, trans-11
(Dannenberger et al., 2005). French, Stanton, et al.
(2000) also hypothesized the increase in CLAcis-9, trans-
11 in animals on higher grass-based diets was due to a
change in biohydrogenation. However, they concluded
grass diets favored the growth of Butyrivibrio fibrisolvens,
the ruminal bacterium responsible for producing the lino-
leic acid isomerase.
With interest to increase the PUFA in beef, trials with
supplements high in certain fatty acids have been con-
ducted. Most attempts have been made using linseed, lin-
seed oil, sunflower, sunflower oil, and fish oil (Mandell,
Buchanan-Smith, Holub, & Campbell, 1997; Scollan
et al., 2006). Generally, levels of PUFA in the lean have
not been high enough to claim health benefits, but some
negative flavors due to oxidation and shorter shelf-life have
been reported (Miller, 2001). The long chain fatty acids in
fish oil (Richardson et al., 2004) and several long chain
fatty acids from plant oils can bypass rumen biohydrogena-
tion with minimal change (Scollan et al., 2004). This
increase in unsaturation can lead to negative flavor
perception.
After 80–90 d on a corn finishing diet, no further signif-
icant beef flavor changes occur (Melton et al., 1982). Liver
flavor intensity increased up to 86 d of corn-based diets,
and sour flavor intensity decreased to a minimum at
122 d on corn, while metallic and off-flavor intensity were
unaffected by time-on-feed. Fishy and milky-oily flavors
decreased linearly with time-on-feed. Melton et al. (1982)
hypothesized increased beef fat and liver flavor with
decreased milky-oily, sour and fishy flavor gave a more
desirable beef flavor in corn-fed beef. Mandell et al.
(1998) disagreed with this hypothesis as they found liver
flavor was positively correlated to metallic and grassy
aroma, sour flavor, and metallic and grassy aftertaste and
negatively correlated to beef flavor. They also found sour
flavor notes were not affected by production type, but
metallic aroma was affected due to the differences in 18:1
and 18:3 in the meat from the different feed sources.
The biggest difference in the flavor of meat from grass-
and grain-fed beef animals probably is due to fatty acid
concentration and type as fatty acids are the primary
source of carbonyl compounds (Melton, 1983). Oleic and
linoleic acid are found in higher concentrations in grain-
fed diets than in grass-fed diets (Enser et al., 1998; Vasta
& Priolo, 2006) while a-linolenic is higher in grass-based
diets (Enser et al., 1998). Therefore, compounds which
are derived from linolenic acid (4-heptenal, 2,4-heptadie-
nal, and 2,6-nonadienal) are usually in higher concentra-
tion in meat from grass-fed animals while hexanal,
2-heptenal, and 2,4-decadienal (products of linoleic acid)
are typically found in higher concentrations in meat from
grain-fed animals (Larick et al., 1987). Those compounds
have been shown to be higher in beef with higher levels
of oxidation. Furthermore, beef from corn-based diets
has higher levels of glucose (Brown et al., 1979; Melton
et al., 1982), c- and a-tocopherol (Yang, Lanari, Brewster,
& Tume, 2002), and carotenoids (Melton, 1990; Yang
et al., 2002).
It is important to note most of these findings on flavor
were studied in the US. An individual usually comes to pre-
fer the foods he/she grew up eating. Sitz et al. (2005) and
 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
Killinger et al. (2004) found the greatest sensory difference
between Australian or Argentine (respectively) grass-fed
and USA grain-fed beef to be flavor when Warner-Bratzler
Shear force values were kept constant. Canadian, barley-
finished cattle were also rated less desirable for flavor than
domestic beef (Sitz et al., 2005). However, 19% of the con-
sumers in the study preferred the Australian meat when
compared to domestic beef while 29.3% preferred the
Canadian-fed beef when compared to domestic beef. Con-
sumers in both studies were willing to pay a premium for
their preference, which was heavily influenced by flavor.
4.3. Volatile compounds
Larick et al. (1987) investigated differences in volatile
compounds in forage systems and grain-fed cattle. Fat
from animals finished on tall fescue, brome grass-red clover
versus orchard grass-red clover pastures were not different
in volatile compounds, but 31 volatiles were in different
concentration in the meat from grain-fed animals. Volatiles
higher in the meat fat from grass-fed animals included pen-
tanoic, heptanoic, octanoic, nonanoic, decanoic, and
dodecanoic acids; heptanal, 2,3-octanedione, 3-hydroxy-
octan-2-one, 2-decenal, 2-tridecanone, hexadecane, hepta-
decane, octodecane, d-dodecalactone, phyt-1-ene, neophyt-
adiene, phyt-2-ene, an isomer of neophytadiene,
2-heptadecanone, dihydrophytol, and phytol with the terp-
enoids in much higher concentration due to rumen-fer-
mented chlorophyll (Suzuki & Bailey, 1985). The fat from
the grain-produced animals was higher in d-tetradecalac-
tone and d-hexadecalactone (Larick et al., 1987). These lac-
tones are derived in the rumen by the oxidation of linoleic
and oleic acids (Vasta & Priolo, 2006). In the study, Larick
et al. (1987) found phyt-2-ene to be highly correlated to
beef flavor intensity while d-tetradecalactone and d-hexa-
decalactone were negatively correlated to grassy flavor.
Pentanal, toluene, 1-ethyl-2-methylbenzene, and an
unknown compound explained 51% of the variation of beef
fat flavor intensity between grass and grain finished (Mel-
ton, 1990). As days on feed increased, pentanal, hexanal,
4-methyl-3-penten-3-one, nonane, acetone, nononal, and
two unknown compounds increased while trans-3-octene,
cis-2-octene, toluene, 3-penten-2-one, 3-hydroxy-2-buta-
none, and five unknown compounds decreased (Melton,
1990).
Several classes of compounds are affected by the ani-
mal’s diet. Descalzo et al. (2005) found more aldehydes
in meat from animals eating concentrate diets rather than
grass. Phenolic compounds are secondary metabolites of
plants so they are typically found in higher concentration
in meat of forage-finished animals compared with grain-
finished with the exception of 4-ethylphenol and cresols
(Vasta & Priolo, 2006). Diet plays a large role on indoles
and their derivatives with grass-fed animals having much
higher levels, especially skatole. Production of these indoles
from ruminal microorganisms can be reduced by feeding
feedstuffs with higher levels of tannins for a few days
69
(Vasta & Priolo, 2006). The volatile 2,3-octanedione has
been suggested as an indicator of grass-fed animals since
the compound is produced by a lipoxygenase on leafy
plants (not seeds) (Young, Berdague, Viallon, Roussett-
Akrim, & Theriez, 1997) and soybeans (Elmore et al.,
2004). This compound can also be derived from heating
and breaking down linoleic acid (Elmore et al., 2002) so
care is needed if using the compound as an indicator of
grass-fed animals. Terpenoids are directly transferred from
grass to animal tissue so these compounds are also consid-
ered a green forage indicator except for b-gurjunene and
limonene, which are higher in concentrate-fed animals
(Vasta & Priolo, 2006). Cornu, Kondjoyan, Frencia, and
Berdague (2001) discovered several terpenoids, including
b-pinene, in beef could be used to determine the region that
an animal came from based on volatile compounds from
the forages in the geographic area that were ingested by
the animal.
Because of sulfur’s low threshold, the small amount of
these volatile compounds in meat plays a significant role
in meat flavor (Drumm & Spanier, 1991) with aldehydes
from PUFA playing a role in the synthesis of these hetero-
cyclic compounds (Vasta & Priolo, 2006). Typically, sulfur
compounds are in higher concentration in grass-fed cattle
because of the tendency of the fatty acids to convert to
aldehydes during thermal processing (Elmore, Mottram,
Enser, & Wood, 1999).
4.4. Oxidative stability
The source of feed can play a role in the oxidative stabil-
ity of beef. When cattle were finished on a mixed diet of
silage, hay, and concentrate (corn, beet pulp, and linseed
cattle-cake), TBARS were always significantly higher than
grass-fed animals regardless of age of animal or storage
condition of the meat (Gatellier, Mercier, Juin, & Renerre,
2005). Brown et al. (1979) also found ground beef from
steers on low energy diets had more free fatty acids and
lower TBARS than meat from animals that consumed a
high energy diet. This was attributed to the increased levels
of vitamin E in biological membranes and fat of grass-fed
animals although it was noted the grain diet also contained
antioxidants of proanthocyanidins and phytic acid (Gatel-
lier et al., 2005). In the same study, a higher heme iron con-
tent (considered to be a pro-oxidant) was found in the
heifer and cow meat on the mixed diets compared to the
grass diets and the steers, which they concluded also
affected the increased oxidation. Interestingly, when
grain-fed animals are supplemented with vitamin E, the
same level of tocopherol is achieved in the lean tissue,
and the meat is more stable following 47 d vacuum pack-
aged storage than grass-fed beef with or without supple-
mentation. Therefore, 4–6 lg/g of a-tocopherol in the
meat of supplemented grain-fed animals appears adequate
to minimize lipid oxidation, but not in grass-fed beef (Yang
et al., 2002).
70 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
5. Myoglobin and heme iron concentration in beef
It has been suggested myoglobin has a close relationship
with lipid oxidation (Renerre & Labadie, 1993) which
could influence flavor. However, little evidence is present
to support these theories even though heme iron, myoglo-
bin, and total iron content have been shown to differ in
grass-fed (iron = 3.7 mg/100 g muscle) and grain-fed ani-
mals (iron = 2.5–2.7 mg/100 g muscle; Srinivasan, Xiong,
Blanchard, & Moody, 1998), different animal maturity lev-
els (older animals have higher levels; Boleman, Miller,
Buyck, Cross, & Savell, 1996), and gender (steers have
lower levels of heme iron than heifers; Gatellier et al.,
2005). Yancey et al. (2006) found livery flavor in beef
increased and beef flavor decreased when total iron levels
increased in the M. gluteus medius. In general, total iron
was not a good indicator of beef flavor attributes. This con-
clusion for total iron was also seen in Jenschke, Eskridge,
and Calkins (2006) for normal-flavored beef samples versus
liver flavored samples (P = 0.23). Small correlations
between liver-like flavor and myoglobin concentrations in
beef were seen, while hemoglobin relationship to livery fla-
vor was not significant (Yancey et al., 2006).
Heme iron was significantly correlated (0.51) to off-fla-
vor intensity for just one (M. vastus lateralis) of seven mus-
cles from the chuck and round (Meisinger, James, &
Calkins, 2006). In this study, regression equations contain-
ing linear and quadratic functions of heme iron concentra-
tion and muscle pH were established. Some relationships
between specific flavors in certain muscles were observed,
but with the small number of off-flavored samples no con-
crete trends were established. However, heme iron has been
thought to be a pro-oxidant that may influence meat flavor
due to its reaction to produce radicals that can promote
lipid oxidation in biological systems (Batifoulier, Mercier,
Gatellier, & Renerre, 2002; Kanner & Harel, 1985).
6. Effect of pH on flavor
The pH of food plays a role in the development of fla-
vors in the Maillard reaction. As pH increases, color and
polymeric compounds increase and nitrogen-containing
compounds like pyrazines are favored (Mottram & Mad-
ruga, 1994). Since fresh meat only has a pH range of
around 5.5–6.0 with a good buffering ability, little work
has been done to investigate the effect of pH on Maillard
products although meat above the normal pH range is per-
ceived to have a decrease in meat flavor intensity. There has
been some thought the higher ultimate pH in meat from
grass-fed animals might favor the formation of thiazoles
and thiophenones because of the availability of amino acid
degradation products while decreasing other sulfur vola-
tiles that favor lower pH (furanthiols, mercaptokin, ali-
phatic sulfides, and thiopenes: breakdown products of
cysteine; Mottram & Madruga, 1994).
Many compounds contributing to beef flavor are water-
soluble. As pH increases in meat, the proteins have
increased water binding properties. During cooking fewer
water-soluble proteins are lost from high pH meat since
there is less cooking loss (Miller, 2001). Dark, firm and
dry (DFD) meat is said to have a musty/moldy, cowy/
grassy, or bloody/serumy aromatic flavors and very high
beef flavor intensity (Miller, 2001). High levels of sodium
and phosphate in manufactured meats can lead to some
of the same flavor perceptions that are in DFD meat. While
beef carcasses with lower than average pH are not com-
mon, the meat from these animals is usually blander.
Lower pH levels can also be attained through the use of
acidic ingredients (Miller, 2001).
7. Flavor relationships among muscles
Most of the research comparing muscles has dealt with
tenderness because there is approximately 3–4 times the
variation between muscles in tenderness compared to fla-
vor (Shackelford, Wheeler, & Koohmaraie, 1995; Wulf &
Page, 2000), especially in the M. longissimus dorsi. In most
studies where flavor or off-flavor scores are determined, the
muscle used is the M. longissimus dorsi. Over the years
some researchers have investigated palatability characteris-
tics, including flavor, from more than one muscle to try to
increase the value and usage of specific muscles. Tables 2
and 3 list the rankings of muscles for flavor intensity and
off-flavor intensity, respectively, from various studies. In
most of the studies the differences in beef flavor intensity
between muscles were relatively small.
7.1. Flavor desirability
Beef flavor intensity was positively correlated to off-fla-
vor intensity (r = 0.71) and weakly correlated to all other
traits (tenderness, r = 0.14; amount of connective tissue,
r = 0.11; juiciness, r = 0.13; sarcomere length, r =
0.31; percentage of desmin degraded, r = 0.34; cooking
loss, r = 0.20) except collagen concentration and shear
force (Rhee, Wheeler, Shackelford, & Koohmaraie,
2004). When simple correlations were run within muscle,
every muscle had significant correlations between flavor
intensity and off-flavor intensity; the only other correla-
tions with off-flavor intensity were the infraspinatus’s cor-
relation to collagen concentration (r = 0.38) and the
longissimus correlation to juiciness (r = 0.44). Jeremiah,
Dugan, Aalhus, and Gibson (2003) found no correlations
between beef flavor intensity and chemical characteristics
for 33 muscles.
Meisinger et al. (2006) found the M. infraspinatus had
the least off-flavors and the lowest frequency of sour notes
of the six chuck and round muscles tested (Table 4). The
M. vastus medialis had the most intense off-flavor ratings
with a high frequency of sour, charred, and oxidized flavor
notes. Liver-like, bloody, and rancid flavor notes and heme
iron content did not differ among muscles. Off-flavor inten-
sity within the M. rectus femoris, M. teres major, M. vastus
lateralis, and M. vastus medialis was significantly related to
 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80 71

Table 2
Ranking of musclesa for flavor intensity from different studies
Rankb 1 2 3 4 5 6 7 8 9 10 11
c c c c c c c c
1 LD IF DI BF TB CP LD LD SV SV LD
2 QFcd SVc IF PMcd SMc SP BFd SMc VIcd MDcd GM
3 STde CPc BF GMcd LDc TB SMde BFc CPcde GRcde SM
4 GMdef LDc SV SMcde SV TBdef GMc MDcdef PPcde
5 SMefg PPcd IP TBcdef RB RFefg IFcdef TBcde
6 TBefg TBcd PM RFdefg SL STfg TBdefg IFcde
7 BFfgh SPde TM LDdefg LT SPfg RBdefg BFcdef
8 SPfgh RBe TB SVdefg SS GMfg VMdefg CPcdef
9 PMgh BBe SP IFefg ADg SSefgh VIcdef
10 IFh AD STefg IFg LTefgh SLcdefg
11 TF PPfg PMh SPefgh LTcdefg
12 GM SPg STefgh VLcdefgh
13 GR PPefgh SSdefgh
14 RA VLefgh RBdefgh
15 PP SFefgh SMdefgh
16 SD BFefgh SPefgh
17 OA GRfgh VMefgh
18 SS RFfgh SFefgh
19 IC BTgh RFefgh
20 VL SLh ADfgh
21 TR SMhi BTgh
22 SM ADi STh
23 RF
24 LD
25 ST
1: Shackelford et al. (1995); 2: Paterson and Parrish (1986); 3: Jeremiah, Dugan, et al., 2003 and Jeremiah, Gibson, et al., 2003; 4: Carmack et al. (1995); 5:
Jeremiah et al. (1985); 6: Molina et al. (2005); 7: Rhee et al. (2004); 8: Wheeler et al. (2000); 9: Brickler (2000), dry cookery; 10: Brickler (2000), wet
cookery; 11: Wulf and Page (2000).
a
AD = M. adductor; BB = M. biceps brachii; BF = M. biceps femoris; BT = M. brachiocephalicus; CP = M. complexus; DI = Diaphragm; GM = M.
gluteus medius; GR = M. gracillis; IC = intercostal muscles; IF = M. infraspinatus; IP = M. ilio psoas; LD = M. longissmus dorsi; LT = M. latissimus
dorsi; MD = M. multifidus dorsi; OA = M. obliquus abdominus internus; PM = M. psoas major; PP = M. pectoralis profundi; QF = M. quadriceps femoris;
RA = M. rectus abdominis; RB = M. rhomboideus; RF = M. rectus femoris; SD = M. spinalis dorsi; SF = M. superficial pectoral; SL = M. splenius;
SM = M. semimembranosus; SP = M. supraspinatus; SS = M. subscapularis; ST = M. semitendinosus; SV = M. serratus ventralis; TB = M. triceps brachii;
TF = M. tensor faciae latae; TM = M. teres major; TR = M. trapezius; VI = M. vastus intermedius; VL = M. vastus lateralis; VM = M. vastus medialis.
b
Samples are ordered from the most intense beef flavor intensity to the least (bland).
c–i
Means within column without common superscript differ.

pH and heme iron, but pH and heme iron were not related
to specific off-flavor notes.
Flavor desirability has been used by some researchers in
addition to or in lieu of flavor intensity. The Canadian Cat-
tlemen’s Association established a goal of 95% consumer
acceptance of beef, and seven muscles or muscle groups –
M. teres major, M. psoas major, M. longissimus thoracis,
M. longissimus lumborum, M. ilio psoas, M. spinalis dorsi,
and M. subscapularis – fell into that category for flavor
desirability (Jeremiah, Gibson, Aalhus, & Dugan, 2003).
However, the two longissimus muscles were rated as the
second and third lowest for beef flavor intensity, although
the range for flavor intensity was fairly small (5.00–6.07 on
a 9-point scale) in that study. Five other muscles or muscle
groups were approaching 95% desirability for flavor as
well.
McKeith, De Vol, Miles, Bechtel, and Carr (1985) found
the M. psoas major, M. infraspinatus, M. longissimus thora-
cis, M. longissimus lumborum, and the M. rectus femoris to
be rated significantly higher than M. supraspinatus, M.
semimembranosus, M. semitendinosus, M. adductor, and
M. pectoralis profundi for flavor desirability with the M.
biceps femoris, M. gluteus medius, and M. triceps brachii
being similar to the two groups. Similar findings from Wulf
and Page (2000) revealed the M. longissimus dorsi and M.
gluteus medius had the same mean for flavor desirability
(5.73 with 8 = intense) while the M. semimembranosus
was less desirable. Muscle differences in flavor intensity
were not due to glycolytic potential, but less than
80 lmol/g affected the M. longissimus dorsi for flavor desir-
ability while the M. gluteus medius saw a linear increase in
flavor desirability with an increase in glycolytic potential
(Wulf, Emnett, Leheska, & Moeller, 2002), likely due to
ultimate pH. Flavor desirability was highly, negatively cor-
related (P < 0.001) to insoluble collagen in a study that
analyzed 33 muscles or muscle groups from 25 Canada
AA steer carcasses (Jeremiah, Dugan, et al., 2003).
7.2. Differences in flavor compounds among muscles
In a study using purge and trap mass spectrometry, dif-
ferences in volatile compounds between raw muscles from
72 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80

Table 3 dius possessed four compounds that were unique to that

Ranking of musclesa for off-flavor intensity from different studies muscle, with one of the compounds being unknown. The
Rankb 1 2 3 4 relative quantity of piperazine and (Z)-3-octene in the M.
1 LD c
LD c
BT c
CPc vastus intermedius was significantly higher compared to
2 STcd SMd MDcd IFcd the other compounds that make-up the volatile profile of
3 GMcd TBde SScd SLcde the muscle. These compounds may help explain why the
4 QFde BFde IFcd TBcde
5 Mdef STde SLcd RBcde
M. vastus intermedius is perceived as having moderately
6 BFef RFde TBcd PPcde intense flavor (Brickler, 2000; Jones et al., 2004). The M.
7 SPef GMde VMcde MDcde rectus femoris had 10 unique compounds. Additionally,
8 TBefg SPef RFcde BTcdef the M. rectus femoris had higher concentrations of many
9 PMfg ADfg SVcde LTcdef of the compounds in comparison to the other three mus-
10 IFg IFg CPcde BFcdef
11 PMg LTcde GRcdef
cles. The beef myology website reports the M. rectus femo-
12 SPcde SVcdef ris has slightly intense flavor whereas the M. vastus
13 VLcde STcdef lateralis, M. vastus intermedius, and M. triceps brachii have
14 BFcde VLcdef moderately intense flavor (Jones et al., 2004). Apparently,
15 SFcdef SPcdef the additional compounds tone down the M. rectus femoris
16 VIcdef ADcdef
17 STcdef SMcdef
flavor so it is not perceived as intense or off-flavored as
18 RBcdef SSdef other muscles in the chuck and round (Brickler, 2000).
19 GRcdef VIdef Interestingly, mass spectrometry also revealed that the
20 SMdef VMef three muscles from the knuckle have 10 additional com-
21 ADef RFef pounds that are not found in the M. triceps brachii. These
22 PPf SFf
findings may explain why the M. triceps brachii was found
1: Shackelford et al. (1995); 2: Rhee et al. (2004); 3: Brickler (2000), dry to have higher off-flavor ratings than the other three mus-
cookery; 4: Brickler (2000), wet cookery.
a
AD = M. adductor; BB = M. biceps brachii; BF = M. biceps femoris;
cles by Brickler (2000).
BT = M. brachiocephalicus; CP = M. complexus; GM = M. gluteus med- The different combinations of volatiles may help explain
ius; GR = M. gracillis; IF = M. infraspinatus; LD = M. longissmus dorsi; the difference in perceived flavor profile (Hodgen, Cuppett,
LT = M. latissimus dorsi; MD = M. multifidus dorsi; PM = M. psoas et al., 2006). The unique compounds of M. vastus interme-
major; PP = M. pectoralis profundi; QF = M. quadriceps femoris; RB = M. dius create a flavor profile that has higher liver-like, metal-
rhomboideus; RF = M. rectus femoris; SF = M. superficial pectoral;
SM = M. semimembranosus; SL = M. splenius; SP = M. supraspinatus;
lic, and rancid characteristics than the M. triceps brachii,
SS = M. subscapularis; ST = M. semitendinosus; SV = M. serratus ven- M. rectus femoris, and M. vastus lateralis (James & Cal-
tralis; TB = M. triceps brachii; VI = M. vastus intermedius; VL = M. kins, 2005). All the volatile compounds in the M. rectus
vastus lateralis; VM = M. vastus medialis. femoris may help reduce the sour taste and oxidized, ran-
b
Samples are ordered from the lowest off-flavor score to the highest off- cid, and charred aromas (James & Calkins, 2005) when
flavor score.
c–g
Means within column without common superscript differ.
compared to the other muscles in this study. While the
M. vastus lateralis only had two different compounds from
the other muscles, the combination of these compounds
the chuck and the round were evaluated (Hodgen, Cuppett, helped reduce the perceived liver-like characteristic in these
& Calkins, 2006). Twenty-eight volatile compounds were samples. Meisinger et al. (2006) found no difference in
identified in all four normal muscles (Table 5). The M. tri- liver-like between the M. rectus femoris, M. triceps brachii,
ceps brachii did not have any compounds unique from the and M. vastus lateralis.
muscles in the knuckle. The M. vastus lateralis had two Cow muscles have very different flavor characteristics
unique aromatic compounds in its volatile profile. Both compared to meat from A-maturity steer carcasses. A
of these compounds were in low concentrations relative benchmarking study compared sensory properties of mus-
to other compounds in the sample. The M. vastus interme- cles from Select, A-maturity carcasses to muscles from four

Table 4
The effect of muscle on percentage of panelists detecting each off-flavor note
Muscle Liver Sour Metallic Charred Bloody Oxidized Fatty Rancid
Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE Mean SE
a a b ab b
Infra-spinatus 9.3 2.9 23.2 3.7 8.7 2.2 29.9 4.4 1.6 1.0 9.5 2.3 14.0 1.3 8.8 1.6
Rectus femoris 9.7 2.9 44.2b 3.7 13.4a 2.2 20.4ab 4.4 3.4 1.0 7.4a 2.3 3.2a 1.3 4.9 1.6
Teres major 8.8 2.9 48.7b 3.7 15.5ab 2.2 21.6ab 4.4 1.8 1.0 8.5ab 2.3 3.3a 1.3 5.8 1.6
Triceps brachii 7.7 2.9 49.5b 3.7 19.5b 2.2 22.2ab 4.4 0.8 1.0 13.3abc 2.3 1.6a 1.3 5.6 1.6
Vastus lateralis 9.1 2.9 48.4b 3.7 15.0ab 2.2 30.5b 4.4 1.3 1.0 17.5c 2.3 1.4a 1.3 6.8 1.6
Vastus medialis 10.8 3.0 49.0b 3.8 17.3ab 2.2 14.8a 4.6 2.9 1.0 14.6bc 2.3 2.3a 1.4 7.2 1.6
Taken from Meisinger et al. (2006).
a,b,c
Means within a column (for off-flavor notes) with different superscripts are significantly (P < 0.05) different.
 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80 73

Table 5
Compound concentration differences for between normal-flavored beef musclesa determined by gas chromatography–mass spectrometry analysis
Retention timeb Compoundc,d TRI REC VAL VAI
1.93 UNK · · ·
2.03 UNK · · ·
2.48 Oxirane, 2,3-dimethyl- · · · ›
2.81 UNK ·
3.07 1,2-Ethanediamine, N,N 0 -dimethyl · ·
3.09 Pentanal · ·
3.12 sec-Butanamine ·
3.15 UNK ·
3.17 Cyclobutanol ·
3.20 UNK · ·
3.45 2-Heptanamine, 5-methyl ·
3.50 UNK · ·
3.65 Pentanal ·
3.71 Pentanal › · · ·
4.41 Piperazine ·
4.77 Cyclobutanol · ·
4.88 UNK · ·
5.29 1-Pentanol · › · ·
5.56 1-Pentanol, 5-amino ·
5.82 UNK · ·
5.86 UNK · ·
6.11 Hexanal · › fl ·
6.27 3-Octene, (Z)- ·
6.44 1,3-Propanediamine, N-methyl or hexanal ·
6.49 2-Heptanamine, 5-methyl · ·
7.60 UNK · ·
8.05 1-Hexanol › fl › fl
8.18 UNK ·
8.25 1-Hexanol ·
8.62 2-Heptanone · · · ·
8.88 Heptanal fl › · ·
9.72 a-Pinene · · · ·
10.32 2-Heptenal · › · ·
10.43 Benzaldehyde · · · ·
10.60 2-Hexen-1-ol, (E)- ·
10.67 1-Heptanol fl › · fl
10.83 b-Pinene · · ·
10.90 1-Octen-3-ol · · · fl
11.01 n-Caprioc acid vinyl ester or 3-octanone, 2-methyl · › · ·
11.19 Furan, 2-pentyl fl · › fl
11.47 Octanal fl › · fl
11.66 3-Carene · ·
12.10 1-Hexanol, 2-ethyl- · · · ·
12.20 1,3-Hexadiene, 3-ethyl-2-methyl- · ·
12.35 3-Octen-2-one ·
12.49 UNK · · · ·
12.77 2-Octenal, (E)- · › · ·
13.00 2-Octen-1-ol or 1,3-octadiene · · · ·
13.05 1-Octanol fl › · fl
13.41 UNK · ·
13.52 2-Nonanone ·
13.63 Octadecanal ·
13.80 Nonanal fl › · ·
14.28 UNK ·
14.83 3-Hydroxymandelic acid, ethyl ester, di-TMS or benzaldehyde, 2,4-bis(trimethylsiloxy)- · · · ·
14.99 2-Nonenal or 1-tetradecene · › · ·
15.13 UNK · ·
15.69 Undec-4-enal ·
15.92 Decanal · · · ·
16.11 2,4-Nonadienal ·
16.93 Benzene, 1,3-bis(1,1-dimethylethyl)- ·
17.03 2-Decenal · › · ·
17.39 UNK · ·
(continued on next page)
74 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80

Table 5 (continued)
Retention timeb Compoundc,d TRI REC VAL VAI
17.66 2,4-Decadienal · ·
17.71 Tridecane · · · ·
18.09 2,4-Decadienal · · ·
18.19 Malonic acid, bis(2-trimethylsilylethyl ester) fl › · ›
18.93 2-Tridecenal · · ·
19.51 Tetradecane · ·
19.71 Oxirane, hexadecyl or UNK · · · ·
21.20 UNK · · ·
21.43 UNK · · ·
Taken from Hodgen (2006).
a
TRI = M. triceps brachii, REC = M. rectus femoris, VAL = M. vastus lateralis, and VAI = M. vastus intermedius.
b
Retention time was obtained from GC–MS report for each sample.
c
Compounds listed were matches found in the mass spectrometry database.
d
› indicates that a higher concentration of the compound was found in that type of sample; fl indicates that a lower concentration of the compound was
found in that type of sample; · indicates the compound was present and/or fell in between higher and lower than ones listed with an arrow.

cow populations (beef-fed, beef-nonfed, dairy-fed, dairy- 8. Liver-like off-flavor in beef

nonfed) (Stelzleni, Johnson, Calkins, Mink, & Gwartney,
2005). All muscles from cows had more intense beef flavor
and greater off-flavor than muscles from younger, A-matu-
rity Select-grade beef carcasses.
In a study conducted in our laboratory (Meisinger et al.,
2006), seven muscles from the chuck and knuckle were
obtained from 30 animals, and animal identity was main-
tained. Three animals had the majority of their muscles
ranked 5 or below on an 8 point hedonic scale for off-flavor
intensity (Table 6) which meant the off-flavor was
approaching or below what average consumers would find
acceptable. Theses data suggest that when an animal has
one muscle in the chuck or knuckle that has off-flavor,
other muscles in the chuck or knuckle are likely to be off-
flavored as well. This suggests there is an ‘animal effect’
to off-flavors found in the chuck and knuckle, meaning
the animal has been exposed to something different than
a normal animal. After investigating several angles, our
laboratory’s current hypothesis is the difference is caused
at least partially by the diet of the animal.
This ‘animal effect’ on flavor leads to other questions
that need to be addressed. Do all muscles in a candidate
animal for off-flavor possess off-flavor? Is it just locomo-
tion muscles or muscles that lay on or near a bone? Most
importantly, what is causing these undesirable flavors
and how can producers, processors, or individuals prepar-
ing the meals mitigate or eliminate the effect?
With the increased use of muscles from the chuck and
round in the US for steak items, rather than roasts or
ground beef, restaurant personnel and foodservice employ-
ees that use these steaks have reported an increased number
of complaints regarding the flavor of the beef they serve.
The majority of the complaints are referred to as liver-like
flavor. Because of the economic impact these steaks have
on the beef industry, it is important to identify and under-
stand the causes of this particular off-flavor.
8.1. Cooking rate and holding time
Most of the anecdotal reports of incidences of off-flavors
our laboratory received were from foodservice personnel. It
was hypothesized that since foodservice preparation tradi-
tionally cooked the meat quickly and then held the product
in warming ovens until the food was presented to the con-
sumer these conditions might promote the liver-like flavor.
Seven muscles located in the chuck and knuckle were
slowly or rapidly cooked (on a 149EC commercial gas grill
or a 249EC grill) and held for 0 or 1 h prior to consump-
tion. Results clearly demonstrated that cooking at a slower
rate and holding for a longer time reduce the intensity of
off-flavor. It is hypothesized the slower cooking and longer
hold time allow the undesirable volatile compounds to dis-
sipate (James & Calkins, 2005).

Table 6
Mean off-flavor intensity scores and the percentage of panelist’s who recognized livery-like flavors
Animal Off-flavor intensity scores Percentage of panelist’s who recognized livery-like flavor
Infra- Rectus Teres Triceps Vastus Vastus Infra- Rectus Teres Triceps Vastus Vastus
spinatus femoris major brachii lateralis medialis spinatus femoris major brachii lateralis medialis
6 3.78 3.78 1.60 2.22 3.00 2.30 66.70 44.40 30.00 44.40 62.50 60.00
7 4.88 3.38 2.67 4.00 3.57 4.60 12.50 50.00 66.70 11.10 42.90 60.00
8 3.88 4.88 4.90 4.50 3.57 3.90 100.00 50.00 50.00 37.50 37.50 60.00
Adapted from Meisinger et al. (2006).
 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
During this study the observation was made that the
samples that would be rated as having intense off-flavor
by the trained taste panelists could be smelled during the
cooking process. With the knowledge that there appears
to be an animal effect and that the off-flavor is at least
somewhat volatile, our laboratory has taken two directions
in determining the cause of this off-flavor – pre-harvest
conditions and compounds differences between lean tissue
that is normal and lean tissue that is liver-like in flavor.
8.2. Frequency
The frequency of this off-flavor appears to relatively
low, although the initial speculation was the incidence
was between 7% and 10% based on Yancey et al. (2006)
and Meisinger et al. (2006). In a survey of urban Nebraska
residents, 6.9% (11 out of 159) of respondents who had
eaten a M. infraspinatus steak stated they had a poor eating
experience due to flavor.
8.3. Sniff test
To develop a sampling method for off-flavored beef
samples, 10 g slices of the knuckle were sliced off on the
fabrication line in a packing plant, cooked immediately
in a room with air purifiers and smelled and/or tasted to
determine if the sample was off-flavored. Samples that were
thought to be off-flavored by this test were taken to the
meat laboratory at the University of Nebraska and reeval-
uated by the panelists that had performed the evaluation at
the plant. Contingency tables revealed the panelists were in
close agreement, but there were few samples with the liver
flavor (Jenschke, Hamling, et al., 2006). Further work is
needed to confirm the usefulness of this method to study
liver-like off-flavors.
8.4. Feedlots and pre-harvest conditions
A total of 328 knuckles, which came from five different
feedlots, were tested by the two panelists. Utilizing smelling
and tasting, feedlot A had 5.0% off-flavored samples (1 of
20), feedlot B had 9.4% (5 of 53), feedlot C had 6.7% (10
of 148), feedlot D had 12.5% (9 of 72) and feedlot E had
14.3% (5 of 35) (Jenschke, Hamling, et al., 2006).
No relationships between any of the pre-harvest traits or
liver scores and the prevalence of off-flavors existed. There
were also no trends in muscles having liver-like flavor and
medical treatment or feed additives received (data not
shown). Since distribution of off-flavors varied among feed-
lots, future research should investigate pre-harvest factors
that might lead to off-flavor development.
8.5. Minerals and fatty acids
The liver-like flavor in beef is often associated with an
increase in metallic flavor (Miller, 2001) and increased
75
cooler aging (Campo, Sanudo, Panea, & Alberti, 1999).
Our laboratory investigated the relationship of eight miner-
als and fatty acids to the liver-like off-flavor. Sodium, 16:1,
18:1n7 (vaccenic acid), 20:2n6, and 20:3 accounted for 46%
of the variation between liver-like samples and controls.
Sodium relationship was low in this study, but we hypoth-
esize sodium may help accentuate the liver-like off-flavor
and that is why it fell into the model. Iron has often been
thought to play a role in the liver-like off-flavor (Yancey
et al., 2006) but this mineral did not show up in any of
our analyses. Our fatty acid analysis revealed similar
results to others (Camfield et al., 1997; Yancey et al.,
2006) with medium and long chain unsaturated fatty acids
playing a role in creating the liver-like off-flavor. Regres-
sion models of unsaturated fatty acids alone accounted
for 40% of the variation in liver-like off-flavor. Further
understanding of manipulation of unsaturated fatty acids
in muscle tissue could prove beneficial in reducing the inci-
dence of the liver-like off-flavor. Current research is being
undertaken to consider a more complete picture of the fatty
acid profiles between off-flavored and normal samples. Sul-
fur level comparisons between the liver-like and the normal
samples need to be determined since this mineral could cre-
ate undesirable flavors.
8.6. Gas chromatography–mass spectrometry
With the evidence the liver-like off-flavor is partially due
to a volatile compound, gas chromatography and mass
spectrometry work was conducted to evaluate the differ-
ences between compounds in liver-like beef samples and
normal-flavored samples. Raw meat samples (5 g) were
mixed with 100 mL of double distilled water in a closed
container with an inlet to push nitrogen through and an
outlet to smell aromas emitted from the sample when
heated in a 40 C water bath (Table 8; Hodgen, Calkins,
& Cuppett, 2006). Undesirable aromas were emitted only
for the samples that were ranked by taste panelists as being
liver-like while normal beef aromas were emitted from the
control samples. When these compounds were collected
and injected into the gas chromatograph, differences in
chromatograms were observed between the normal and
liver-like samples.
The raw, pulverized samples were run on a purge and
trap mass spectrometer (Hodgen, Cuppett, et al., 2006).
Because muscle tissue heated only to 40 C showed differ-
ences in volatile composition between normal and liver-like
samples, it was decided to eliminate much of the effect of
the cooking process by using raw tissue samples to reduce
the amount of compounds produced. This would reduce
the number of compounds produced by the Maillard reac-
tion, but compounds that help give rise to the liver-like fla-
vor would still be present. Differences in the presence of
numerous compounds were apparent between normal and
liver-like samples. Most of the compounds present in
higher concentration in the liver-like samples, like penta-
76 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80

nol, hexanal, hexanol, 1-octen-3-ol, and nonanal, are asso-
ciated with lipid oxidation (Table 7). The compounds b-
pinene, 1-octen-3-ol, and 2,4-decadienal were in higher
concentration in the liver-like samples in all muscles tested.
Several of these same compounds were implicated by Yan-
cey et al. (2006) when they used cooked samples for evalu-
ation between liver-like and normal samples. Interestingly,
these compounds (especially 2,4-decadienal) are associated
with the offensive odor of Eocanthecona furcellata (Wolff) –
commonly known as the stink bug (Ho, Kou, & Tseng,
2003). Limonene, a citrus aroma, was observed in higher
concentrations in the liver-like samples in their study and
in later studies in our laboratory (Hodgen, 2006). In the
original study (Hodgen, Cuppett, et al., 2006), the com-
pound with the same retention time as limonene was iden-
tified by the mass spectrometer database as a hexanol
isomer so it is likely this compound also contributes to
the liver-like aromatic profile in samples.
A larger pool of samples are needed to further investi-
gate the volatile make-up of liver-like samples. Further
investigation into oxidation also needs to be addressed in
regards to liver-like flavor. Both Hodgen, Cuppett, et al.
(2006) and Yancey et al. (2006) found compounds associ-
ated with lipid oxidation, yet thiobarbituric acid assays
found no relationship (Yancey et al., 2006) between normal
and liver-like samples.
8.7. A candidate cause
While the liver-like off-flavor has been around for many
years, there is speculation the increase in incidence of liver-
like samples may be attributed to changes in feeding prac-
tices in the US. Some evidence points to increased levels of
unsaturated fatty acids in the muscle in liver-like samples,
so feedstuffs that have higher levels of unsaturation have
been suspect. The biohydrogenation process in ruminant
animals minimizes passage of unsaturated fatty acids
through the digestive system. One possible candidate out
laboratory has investigated is wet distillers grain plus solu-
bles (WDGS) since the increased use of distillers grain
seemed to coincide with the increased complaints about
liver-like flavor in meat. Feedlots receiving this by-product
of ethanol production also have to ensure sulfur levels in
the animal feed will not reach undesirable levels as the eth-
anol processing method uses sulfur. Sensory results in meat
from a feeding trial in which steers were fed 0%, 10%, 20%,
30%, 40%, and 50% levels of WDGS did not show any sig-
nificant relationship (P = 0.07) to the perception of liver-
like off-flavor in the M. rectus femoris although it was
approaching significance (Table 8; Jenschke, Hodgen, Van-
der Pol, Calkins, & Klopfenstein, 2006). However, this
work was conducted on meat aged 7 d. Storage (aging) of
meat would allow oxidation to occur.

Table 7
Compound concentration differences between the liver-like and normal-flavored beef muscles determined by gas chromatography–more spectrometry
analysis
Compounda,b TRIc RECc VALc VAIc
Liver-like Normal Liver-like Normal Liver-like Normal Liver-like Normal
2,3-Dimethyl oxirane › fl
Pentanal › fl › fl
Heptanol › fl
Hexanal › fl
Hexanol
2-Heptanone › fl
Heptanal › fl
Benzaldehyde › fl
b-Pinene › fl › fl › NP › fl
1-Octen-3-ol › fl › fl › fl › fl
2-Methyl-3-octanone or n-caproic acid, vinyl ester › fl › fl › fl
2-Pentyl furan › fl › fl
Octanal › fl › fl
a-Pinene › fl › fl
2-Octenal › fl › fl
1-Octanol › fl › fl
Nonanal › fl › fl › fl
Hydroxymandelic acid › fl
2-Nonenal › fl
1,3-Bis(1,1-dimethylethyl) benzene › fl
2,4-Decadienal › NP › fl › fl › fl
2,4-Decadienal › NP › fl › fl › fl
Taken from Hodgen (2006).
a
Compounds listed revealed visual concentration differences in chromatograms.
b
› indicates that a higher concentration of the compound was found in that type of sample; fl indicates that a lower concentration of the compound was
found in that type of sample; NP = not present.
c
TRI = M. triceps brachii, REC = M. rectus femoris, VAL = M. vastus lateralis, and VAI = M. vastus intermedius.
 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80 77

Table 8
Least squares means for main effects for off-flavor intensity, liver-like and metallic off-flavors
Effect Off-flavor intensitya Liver-likeb Metallicb
Treatmentc 0.47d 0.07d 0.73d
0 5.72 14.44 34.07
10 5.49 19.63 27.41
20 5.69 11.85 31.85
30 5.74 7.41 31.85
40 5.54 12.22 34.81
50 5.73 8.52 36.30
Standard error of the mean 0.11 3.07 4.17
Quality grade 0.002d 0.03d <0.001d
Choice 5.51e 15.19f 39.26f
Select 5.80f 9.51e 26.17e
Standard error of the mean 0.07 1.77 2.41
Adapted from Jenschke, Hodgen, et al. (2006).
a
Off-flavor intensity: 1 = extreme off-flavor; 8 = no off-flavor.
b
Off-flavors are expressed as a percentage of panelists that identified the off-flavor.
c
Treatments: percentage of wet distillers grains plus solubles included in diet.
d
P-value from analysis of variance tables.
e,f
Mean values within a column and followed by the same letter are not significantly different (P > 0.05).

Further research has been conducted in our laboratory
to investigate the relationship of flavor to meat proper-
ties from steers fed WDGS. Increases in polyunsaturated
fatty acids were seen with an increase in WDGS levels in
the diet (4.90% of fatty acid methyl esters versus 5.91%
and 6.23% of fatty acid methyl esters for cattle fed
15% or 30% WDGS on a dry matter basis). Our results
(unpublished) also show an increase in oxidation and a
decrease in color stability when WDGS are fed. Simi-
larly, Roeber, Gill, and DiCostanzo (2005) found some
reduction in color stability upon storage for beef from
cattle fed distillers grains.
Approaching the cause of liver-like off-flavors in beef
chuck and round muscles from a pre- and post-harvest
view has increased our knowledge about the liver-like fla-
vor, but further research is still needed to be able to prevent
the occurrence of this undesirable flavor from occurring.
9. Conclusions
Development of the flavor of meat is a very complex sys-
tem. Factors that make some meat desirable and other
meat undesirable is influenced by many conditions, includ-
ing pre-harvest animal environment and diet, post-harvest
handling, and consumers’ individual preferences. The first
two factors can be investigated and solutions can be deter-
mined; the latter is a much harder variable to control since
a person’s internal and external influences help develop
personal preferences. The role of the meat industry should
be to try to produce the highest quality product for the
market they are targeting.
References
Artz, W. E., Perkins, E. G., & Salvador-Henson, L. (1993). Character-
ization of the volatile decomposition products of oxidized methyl
arachidonate. Journal of the American Oil Chemists’ Society, 70(4),
377–382.
Batifoulier, F., Mercier, Y., Gatellier, P., & Renerre, M. (2002). Influence
of vitamin E on lipid and protein oxidation induced by H2O2-activated
MetMb in microsomal membranes from turkey muscle. Meat Science,
61(4), 389–395.
Berry, B. W., Leddy, K. F., Bond, J., Rumsey, T. S., & Hammond, A. C.
(1988). Effects of silage diets and electrical stimulation on the
palatability, cooking and pH characteristics of beef loin steaks.
Journal of Animal Science, 66(4), 892–900.
Bidner, T. D., Montgomery, R. E., Bagley, C. P., & McMillin, K. W.
(1985). Influence of electrical stimulation, blade tenderization and
postmortem vacuum aging upon the acceptability of beef finished on
forage or grain. Journal of Animal Science, 61(3), 584–589.
Blank, I., Lin, J., Vera, F. A., Welti, D. H., & Fay, L. B. (2001).
Identification of potent odorants formed by autoxidation of arachi-
donic acid: Structure elucidation and synthesis of (E,Z,Z)-2,4,7-
tridecatrienal. Journal of Agricultural and Food Chemistry, 49(6),
2959–2965.
Boleman, S. J., Miller, R. K., Buyck, M. J., Cross, H. R., & Savell, J. W.
(1996). Influence of realimentation of mature cows on maturity, color,
collagen solubility, and sensory characteristics. Journal of Animal
Science, 74(9), 2187–2194.
Bowling, R. A., Riggs, J. K., Smith, G. C., Carpenter, Z. L., Reddish, R.
L., & Butler, O. D. (1978). Production, carcass and palatability
characteristics of steers produced by different management systems.
Journal of Animal Science, 46(2), 333–340.
Brickler, J. E. (2000). The effect of carcass weight, yield grade, quality
grade, and cooking method on physical attributes, Warner-Bratzler
Shear Force, and sensory panel characteristics on muscles of the beef
chuck and round. Masters thesis, University of Florida, Gainesville,
FL.
Brown, H. G., Melton, S. L., Riemann, M. J., & Backus, W. R. (1979).
Effects of energy intake and feed source on chemical changes and
flavor of ground beef during frozen storage. Journal of Animal Science,
48(2), 338–347.
Camfield, P. K., Brown, A. H., Jr., Lewis, P. K., Rakes, L. Y., & Johnson,
Z. B. (1997). Effects of frame size and time-on-feed on carcass
characteristics, sensory attributes, and fatty acid profiles of steers.
Journal of Animal Science, 75(7), 1837–1844.
Campo, M. M., Sanudo, C., Panea, B., & Alberti, P. (1999). Breed type
and ageing time effects on sensory characteristics of beef strip loin
steaks. Meat Science, 51(4), 383–390.
78 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
Carmack, C. F., Kastner, C. L., Dikeman, M. E., Schwenke, J. R., &
Garcia Zepeda, C. M. (1995). Sensory evaluation of beef flavor-
intensity, tenderness, and juiciness among major muscles. Meat
Science, 39(1), 143–147.
Cornu, A., Kondjoyan, N., Frencia, J. P., & Berdague, J. L. (2001).
Deciphering the message from volatile components of fat tissues.
Viandes and Product Carnes, 22(2), 35–38.
Dannenberger, D., Lorenz, S., Nuernberg, G., Scollan, N., Ender, K., &
Nuernberg, K. (2006). Analysis of fatty aldehydes composition,
including 12-methyltridecanal, in plasmalogens from longissimus
muscle of concentrate- and pasture-fed bulls. Journal of Agricultural
and Food Chemistry, 54(1), 182–188.
Dannenberger, D., Nuernberg, G., Scollan, N., Schabbel, W., Steinhart,
H., Ender, K., et al. (2004). Effect of diet on the deposition of n-3 fatty
acids, conjugated linoleic and c18:1 trans fatty acid isomers in muscle
lipids of German Holstein bulls. Journal of Aquatic Food Product
Technology, 52(21), 6607–6615.
Dannenberger, D., Nuernberg, K., Nuernberg, G., Scollan, N., Steinhart,
H., & Ender, K. (2005). Effect of pasture vs. Concentrate diet on CLA
isomer distribution in different tissue lipids of beef cattle. Lipids, 40(6),
589–598.
Descalzo, A. M., Garcia, P. T., Pensel, N. A., Josifovich, J. A., Insani, E.
M., Biolatto, A., et al. (2005). Influence of pasture or grain-based diets
supplemented with vitamin E on antioxidant/oxidative balance of
Argentine beef. Meat Science, 70(1), 35–44.
Drumm, T. D., & Spanier, A. M. (1991). Changes in the content of lipid
autoxidation and sulfur-containing compounds in cooked beef during
storage. Journal of Agricultural and Food Chemistry, 39(2), 336–343.
Dube, G., Bramblett, V. D., Howard, R. D., Homler, B. E., Johnson, H.
R., Harrington, R. B., et al. (1971). Dietary effects on beef compo-
sition. 4. Processing and palatability attributes. Journal of Food
Science, 36(1), 147–154.
Elmore, J. S., Campo, M. M., Enser, M., & Mottram, D. S. (2002). Effect
of lipid composition on meat-like model systems containing cysteine,
ribose, and polyunsaturated fatty acids. Journal of Agricultural and
Food Chemistry, 50(5), 1126–1132.
Elmore, J. S., Mottram, D. S., Enser, M., & Wood, J. D. (1999). Effect of
the polyunsaturated fatty acid composition of beef muscle on the
profile of aroma volatiles. Journal of Agricultural and Food Chemistry,
47(4), 1619–1625.
Elmore, J. S., Warren, H. E., Mottram, D. S., Scollan, N. D., Enser, M.,
Richardson, R. I., et al. (2004). A comparison of the aroma volatiles
and fatty acid compositions of grilled beef muscle from Aberdeen
angus and Holstein-Friesian steers fed diets based on silage or
concentrates. Meat Science, 68(1), 27–33.
Enser, M., Hallett, K. G., Hewitt, B., Fursey, A. J., Wood, J. D., &
Harrington, G. (1998). Fatty acid content and composition of UK beef
and lamb muscle in relation to production system and implications for
human nutrition. Meat Science, 49(3), 329–341.
Farmer, L. J., & Patterson, R. L. S. (1991). Compounds contributing to
meat flavour. Food Chemistry, 40(2), 201–205.
Fay, L. B., & Brevard, H. (2005). Contribution of mass spectrometry to
the study of the Maillard reaction in food. Mass Spectrometry Review,
24(4), 487–507.
Francis, J. J., Romans, J. R., & Norton, H. W. (1977). Consumer rating of
two beef marbling levels. Journal of Animal Science, 45(1), 67–70.
French, P., O’Riordan, E. G., Monahan, F. J., Caffrey, P. J., Mooney, M.
T., Troy, D. J., et al. (2001). The eating quality of meat of steers fed
grass and/or concentrates. Meat Science, 57(4), 379–386.
French, P., O’Riordan, E. G. O., Monahan, F. J., Caffrey, P. J., Vidal, M.,
Mooney, M. T., et al. (2000). Meat quality of steers finished on
autumn grass, grass silage or concentrate-based diets. Meat Science,
56(2), 173–180.
French, P., Stanton, C., Lawless, F., O’Riordan, E. G., Monahan, F. J.,
Caffrey, P. J., et al. (2000). Fatty acid composition, including
conjugated linoleic acid, of intramuscular fat from steers offered
grazed grass, grass silage, or concentrate-based diets. Journal of Animal
Science, 78(11), 2849–2855.
Gasser, U., & Grosch, W. (1988). Identification of volatile flavour
compounds with high aroma values from cooked beef. Zeitschrift fuer
Lebensmittel-Untersuchung und-Forschung A, 186(6), 489–494.
Gatellier, P., Mercier, Y., Juin, H., & Renerre, M. (2005). Effect of
finishing mode (pasture- or mixed-diet) on lipid composition, colour
stability and lipid oxidation in meat from Charlois cattle. Meat
Science, 69(5), 175–186.
Goodson, K. J., Morgan, W. W., Reagan, J. O., Gwartney, B. L.,
Courington, S. M., Wise, J. W., et al. (2002). Beef customer
satisfaction: Factors affecting consumer evaluations of clod steaks.
Journal of Animal Science, 80(2), 401–408.
Guth, H., & Grosch, W. (1993). 12-Methyltridecanal, a species-specific
odorant of stewed beef. Lebensmittal-Wissenschaft und Technologie,
26(2), 171–177.
Ho, H. Y., Kou, R., & Tseng, H. K. (2003). Semiochemicals from the
predatory stink bug Eocantheocona furcella (Wolff): Components of
metathoracic gland, dorsal abdominal gland, and sternal gland
secretions. Journal of Chemical Ecology, 29(9), 2101–2114.
Ho, C. T., Oh, Y. C., & Bae-Lee, M. (1994). The flavour of pork. In F.
Shahidi (Ed.), Flavor of meat and meat products (2nd ed., pp. 38–51).
London: Blackie Academic and Professional.
Hodgen, J. M. (2006). Factors influencing off-flavor in beef. PhD
dissertation. University of Nebraska, Lincoln, NE.
Hodgen, J. M., Calkins, C. R., & Cuppett, S. L. (2006). Protocol for
determining volatile compound differences between liver-like and
normal beef samples using gas chromatography. In Proceedings of
the American society of animal science midwestern annual meeting, Des
Moines, IA (Abstr.).
Hodgen, J. M., Cuppett, S. L., & Calkins, C. R. (2006). Identification of
off-flavor compounds in beef. In Proceedings of the American meat
science association reciprocal meat conference, Champagne-Urbana, IL.
Huffman, K. L., Miller, M. F., Hoover, L. C., Wu, C. K., Brittin, H. C., &
Ramsey, C. B. (1996). Effect of beef tenderness on consumer
satisfaction with steaks consumed in the home and restaurant. Journal
of Animal Science, 74(1), 91–97.
James, J. M., & Calkins, C. R. (2005). Cooking rate and holding time
effect on the flavor of muscles from the beef chuck and the round. In
Proceedings of the international congress of meat science and technol-
ogy, Baltimore, MD.
Jenschke, B. E., Eskridge, K. M., Calkins, C. R. (2006). Factors associated
with liver-like off flavor in beef. In Proceedings of the American meat
science association reciprocal meat conference, Champagne-Urbana, IL.
Jenschke, B. E., Hamling, A. E., Hodgen, J. M., Meisinger, J. L., Moss, D.
A., Lundesjö Ahnström, M., et al. (2006). A novel procedure to
identify off-flavored beef steaks. Journal of Animal Science, 84(Suppl.
2), 170 (Abstr.).
Jenschke, B. E., Hodgen, J. M., Vander Pol, K. J., Calkins, C. R., &
Klopfenstein, T. J. (2006). Wet distillers grains plus solubles do not
increase liver-like off-flavors in cooked beef. 2006 Nebraska beef
reports, Lincoln, NE.
Jeremiah, L. E., Beauchemin, K. A., Jones, S. D. M., Gibson, L. L., &
Rode, L. M. (1998). The influence of dietary cereal grain source and
feed enzymes on the cooking properties and palatability attributes of
beef. Canadian Journal of Animal Science, 78, 271–275.
Jeremiah, L. E., Dugan, M. E., Aalhus, J. L., & Gibson, L. L. (2003).
Assessment of the relationship between chemical components and
palatability of major beef muscles and muscle groups. Meat Science,
65(3), 1013–1019.
Jeremiah, L. E., Gibson, L. L., Aalhus, J. L., & Dugan, M. E. (2003).
Assessment of palatability attributes of the major beef muscles. Meat
Science, 65(3), 949–958.
Jeremiah, L. E., Martin, A. H., & Murray, A. C. (1985). The effects of
various post-mortem treatments on certain physical and sensory
properties of three different bovine muscles. Meat Science, 12(3),
155–176.
Jones, S. J., Calkins, C. R., Johnson, D. D., Carpenter, B., Guru, A., &
Singh, V. (2004). Beef myology and muscle profiling. http://bovi-
ne.unl.edu/.
 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
Kanner, J., & Harel, S. (1985). Initiation of membranal lipid peroxidation
by activated metmyoglobin and methemoglobin. Archives of Biochem-
istry and Biophysics, 237, 314–321.
Killinger, K. M., Calkins, C. R., Umberger, W. J., Feuz, D. M., &
Eskridge, K. M. (2004). A comparison of consumer sensory acceptance
and value of domestic beef steaks and steaks from a branded, argentine
beef program. Journal of Animal Science, 82(11), 3302–3307.
Ladikos, D., & Lougovois, V. (1990). Lipid oxidation in muscle foods: A
review. Food Chemistry, 35(4), 295–314.
Larick, D. K., Hedrick, H. B., Bailey, M. E., Williams, J. E., Hancock, D. L.,
Garner, G. B., et al. (1987). Flavor constituents of beef as influenced by
forage- and grain-feeding. Journal of Food Science, 52(2), 245–251.
Lee, M. R. F., Winters, A. L., Scollan, N. D., Dewhurst, R. J.,
Theodorou, M. K., & Minchin, F. R. (2004). Plant-mediated lipolysis
and proteolysis in red clover with different polyphenol oxidase
activities. Journal of the Science of Food and Agriculture, 84(13),
1639–1645.
Listrat, A., Rakadjiyski, N., Jurie, C., Picard, B., Touraille, C., & Geay,
Y. (1999). Effect of the type of diet on muscle characteristics and meat
palatability of growing salers bulls. Meat Science, 53(2), 115–124.
Lorenzen, C. L., Neely, T. R., Miller, R. K., Tatum, J. D., Wise, J. W.,
Taylor, J. F., et al. (1999). Beef customer satisfaction: Cooking
method and degree of doneness effects on the top loin steak. Journal of
Animal Science, 77(3), 637–644.
MacLeod, G. (1986). The scientific and technological basis of meat
flavours. In G. G. Birch & M. G. Lindley (Eds.), Developments in food
flavors (pp. 191–223). New York: Elsevier Applied Science.
MacLeod, G. (1994). The flavor of beef. In F. Shahidi (Ed.), Flavor of
meat and meat products (pp. 4–37). Glasgow, UK: Blackie Academic
and Professional.
Mandell, I. B., Buchanan-Smith, J. G., & Campbell, C. P. (1998). Effects
of forage vs grain feeding on carcass characteristics, fatty acid
composition, and beef quality in limousin-cross steers when time on
feed is controlled. Journal of Animal Science, 76(10), 2619–2630.
Mandell, I. B., Buchanan-Smith, J. G., Holub, B. J., & Campbell, C. P.
(1997). Effects of fish meal in beef cattle diets on growth performance,
carcass characteristics, and fatty acid composition of longissimus
muscle. Journal of Animal Science, 75(4), 910–919.
Manley, C. H., & Choudhury, B. H. (1999). Thermal process flavorings. In
P. R. Ashurst (Ed.), Food flavorings (3rd ed.) (pp. 283–326). Boston:
Kluwer Academic Publisher.
McKeith, F. K., De Vol, D. L., Miles, R. S., Bechtel, P. J., & Carr, T. R.
(1985). Chemical and sensory properties of thirteen major beef
muscles. Journal of Food Science, 50(4), 869–872.
Meisinger, J. L., James, J. M., & Calkins, C. R. (2006). Flavor
relationships among muscles from the beef chuck and round. Journal
of Animal Science, 84(10), 2826–2833.
Melton, S. L. (1983). Effect of forage feeding on beef flavor. Food
Technology, 37(5), 239–248.
Melton, S. L. (1990). Effects of feeds on flavor of red meat: A review.
Journal of Animal Science, 68(12), 4421–4435.
Melton, S. L., Black, J. M., Davis, G. W., & Backus, W. R. (1982). Flavor
and selected chemical components of ground beef from steers
backgrounded on pasture and fed corn up to 140 days. Journal of
Food Science, 47(3), 699–704.
Miller, R. K. (2001). Beef flavor: A white paper. Centennial, CO: National
Cattlemens’ Beef Association.
Miller, R. K., Moeller, S. J., Goodwin, R. N., Lorenzen, C. L., & Savell, J.
W. (2000). Consistency in meat quality. In Proceedings of the 46th
international congress of meat science & technology, Buenos Aires,
Argentina (pp. 566–580).
Miller, R. K., Rockwell, L. C., Lunt, D. K., & Carstens, G. E. (1996).
Determination of the flavor attributes of cooked beef from cross-bred
angus steers fed corn- or barley-based diets. Meat Science, 44(4),
235–243.
Molina, M. E., Johnson, D. D., West, R. L., & Gwartney, B. L. (2005).
Enhancing palatability traits in beef chuck muscles. Meat Science,
71(1), 52–61.
79
Moody, W. G. (1983). Beef flavor – A review. Food Technology, 37(5),
227–232, 238.
Mottram, D. S., & Edwards, R. A. (1983). The role of triglycerides and
phospholipids in the aroma of cooked beef. Journal of the Science of
Food and Agriculture, 34(5), 517–522.
Mottram, D. S., & Madruga, M. S. (1994). Important sulfur containing
aroma volatiles in meat. In C. J. Mussinan & M. E. Keelan (Eds.),
Sulfur compounds in foods. ACS symposium series (Vol. 564,
pp. 180–187). Washington, DC: American Chemical Society.
Mottram, D. S., & Whitfield, F. B. (1994). Aroma volatiles from meat-like
Maillard systems. In T. H. Parliment, M. J. Morello, & R. J.
McGorrin (Eds.), Thermally generated flavors – Maillard, microwave,
and extrusion processes. ACS symposium series (Vol. 543, pp. 180–191).
Washington, DC: American Chemical Society.
Muir, P. D., Deaker, J. M., & Brown, M. D. (2003). Effects of forage- and
grain-based feeding systems on beef quality: A review. New Zealand
Journal of Agricultural Research, 41, 623–635.
Myer, R. O., Johnson, D. D., Knauft, D. A., Gorbet, D. W., Brendemuhl,
J. H., & Walker, W. R. (1992). Effect of feeding high-oleic-acid peanuts
to growing-finishing swine on resulting carcass fatty acid profile and on
carcass and meat quality characteristics. Journal of Animal Science, 70,
3734–3741.
Neely, T. R., Lorenzen, C. L., Miller, R. K., Tatum, J. D., Wise, J. W.,
Taylor, J. F., et al. (1998). Beef customer satisfaction: Role of cut,
USDA quality grade, and city on in-home consumer ratings. Journal of
Animal Science, 76(4), 1027–1033.
Neely, T. R., Lorenzen, C. L., Miller, R. K., Tatum, J. D., Wise, J. W.,
Taylor, J. F., et al. (1999). Beef customer satisfaction: Cooking
method and degree of doneness effects on the top round steak. Journal
of Animal Science, 77(3), 653–660.
Nuernberg, K., Wood, J. D., Scollan, N. D., Richardson, R. I., Nute, G.
R., Nuernberg, G., et al. (2005). Effect of a grass-based and a
concentrate feeding system on meat quality characteristics and fatty
acid composition of longissimus muscle in different cattle breeds.
Livestock Production Science, 94(2), 137–147.
Oltjen, R. R., Rumsey, T. S., & Putnam, P. A. (1971). All-forage diets for
finishing beef cattle. Journal of Animal Science, 32(2), 327–333.
Paterson, B. C., & Parrish, F. C. Jr., (1986). A sensory panel and chemical
analysis of certain beef chuck muscles. Journal of Food Science, 51(4),
876–879, 896.
Platter, W. J., Tatum, J. D., Belk, K. E., Chapman, P. L., Scanga, J. A., &
Smith, G. C. (2003). Relationships of consumer sensory ratings,
marbling score, and shear force value to consumer acceptance of beef
strip loin steaks. Journal of Animal Science, 81(11), 2741–2750.
Raes, K., Balcaen, A., Dirinck, P., De Winne, A., Claeys, E., Demeyer,
D., et al. (2003). Meat quality, fatty acid composition and flavour
analysis in Belgian retail beef. Meat Science, 65(4), 1237–1246.
Renerre, M., & Labadie, J. (1993). Fresh red meat packaging and meat
quality. In Proceedings of the 39th international congress of meat
science & technology, Alta., Canada (p. 361).
Rhee, K. S., Davidson, T. L., Cross, H. R., & Ziprin, Y. A. (1990).
Characteristics of pork products from swine fed a high monounsat-
urated fat diet: Part 1 – Whole muscle products. Meat Science, 27(4),
329–341.
Rhee, M. S., Wheeler, T. L., Shackelford, S. D., & Koohmaraie, M.
(2004). Variation in palatability and biochemical traits within and
among eleven beef muscles. Journal of Animal Science, 82(2), 534–550.
Richardson, R. I., Costa, P., Nute, G. R., Enser, M., Wood, J. D., &
Scollan, N. (2004). Effect of free and ruminally-protected fish oils on
fatty acid composition, sensory and oxidative characteristics of beef
loin muscle. In Proceedings of the 50th international congress of meat
science and technology, Helsinki, Finland (p. 2.43).
Roeber, D. L., Gill, R. K., & DiCostanzo, A. (2005). Meat quality
responses to feeding distiller’s grains to finishing Holstein steers.
Journal of Animal Science, 83(10), 2455–2460.
Romans, J. R., Johnson, R. C., Wulf, D. M., Libal, G. W., & Costello, W.
J. (1995). Effects of ground flaxseed in swine diets on pig performance
and on physical and sensory characteristics and omega-3 fatty acid
80 C.R. Calkins, J.M. Hodgen / Meat Science 77 (2007) 63–80
content of pork: I. Dietary level of flaxseed. Journal of Animal Science,
73(7), 1982–1986.
Romans, J. R., Wulf, D. M., Johnson, R. C., Libal, G. W., & Costello, W.
J. (1995). Effects of ground flaxseed in swine diets on pig performance
and on physical and sensory characteristics and omega-3 fatty acid
content of pork: II. Duration of 15% dietary flaxseed. Journal of
Animal Science, 73(7), 1987–1999.
Sanderson, G. W., Co, H., & Gonzalez, J. G. (1971). Biochemistry of tea
fermentation: The role of carotenes in black tea aroma formation.
Journal of Food Science, 36(2), 231–236.
Savell, J. W., Lorenzen, C. L., Neely, T. R., Miller, R. K., Tatum, J. D.,
Wise, J. W., et al. (1999). Beef customer satisfaction: Cooking method
and degree of doneness effects on the top sirloin steak. Journal of
Animal Science, 77(3), 645–652.
Scollan, N., Enser, M., Richardson, R. I., Gulati, S., Hallett, K. G., Nute,
G. R., et al. (2004). The effects of ruminally protected dietary lipid on
the lipid composition and quality of beef muscle. In Proceedings of the
50th international congress of meat science and technology, Helsinki,
Finland (p. 2.50).
Scollan, N., Richardson, I., Moloney, A., Dannenberger, D., Hocquette,
J. F., & Nuernberg, K. (2006). Innovations in beef production systems
that enhance the nutritional and health value of beef lipids and their
relationship with meat quality. Meat science, 74(1), 17–33.
Shackelford, S. D., Reagan, J. O., Haydon, K. D., & Miller, M. F. (1990).
Effects of feeding elevated levels of monounsaturated fats to growing-
finishing swine on acceptability of boneless hams. Journal of Food
Science, 55(6), 1485–1487, 1517.
Shackelford, S. D., Wheeler, T. L., & Koohmaraie, M. (1995). Relation-
ship between shear force and trained sensory panel tenderness ratings
of 10 major muscles from Bos indicus and Bos taurus cattle. Journal of
Animal Science, 73(11), 3333–3340.
Shahidi, F. (1989). Flavor of cooked meat. In R. Teranishi, R. G. Buttery,
& F. Shahidi (Eds.), Flavor chemistry trends and developments
(pp. 188–201). Washington, DC: American Chemical Society.
Shahidi, F., & Pegg, R. B. (1994). Hexanal as an indicator of the flavor
deterioration. Journal of Food Lipids, 1(1), 177–186.
Sitz, B. M., Calkins, C. R., Feuz, D. M., Umberger, W. J., & Eskridge, K.
M. (2005). Consumer sensory acceptance and value of domestic,
Canadian, and Australian grass-fed beef steaks. Journal of Animal
Science, 83(12), 2863–2868.
Smith, G. C., Savell, J. W., Cross, H. R., & Carpenter, Z. L. (1983). The
relationship of USDA quality grade to beef flavor. Food Technology,
37(5), 233–238.
Srinivasan, S., Xiong, Y. L., Blanchard, S. P., & Moody, W. G. (1998).
Proximate, mineral, and fatty acid composition of semimembranosus
and cardiac muscles from grass- and grain-fed and zeronal-implanted
cattle. Food Chemistry, 63(1), 543–547.
Stelzleni, A., Johnson, D., Calkins, C., Mink, B., & Gwartney, B. (2005).
Benchmarking muscles from various commercially available cow
carcasses. 2005 Florida beef report, Gainesville, FL.
Suzuki, J., & Bailey, M. E. (1985). Direct sampling capillary GLC analysis
of flavor volatiles from ovine. Food Chemistry, 33(3), 343–347.
Tai, C.-Y., & Ho, C. T. (1997). Influence of cysteine oxidation on thermal
formation of Maillard aromas. Journal of Agricultural and Food
Chemistry, 45(9), 3586–3589.
Tai, C.-Y., & Ho, C. T. (1998). Influence of glutathione oxidation and ph
on thermal formation of Maillard-type volatile compounds. Journal of
Agricultural and Food Chemistry, 46(6), 2260–2265.
Ullrich, F., & Grosch, W. (1987). Identification of the most intense volatile
flavour compounds formed during autoxidation of linoleic acid.
Zeitschrift fuer Lebensmittel-Untersuchung und-Forschung A, 184(4),
277–282.
Vasta, V., & Priolo, A. (2006). Ruminant fat volatiles as affected by diet –
A review. Meat Science, 73(2), 218–228.
Wheeler, T. L., Shackelford, S. D., & Koohmaraie, M. (2000). Relation-
ship of beef longissimus tenderness classes to tenderness of gluteus
medius, semimembranosus, and biceps femoris. Journal of Animal
Science, 78(11), 2856–2861.
Wood, J. D., & Enser, M. (1997). Factors influencing fatty acids in meat
and the role of antioxidants in improving meat quality. British Journal
of Nutrition, 78(Suppl. 1), S49–S60.
Wulf, D. M., Emnett, R. S., Leheska, J. M., & Moeller, S. J. (2002).
Relationships among glycolytic potential, dark cutting (dark, firm, and
dry) beef, and cooked beef palatability. Journal of Animal Science,
80(7), 1895–1903.
Wulf, D. M., & Page, J. K. (2000). Using measurements of muscle color,
pH, and electrical impedance to augment the current USDA beef
quality grading standards and improve the accuracy and precision of
sorting carcasses into palatability groups. Journal of Animal Science,
78(10), 2595–2607.
Yancey, E. J., Grobbel, J. P., Dikeman, M. E., Smith, J. S., Hachmeister,
K. A., Chambers, E. C. IV,, et al. (2006). Effects of total iron,
myoglobin, hemoglobin, and lipid oxidation of uncooked muscles on
livery flavor development and volatiles of cooked beef steaks. Meat
Science, 73(4), 680–686.
Yang, A., Lanari, M. C., Brewster, M., & Tume, R. K. (2002). Lipid
stability and meat colour of beef from pasture- and grain-fed cattle
with or without vitamin E supplement. Meat Science, 60(1), 41–50.
Young, O. A., & Baumeister, B. M. B. (1999). The effect of diet on the
flavour of cooked beef and the odour compounds in beef fat. New
Zealand Journal of Agricultural Research, 42(4), 297–304.
Young, O. A., Berdague, J.-L., Viallon, C., Roussett-Akrim, S., & Theriez,
M. (1997). Fat-borne volatiles and sheepmeat odour. Meat Science,
45(2), 183–200.
 MEAT
SCIENCE
Meat Science 77 (2007) 81–89
www.elsevier.com/locate/meatsci

Technologies to shorten the drying period of dry-cured meat products

J. Arnau *, X. Serra, J. Comaposada, P. Gou, M. Garriga
IRTA, Finca Camps i Armet, E-17121 Monells (Girona), Spain

Received 15 March 2007; accepted 15 March 2007

Abstract

Dry-cured meat products are well-known for their unique sensory characteristics. However, the traditional process is very time con-
suming. The process can be shortened especially by accelerating the drying period, which is the most time consuming. This paper deals
with some technological, safety and sensorial aspects for producing fermented sausages and dry-cured hams when the process time is
shortened. Different techniques, such as temperature increase and thickness reduction, and the effects of some ingredients and additives
are discussed. A Quick-Dry-Slice process based on a continuous system that combines both convective and vacuum drying could accel-
erate the drying of slices after the desired pH is reached in fermented sausages.
There are safety concerns when processes are shortened, but possible additional hurdles, such as the introduction of bacteriocin-pro-
ducing starter cultures and high-pressure treatments at the end of the process, could reduce them. Methods to speed up the development
of typical colour, texture and flavour and their limitations are also discussed.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Fermented sausages; Dry-cured ham; Accelerated production; Technology; Safety; Sensory properties

1. Introduction The aim of this paper is to present some technological,

safety and sensorial aspects for producing fermented sau-
Fermented sausages and dry-cured hams are traditional sages and dry-cured hams when the process time is
products prepared since the earliest civilizations to preserve reduced.
meat. Although nowadays meat can be preserved by freez-
ing, refrigeration and thermal processing, the traditional
2. Technological factors to shorten dry-cured meat
fermented sausages and dry-cured hams are still produced
production process
in large quantities because they have a unique and much
appreciated flavour. However, the traditional processes
The process for manufacturing fermented sausages
are very time consuming. They range from 1–2 weeks in
involves grinding of meat and fat, mixing with salt, spices
small calibre fermented sausages to 1.5–3 years in Iberian
and other ingredients and additives, and stuffing into cas-
hams (obtained from Iberian pigs fed and fattened with
ings. Natural casings are typically used in traditional prod-
acorns). Drying is the limiting step of the process in terms
ucts, whereas artificial casings are used in products to be
of time. A shortening of the drying period would result in a
sliced because they show higher water permeability and
reduction of the drying facilities, capital and labour, and
resistance, constant diameter and they are easily removed.
would increase the profit margin and the product compet-
Once stuffed, sausages are fermented to the desired pH and
itiveness while reducing some safety concerns, such as
finally dried to the target water content, at the appropriate
mould growth, lipid oxidation and mite infestation.
temperature and air humidity.
The dry-cured ham process is based on a salting-curing
*
Corresponding author. Tel.: +34 972 630 052; fax: +34 972 630 373. step where salt and other additives are absorbed, followed
E-mail address: jacint.arnau@irta.es (J. Arnau). by a resting period at a temperature below 5 C until water

0309-1740/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.03.015
82 J. Arnau et al. / Meat Science 77 (2007) 81–89
activity (aw) decreases below 0.96 to prevent growth of
undesirable microorganisms (Leistner, 1985). Finally, a
drying period follows where the temperature is gradually
raised as aw decreases, to accelerate the drying process
and the development of the typical aged flavour.
2.1. Fermentation step
In fermented sausages, the pH is a very important hur-
dle to stabilize the product and affects flavour and texture.
The faster the pH decreases, the shorter the drying process
can be. Carbohydrates, mainly dextrose, are the substrates
used by lactic acid bacteria (LAB), added or not as starter
cultures, to produce organic acids (e.g. lactic acid). To
shorten manufacturing time, acidification can be produced
by the addition of chemical additives, such as gluconodelt-
alactone – that hydrolyzes to gluconic acid, or encapsu-
lated acids (e.g. lactic and citric acids), that have a
mechanism for slow and targeted acid release by employing
coating materials that take time to dissolve/break down
(Gibbs, Kermasha, Alli, & Mulligan, 1999). The main
advantage of using chemical acidification is in eliminating
the 12–48 h fermentation period. However, it should be
noted that using LAB cultures is very common because
of its contribution to flavour.
Some ingredients help to speed up the fermentation pro-
cess. For instance, addition of soy protein isolates has been
shown to stimulate growth of starter cultures and thereby
to accelerate fermentation (Hagen, Næs, & Holck, 2000).
Moreover, the Mn+2 present in some spices accelerates
pH drop and stimulates growth of lactobacilli (Vanden-
driessche, Vandekerckhove, & Demeyer, 1980; Zaika &
Kissinger, 1984). The magnitude and perseverance of the
stimulating effect produced by Mn+2 differs with the type
of LAB starter (Hagen et al., 2000). Some LAB strains
may also act as bioprotective cultures by producing antimi-
crobial compounds (bacteriocins), thus enhancing the
safety of fermented sausages (Hugas, Garriga, Aymerich,
& Monfort, 1995; Hugas, Neumeyer, Pagès, Garriga, &
Hammes, 1996).
Fermented sausages with small calibre (initial
B < 20 mm) have a high surface/volume relationship that
facilitates oxygen penetration, increases redox potential
and hinders LAB growth and pH decline. In these prod-
ucts, the lower effectiveness of the so called traditional hur-
dles could be compensated by other hurdles, e.g. addition
of organic acids and selected starter cultures, decrease of
aw and high-pressure treatment at the end of the process
(Garriga et al., 2005).
In some Chinese traditional sausages, such as ‘‘Lup
Cheong’’, which contains up to 20% of sugar, the pH
remains relatively high (between 5.8 and 6.2) and is influ-
enced by the amount of added soy sauce. ‘‘Lup Cheong’’
is neither fermented nor ripened, but dried quickly at tem-
peratures up to 50 C (Savic, Sheng, & Savic, 1988). The
‘‘Fuet Dolç’’ (a Spanish dry sausage) is another non-fer-
mented product that is rapidly stabilized because it is pro-
duced by adding 10–20 g of salt, lemon zest and 800–1000 g
of sacarose per kg of meat and attains a water activity
between 0.83 and 0.85 just after stuffing. This allows a dry-
ing process at a high temperature without safety concerns
(Arnau & Matas, 2004).
Fermentation is not typical in the traditional dry-cured
ham process because no carbohydrates (or sometimes only
small amounts) are added, and LAB are not the predomi-
nant microorganisms. In traditional hams, pH tends to
increase during the process (Arnau, Guerrero, Casade-
mont, & Gou, 1995), especially in the surface during the
resting period if hams are submitted to high relative humid-
ity (Arnau, Gou, & Comaposada, 2003).
2.2. Salting
Processing techniques for the Northern European ham
types are characterised by brine immersion, followed by a
drying and/or smoking period and ageing for 3–12 months.
Other techniques, such as brine injection and vacuum tum-
bling, are also used for basic quality products. In contrast,
Southern European products are non-smoked and mainly
produced by means of salt rubbing on the surface. The age-
ing period is about 6 months in rapid processes and
between 1.5 and 3 years in Iberian ham. For bone-in hams
vacuum impregnation has been suggested to speed up salt
diffusion (Chiralt et al., 2001). A simultaneous brine thaw-
ing/salting operation has been proposed for the processing
of frozen hams which are salted after thawing (Barat et al.,
2006; Barat, Grau, Pagán-Moreno, & Fito, 2004). The
results from this study showed a significant reduction of
thawing and salting time needed to reach a NaCl concen-
tration, on a dry weight basis, similar to that obtained in
the traditional pile salting method. In fact, the use of brine
thawing/salting with saturated brine in fresh and thawed
hams allowed 58% and 61% time reductions, respectively,
compared with the traditional process (Barat, Grau, Ibá-
ñez, & Fito, 2005).
In order to speed up the process for dry-cured hams,
several additional production techniques have been pro-
posed: boning and skinning pork legs prior to cure applica-
tion (Montgomery, Kemp, & Fox, 1976), trimming of
subcutaneous and intermuscular fat, blade tenderisation
(Kemp & Fox, 1985) and tumbling (Kemp, Abidoye, Lang-
lois, Franklin, & Fox, 1980; Marriott, Graham, & Claus,
1992; Marriott, Graham, Shaffer, & Phelps, 1987a; Ocker-
man & Organisciak, 1978).
The salting process can also be accelerated by using
boned hams (trimmed of skin and subcutaneous fat and
excised in several pieces) combined with the curing mixture
in a tumbler under vacuum. Once the curing mixture has
been absorbed, the pieces could be treated with transgluta-
minase and vacuum packed to facilitate binding. This pro-
cess is currently under study to accelerate the stabilization
of dry-cured hams treated with potassium lactate and a
reduced amount of added salt, i.e. 15 g/kg of green ham
(Serra, Gou, Fulladosa, Costa, & Arnau, 2007). The
 J. Arnau et al. / Meat Science 77 (2007) 81–89
enzyme transglutaminase (TGase; EC 2.3.2.13) has the
property to form crosslinks between protein molecules
and has been proven useful for binding food proteins (De
Jong & Koppelman, 2002; Kuraishi et al., 1997; Motoki
& Seguro, 1998). Kuraishi et al. (1997) reported a good
binding strength in restructured pork meat obtained by
the addition of TGase with either salt (to enhance myofibr-
illar protein extraction) or caseinate (a good substrate for
TGase and acting as glue between meat pieces), and then
stuffed into casings and kept at 5 C for a period longer
than 2 h.
2.3. Drying-ageing
After the resting period in the dry-cured ham process,
the drying rate can be sped up by increasing the tempera-
ture and reducing the air relative humidity. In restructured
boned ham, drying can be favoured by reducing the ham
thickness and trimming away subcutaneous and intermus-
cular fat. The use of water-permeable plastic bags for
restructured-ham drying has been proposed to minimize
handling to improve hygiene and binding, to start drying
earlier and to prevent crusting, mould growth and mite
infestation (Serra et al., 2007).
Several applications have been reported to shorten dry-
ing time in fermented sausages: reducing the product cali-
bre (i.e. diameter) to reduce the distance for water to
reach the product surface, freeze-drying of the meat to
remove the highest amount of water prior to fermentation
(Lu & Townsend, 1973) and using PSE pork (Pale, Soft,
Exudative) to reduce water holding capacity (Townsend,
Davis, Lyon, & Mescher, 1980). Moreover, Chin, Keeton
and Lacey (1995) observed a 30% drying time reduction
(from 18 days to 12 days), without noticeable quality
defects in pieces of pepperoni dried under vacuum in labo-
ratory conditions. Recently, a new drying-maturing pro-
cess: Quick-Dry-Slice (QDS) has been proposed for sliced
products (Comaposada, Arnau, Gou, & Monfort, 2004).
Sausages are fermented to the desired pH, and are then fro-
zen, sliced and dried in a continuous system that combines
both convective and vacuum drying (Metalquimia, Girona,
Spain; www.metalquimia.com). With the QDS system, the
traditional drying process could be reduced to 30 min. The
safety of fermented sausages manufactured with this sys-
tem was similar to that of traditionally-dried fermented
products. Moreover, bacterial counts (i.e. Clostridium,
Escherichia coli, S. aureus, L. monocytogenes) and Salmo-
nella occurrence in vacuum-packaged slices (15 days of
storage at 4 C ± 1 C) were not different between QDS
and traditional process. Challenge tests with the inocula-
tion (100 CFU/g) of L. monocytogenes, S. aureus and Sal-
monella at the mixing step, showed a similar behaviour of
pathogens in both systems, i.e. traditional and QDS. At
the end of the process L. monocytogenes showed values
<1.30 log UFC/g, fulfilling the microbiological criteria for
these types of products (Commission Regulation, 2005).
However, Salmonella was still detected in 25 g in both types
83
of manufactured sausages. Moreover, promising results
were obtained when Salmonella was inoculated at lower
level (<3 MPN/g) and 2 g/kg of sodium acetate were
included in the formulation (Garriga, unpublished results).
These preliminary results will be further investigated.
Sausages fermented and dried with the QDS process had
a less acid flavour than sausages dried with the traditional
system. The reason for the lower acid flavour in the QDS
process is probably because there is neither a further acid-
ification during drying nor an acid gradient between the
outer part and central part of the slice. Furthermore, vola-
tile acids could be evenly removed during drying. For these
reasons, in the QDS process, the pH could be decreased to
lower values after the fermentation step than in the tradi-
tional process. Colour was more intense in the QDS pro-
cess and some sensitive colorants (e.g. Ponceau 4R) did
not fade during the process. The QDS product flavour
was slightly different from the traditional product and
improved during storage after packaging (Comaposada,
unpublished results).
Osmotic dehydration is used in a continuous drying sys-
tem in which a continuous flattened strip of minced meat
(within a sealed film sheath) goes through a dehydrating
solution bath. On exiting the bath, the sheath film is
removed and the dried minced meat strip is cut and pack-
aged (Sirami & Louthellier, 2002). The drying time
required for this process (Osmofood system) is around
3 h. During the process water activity is reduced so that
the bacteriological count is decreased by 2–3 logs (Anony-
mous, 2004).
3. Stabilization rate and safety concerns in dry-cured meat
products with shorter production times
Pathogenic and spoilage microorganisms are generally
inhibited in fermented sausages due to several hurdles:
nitrite curing salt, decrease of redox potential, growth of
competitive flora, pH decrease, aw decrease and micro-
structure. In dry-cured ham, the hurdles are low pH and
low temperature (in green hams and during resting) and
aw decrease (throughout the process). Both types of prod-
ucts are considered safe when produced and consumed
according to the traditional system. However, some con-
cerns exist about products that are produced more rapidly,
or that are commercialized and/or consumed in a different
way than the traditional system. Rapid fermentation might
be useful to increase the microbial safety of the final prod-
ucts, because it has been shown that L. monocytogenes may
grow when fermentation relies on endogenous flora (Gar-
riga et al., 2005) and both L. monocytogenes and E. coli
O157:H7 may grow during the early stages of the fermen-
tation process even when starter cultures are added (Nissen
& Holck, 1998). From a safety point of view, it may be bet-
ter to store some fermented sausages at room temperature
than in the cold (Leistner, 1995; Nissen & Holck, 1998),
provided that the sensory qualities are retained and that
similar results are obtained with other food pathogens.
84 J. Arnau et al. / Meat Science 77 (2007) 81–89
Several authors have reported the antilisterial effect of
bacteriocin-producing LAB strains in fermented sausages.
A suitable starter culture: Lb. sakei CTC494 (Garriga
et al., 1996), was able to suppress the growth of Listeria
(initially spiked at 9 · 103 CFU/g) and to reduce initial
counts by 2.64 log in dry sausages (Hugas et al., 1995).
Pediocins produced by Pd. acidilactici during fermentation
and enterocins produced by Ent. faecium CTC492,
included in the meat mixture, provide an additional hurdle
against Listeria proliferation in different types of fermented
sausages (Aymerı́ch, Garriga, Jofré, Martı́n, & Monfort,
2006).
High-hydrostatic-pressure (HHP) processing is a non-
thermal food preservation technology with more prospects
nowadays for its application in the meat industry (Hugas,
Garriga, & Monfort, 2002). HHP has been recommended
to produce low-risk and high-quality slightly fermented
sausages (Garriga et al., 2005; Marcos, Aymerich, Guàr-
dia, & Garriga, 2007) and some companies use it as an
additional hurdle to decrease pathogen-related concerns,
both in dry-cured hams and fermented sausages.
In challenge studies, a reduction of Salmonella, inocu-
lated in sausages at the time of manufacturing, was
observed during ripening, irrespective of the addition of
starter cultures. Moreover, HHP treatment at 400 MPa
and 17 C for 10 min on ripened sausages ensured the
absence of the pathogen in slightly fermented sausages.
Selected starter cultures proved to be useful to control
enterococci population and biogenic amine content in both
types of sausages (‘‘fuet’’ and ‘‘chorizo’’) whereas HHP
might be considered an additional hurdle only in ‘‘fuet’’.
L. monocytogenes decreased significantly in starter batches,
but no additional reduction was recorded after HHP treat-
ment. Pressurization after ripening diminished the inocu-
lated Listeria in the non-starter batches, although the
starter cultures used were more effective than HHP treat-
ment (400 MPa) to reduce L. monocytogenes in these types
of sausages (Garriga et al., 2005). Complying with the good
manufacturing practices, the combination of the appropri-
ate starter culture(s) and HHP treatment after ripening will
improve the safety of slightly fermented sausages.
Some studies have also been done on other types of
cured meat products. Vacuum-skin-packaged dry-cured
ham slices treated at 600 MPa for 6 min showed a signifi-
cant reduction in spoilage endogenous microflora and
counts were kept at low levels during the 120-day storage
period. L. monocytogenes was present in one of the control
samples but absent in all HHP-treated samples during the
whole storage period studied. Salmonella and Campylobac-
ter were not detected in any samples either control or
HHP-treated (Garriga, Grèbol, Aymerich, Monfort, &
Hugas, 2004). Different inactivation levels after HHP treat-
ment were observed in dry-cured ham spiked with L. mon-
ocytogenes, depending on the equipment used (Hugas et al.,
2002), which suggest that further research is needed to
achieve the desired levels of microbial inactivation and
shelf life extension.
4. Sensory properties of dry-cured meat products with a short
production process
4.1. Colour
The typical cured colour is usually obtained by the addi-
tion of nitrate for long processes, or by nitrite in fast pro-
cesses. Nitrate by itself does not produce the cured colour
and has to be reduced to nitrite by some nitrate reductase
microorganisms. So, this process could be accelerated by
adding some starter cultures, such as Staphylococcus and
Kocuria. If nitrate reduction is necessary, then fermenta-
tion must occur slowly to allow these bacteria to grow.
The pH of the mince should be from 6.0 to 5.4, because
a lower pH would prevent nitrate reductase activity. Nitrite
is reduced by ascorbate and some reductive substances of
meat to nitric oxide, which finally reacts with myoglobin
(Mb) to produce the nitrosylmyoglobin complex
(MbFe(II)NO). This complex is the main contributor to
the characteristic cured colour. In fermented sausages cur-
ing agents are added to the mixture, but in dry-cured ham
the curing salts have to diffuse through the muscle (Fox,
1980). In order to speed up this diffusion process, vacuum
impregnation (Chiralt et al., 2001) and brine injection
could be used. Nitric oxide has been proposed to reduce
curing time of boneless pork legs (Marriott, Tracy, Kelly,
& Graham, 1983). However, its use needs further study
to determine potential limitations and applicability to exist-
ing industrial processes.
In products prepared using only NaCl, such as Parma
ham and some Iberian hams, colour is due to a slow for-
mation of the Zn-protoporphyrine IX complex, believed
to originate from Mb in which Fe has been substituted
by Zn and the heme group separated from the native
heme-protein (Møller, Adamsen, Catharino, Skibsted, &
Eberlin, 2007; Wakamatsu, Nishimura, & Hattori,
2004a). This complex is not formed if there is any contact
with oxygen or curing substances (Adamsen, Møller,
Laursen, Olsen, & Skibsted, 2006; Wakamatsu, Okui,
Ikeda, Nishimura, & Hattori, 2004b). Three possible sub-
stitution patterns have been suggested: (i) a non-enzy-
matic reaction in which Zn(II) substitutes Fe(II) under
anaerobic conditions with concomitant dissociation from
the heme; (ii) a bacterial enzymatic reaction where bacte-
ria degrade the pigment, or (iii) an enzymatic reaction
where an endogenous ferrochelatase interchanges the
two metals (Adamsen et al., 2006; Wakamatsu et al.,
2004). More knowledge about these mechanisms and
how to speed up the formation of the stable red pigment
in Parma ham would enable the manufacturing of meat
products with a desirable and stable red colour without
the use of nitrite and nitrate.
Apart from the above-mentioned reactions, some ingre-
dients (e.g. paprika) and colorants could contribute to the
colour of fermented sausages. The differences in colour
caused by the use of a variety of ingredients in fermented
sausages (e.g. non-meat proteins) could be compensated
 J. Arnau et al. / Meat Science 77 (2007) 81–89
by using colour-intense raw materials and colouring
agents.
4.2. Texture
The consistency of fermented sausages increases due to
acidification and drying. During fermentation the pH
declines and the myofibrillar proteins aggregate to form
a gel. After gelation, drying is a major factor affecting
binding and rheological properties. Non-meat proteins,
such as wheat, soy isolates and whey proteins, and
freeze-dried meat could also increase consistency (Stie-
bing, 1999). The more rapid the pH declines, the firmer
the sausage is. Therefore, the addition of rapidly acidify-
ing starter cultures and the presence of microelements
promoting microbial development, such as Mn+2, could
shorten the production time because they harden the
sausages.
A relationship between hardness and water content has
been reported in dry-cured ham and loin (Fig. 1). Dry-
cured ham hardness increases slightly with the decrease in
water content until it reaches around 0.6 g H2O/g of dry
matter. However, below this critical water content there
is an important increase in hardness, that could be related
with the crust development at the surface which is the most
important texture problem in the external zones of the dry-
cured hams (Serra, Ruiz-Ramı́rez, Arnau, & Gou, 2005).
The first approach to obtain an acceptable texture with a
short drying period in dry-cured meat products is to accel-
erate the drying process by decreasing the relative humidity
and increasing the temperature of the drying air. However,
this implies a reduction in the water activity, and conse-
quently in the water content at the surface. To avoid the
crusting problem, precise control of water content at the
surface should be maintained.
12.24
10.20
8.16
Hardness (N)
6.12
4.08
2.04
0.00
0.2 0.4
Water content (X = kg H2O / kg dry matter)
Fig. 1. Relationship between water content and TPA hardness: (1) Serra et al. (2005); (2) Ruiz-Ramı́rez et al. (2005b); (3) Ruiz-Ramı́rez et al. (2005a); (4)
Ruiz-Ramı́rez et al. (2006); BF: biceps femoris; SM: semimembranosus.
85
The relationship between hardness and water content in
dry-cured hams is also affected by proteolysis index (PI),
pH and salt content (Ruiz-Ramı́rez, Arnau, Serra, &
Gou, 2006). Hams with higher salt content and hams with
higher initial pH showed lower PI. Hams with lower PI
achieved the same level of hardness with higher water con-
tents than those with higher PI (Fig. 2). Therefore, the crit-
ical water content during drying at the surface will depend
on the initial pH and the added NaCl level.
The most important texture problems in the inner zones
of dry-cured ham are excessive softness (Parolari, Virgili, &
Schivazappa, 1994; Virgili, Parolari, Schivazappa, Bordini,
& Borri, 1995) and pastiness (Arnau, 1991; Arnau, Guer-
rero, & Sárraga, 1998; Garcı́a-Garrido, Quiles-Zafra, Tap-
iador, & Luque de Castro, 2000; Garcı́a-Rey, Garcı́a-
Garrido, Quiles-Zafra, Tapiador, & Luque de Castro,
2004). Softness is associated with proteolysis (Parolari
et al., 1994; Virgili et al., 1995), which depends on moisture
content, salt content and temperature. Therefore the prote-
olysis activity should be considered when the temperature
is increased to accelerate the drying process, especially in
the first steps of the process and when low salt products
are elaborated. High pressure affects the relationship
between hardness and water content and could be useful
to decrease softness and pastiness in dry-cured hams. Serra
et al. (2006) observed that pressurization improved the sen-
sory texture of pasty dry-cured hams by increasing hard-
ness and reducing pastiness. This could be useful for
producing low salt dry-cured hams.
4.3. Flavour
The flavour characteristics of dry sausages are thought
to result from a combination of spices, meat endogenous
enzyme activities, microbial activities, autoxidation
BF muscles of commercial dry-cured hams (1)
Dry-cured loins (2)
BF and SM muscles dried for 45 days (3)
BF and SM mucles from dry-cured hams
(289-day process) (4)
0.6 0.8 1.0 1.2 1.4 1.6 1.8
86 J. Arnau et al. / Meat Science 77 (2007) 81–89
7 and microbial counts are usually normal for these prod-
15.51% PI
6 release by lipases, and increase slightly because of the
19.96% PI
23.70% PI
5
28.44% PI
Hardness (N)
4 mental conditions, or to add either an efficient starter
3 obtained from non-toxigenic moulds), so that volatiles
2 usual (Bruna, Fernández, Ordóñez, & De la Hoz, 2002;
1
0
0.20 0.45 0.70 0.95 1.20 1.45
Water content (X = kgH2O / kg dry matter)
Fig. 2. Predicted TPA hardness versus water content (X) according to the
proteolysis index (PI) of dry-cured hams (Ruiz-Ramı́rez et al., 2006).
processes and the interaction among odorous compounds,
whose relative importance varies from product to product
(Ordóñez, Hierro, Bruna, & De la Hoz, 1999).
Attempts to accelerate flavour development, by means
of enhancing lipid and protein breakdown, have been
carried out on cheese by using genetically modified star-
ter cultures, ripening at high temperature and with the
addition of enzymes or slurry systems (El Soda & Pan-
dian, 1991). In a similar way, current understanding of
reactions occurring during the process of fermented sau-
sages has led to the acceleration of ripening and flavour
through the addition of lipases and proteinases, e.g. lip-
ases from Candida cylindracea and Rhizomucor miehei
(Zalacain, Zapelena, Astiasarán, & Bello, 1995, 1996;
Zalacain, Zapelena, Paz de Peña, Astiasarán, & Bello,
1997a; Zalacain, Zapelena, Paz de Peña, Astiasarán, &
Bello, 1997b), pancreatic lipase (Fernández, De la Hoz,
Dı́az, Cambero, & Ordóñez, 1995), serine proteinase from
Lactobacillus paracasei (Blom et al., 1996; Naes, Holck,
Axelsson, Andersen, & Blom, 1995), pronase from Strep-
tomyces griseus (Dı́az, Fernández, Garcı́a de Fernando,
De la Hoz, & Ordóñez, 1993), aspartyl proteinase from
Aspergillus oryzae and papain from Carica papaya (Dı́az,
Fernández, Garcı́a de Fernando, De la Hoz, & Ordóñez,
1997). Results have shown that it is possible to accelerate
proteolysis and lipolysis, but only a slight flavour
improvement was obtained in some cases. The softening
effect of proteases is more important than their effect
on flavour, and at high concentrations, an excessive soft-
ening is observed. Lipase addition increases the hydroly-
sis of triglycerides to free fatty acids, producing a more
intense oxidation, and when high amounts are added
an oily texture may appear (Fernández, Ordóñez, Bruna,
Herranz, & De la Hoz, 2000). The use of either lipases or
proteases does not affect the drying rate, and pH values
ucts. However, pH can decrease slightly due to fatty acid
addition of proteases (Fernández et al., 2000). So, the
addition of proteinases and lipases alone is not appropri-
ate for shortening the ripening period (Ordóñez et al.,
1999). Thus, it is necessary to find appropriate environ-
culture, or other types of enzymes (e.g. crude extracts
may form from amino acids and fatty acids faster than
Fernández et al., 2000; Herranz et al., 2004). Otherwise,
a much longer time is necessary for transformation of
free amino acids and fatty acids through microbial and
chemical reactions into the compounds responsible for
the flavour.
In the QDS process (Comaposada et al., 2004), the fla-
vour obtained in slices of sausages dried after fermentation
(to the same drying level as obtained by traditional process)
was considered satisfactory: (i) when selected aromas were
added together with the typical ingredients, additives and
starter cultures, or (ii) when these aromas were added at
the end of the ageing process and the sliced product was
then kept packaged in anaerobic conditions for several
days.
The aroma of Serrano and Iberian hams has a lipid oxi-
dation base and many aromatic nuances from amino acids
originated through Maillard reactions and Strecker degra-
dation. Nevertheless, compounds from the feeding and of
microbial origin could also contribute to the overall flavour
(Flores, Grimm, Toldrá, & Spanier, 1997; Garcı́a et al.,
1991; Ruiz, Muriel, & Ventanas, 2002; Toldrá & Flores,
1998; Toldrá, 1998; Ventanas et al., 1992). The leaner prod-
ucts have a shorter drying process. However, the presence
of some fat is desirable because it is an excellent solvent
for some aroma compounds and helps to preserve the typ-
ical aroma of fermented sausages and dry-cured ham.
Rogers, Kemp and Varney (1965) attempted to accel-
erate flavour development in dry-cured ham through
the injection of pancreatic lipase and papain into fresh
pork legs. Conventional dry-cured hams were preferred
to both enzyme-treated samples, and the papain-treated
hams were considered too mushy. The use of starter cul-
tures to improve the quality and safety of dry-cured hams
has also been studied (Bartholomew & Blumer, 1977;
Marriott, Phelps, Graham, & Shaffer, 1987b; Sánchez-
Molinero & Arnau, 1998). In general, the expected effect
of starter cultures on dry-cured hams is lower than in fer-
mented sausages due to the low temperature and superfi-
cial water activity of the product during both salting and
resting periods. In addition, problems could arise in the
ham structure if they were injected. So, greater knowl-
edge about the ability of the starter cultures to develop
appropriate aromas and how they can be used in new
processes in the meat industry will lead to meat products
with improved flavour.
 J. Arnau et al. / Meat Science 77 (2007) 81–89
5. Conclusions
To sum up, a lot of effort has been made over the last
decades, from a technological, safety and sensorial point
of view, to shorten the manufacturing process of fermented
sausages and dry-cured hams. However, further studies
should be focussed on the combination of different technol-
ogies, which seem to offer the possibility of producing rapid
dry-cured meat products and to meet the consumer
demand for quality and safety.
Acknowledgements
Part of this study has been supported by the Spanish
Ministry of Science and Technology (Project AGL2003-
04612) and the TRUEFOOD European Commission Inte-
grated Project within the Sixth RTD Framework Pro-
gramme (Contract n FOOD-CT-2006-016264). The
information in this document reflects only the authors’
views and the Community is not liable for any use that
may be made of the information contained therein.
References
Adamsen, C. E., Møller, J. K. S., Laursen, K., Olsen, K., & Skibsted, L.
H. (2006). Zn-porphyrin formation in cured meat products: effect of
added salt and nitrite. Meat Science, 72(4), 672–679.
Anonymous (2004). IFFA: Further processing. Meat International, 14(3),
24.http://www.agriworld.nl/public/file/pdf/20060502-mecs14.3.pdf.
Arnau, J. (1991). Aportaciones a la calidad tecnológica del jamón curado
elaborado por procesos acelerados. Thesis, Universitat Autònoma de
Barcelona, Facultat de Veterinària, Barcelona, Spain.
Arnau, J., Gou, P., & Comaposada, J. (2003). Effect of relative humidity
of drying air during the resting period on the composition and
appearance of dry-cured ham surface. Meat Science, 65, 1275–1280.
Arnau, J., Guerrero, L., Casademont, G., & Gou, P. (1995). Physical and
chemical changes in different zones of normal and PSE dry-cured ham
during processing. Food Chemistry, 52, 63–69.
Arnau, J., Guerrero, L., & Sárraga, C. (1998). The effect of green ham pH
and NaCl concentration on cathepsin activities and sensory charac-
teristics of dry-cured ham. Journal of the Science of Food and
Agriculture, 77, 387–392.
Arnau, J., & Matas, J. (2004). La ‘‘botifarra dolça’’ i el ‘‘fuet dolç’’:
sistemas de elaboración y formas de consumo. Eurocarne, 126,
129–131.
Aymerı́ch, T., Garriga, M., Jofré, A., Martı́n, B., & Monfort, J. M. (2006).
The use of bacteriocins against meat-borne pathogens. In L. M. L.
Nollet & F. Toldrá (Eds.), Advanced Technologies for Meat Processing
(pp. 371–399). Boca Ratón, FL (USA): CRC Press. Taylor & Francis
Group.
Barat, J. M., Grau, R., Ibáñez, J. B., & Fito, P. (2005). Post-salting studies
in Spanish cured ham manufacturing. Time reduction by using brine
thawing–salting. Meat Science, 69(2), 201–208.
Barat, J. M., Grau, R., Ibáñez, J. B., Pagán, M. J., Flores, M., Toldrá, F.,
et al. (2006). Accelerated processing of dry-cured ham. Part I.
Viability of the use of brine thawing/salting operation. Meat Science,
72(4), 757–765.
Barat, J. M., Grau, R., Pagán-Moreno, M. J., & Fito, P. (2004).
Replacement of pile salting by simultaneous brine thawing–salting in
Spanish cured ham manufacturing. Meat Science, 66(3), 603–608.
Bartholomew, D. T., & Blumer, T. N. (1977). The use of a commercial
Pediococcus cerevisiae starter culture in the production of country-
style hams. Journal of Food Science, 42, 494–497.
87
Blom, H., Hagen, B. F., Pedersen, B. O., Holck, A. L., Axelsson, L., &
Naes, H. (1996). Accelerated production of dry fermented sausages.
Meat Science, 43(S), S229–S242.
Bruna, J. M., Fernández, M., Ordóñez, J. A., & De la Hoz, L. (2002).
Enhancement of the flavour development of dry fermeanted sausages
by using a protease (Pronase E) and a cell-free extract of Penicillium
camemberti. Journal of the Science of Food and Agriculture, 82(5),
526–533.
Chin, K. B., Keeton, J. T., & Lacey, R. E. (1995). Utilization of vacuum-
drying for reduction of pepperoni drying time. In Proceedings 41st
international congress of meat science and technology (pp. 501–502),
San Antonio, Texas, USA.
Chiralt, A., Fito, P., Barat, J. M., Andrés, A., González-Martı́nez, C.,
Escriche, I., et al. (2001). Use of vacuum impregnation in food salting
process. Journal of Food Engineering, 49, 141–151.
Comaposada, J., Arnau, J., Gou, P., & Monfort, J. M. (2004). Accelerated
method for drying and maturing sliced food products. Patent WO/
2005/092109.
Commission Regulation, (EC) No 2073/2005, of 15 November 2005, on
microbiological criteria for foodstuffs. Official Journal of the European
Union, L 338/1-26.
De Jong, G. A. H., & Koppelman, S. J. (2002). Transglutaminase
catalyzed reactions: impact on food applications. Journal of Food
Science, 67, 2798–2806.
Dı́az, O., Fernández, M., Garcı́a de Fernando, G. D., De la Hoz, L.,
& Ordóñez, J. A. (1993). Effect of the addition of pronase E on
the proteolysis in dry-fermented sausages. Meat Science, 34,
205–216.
Dı́az, O., Fernández, M., Garcı́a de Fernando, G. D., De la Hoz, L.,
& Ordóñez, J. A. (1997). Proteolysis in dry-fermented sausages:
the effect of selected exogenous proteases. Meat Science, 46,
115–128.
El Soda, M., & Pandian, S. (1991). Recent developments in accelerated
cheese ripening. Journal of Dairy Science, 74, 2317–2335.
Fernández, M., De la Hoz, L., Dı́az, O., Cambero, I., & Ordóñez, J. A.
(1995). Effect of the addition of pancreatic lipase on the ripening of
dry-fermented sausages. Part I. Microbial, physico-chemical and
lipolitic changes. Meat Science, 40, 159–170.
Fernández, M., Ordóñez, J. A., Bruna, J. M., Herranz, B., & De la Hoz, L.
(2000). Accelerated ripening of dry fermented sausages. Trends in Food
Science and Technology, 11, 201–209.
Flores, M., Grimm, C. C., Toldrá, F., & Spanier, A. M. (1997).
Correlations of sensory and volatile compounds of Spanish ‘‘Serrano’’
dry-cured ham as a function of two processing times. Journal of
Agricultural and Food Chemistry, 45, 2178–2186.
Fox, J. B. Jr., (1980). Diffusion of chloride, nitrite, and nitrate in beef and
pork. Journal of Food Science, 45(6), 1740–1744.
Garcı́a, C., Berdagué, J. L., Antequera, T., López-Bote, C., Córdoba, J. J.,
& Ventanas, J. (1991). Volatile components of dry-cured Iberian ham.
Food Chemistry, 41, 23–32.
Garcı́a-Garrido, J., Quiles-Zafra, R., Tapiador, J., & Luque de Castro, M.
(2000). Activity of cathepsin B, D, H and L in Spanish dry-cured ham
of normal and defective texture. Meat Science, 56, 1–6.
Garcı́a-Rey, R., Garcı́a-Garrido, J., Quiles-Zafra, R., Tapiador, J., &
Luque de Castro, M. (2004). Relationship between pH before salting
and dry-cured ham quality. Meat Science, 67, 625–632.
Garriga, M., Grèbol, N., Aymerich, M. T., Monfort, J. M., & Hugas, M.
(2004). Microbial inactivation after high pressure processing at 600
MPa in commercial meat products over its shelf life. Innovative Food
Science and Emerging Technologies, 5(4), 451–457.
Garriga, M., Hugas, M., Gou, P., Aymerich, T., Arnau, J., & Monfort, J.
M. (1996). Technological and sensorial evaluation of Lactobacillus
strains as starter cultures in fermented sausages. International Journal
of Food Microbiology, 32, 173–183.
Garriga, M., Marcos, B., Martı́n, B., Veciana-Nogués, T., Bover-Cid, S.,
Hugas, M., et al. (2005). Starter cultures and high pressure processing
to improve the hygiene and safety of slightly fermented sausages.
Journal of Food Protection, 68(11), 2341–2348.
88 J. Arnau et al. / Meat Science 77 (2007) 81–89
Gibbs, B. F., Kermasha, S., Alli, I., & Mulligan, C. N. (1999).
Encapsulation in food industry: a review. International Journal of
Food Science and Nutrition, 50(3), 213–224.
Hagen, B. F., Næs, H., & Holck, A. L. (2000). Meat starters have
individual requirements for Mn+2. Meat Science, 55(2), 161–168.
Herranz, B., Fernández, M., Hierro, E., Bruna, J. M., Ordóñez, J. A., &
De la Hoz, L. (2004). Use of Lactococcus lactis subsp. Cremoris
NCDO 763 and a-ketoglutarate to improve the sensory quality of dry
fermented sausages. Meat Science, 66(1), 151–163.
Hugas, M., Garriga, M., Aymerich, T., & Monfort, J. M. (1995). Inhition
of Listeria in dry fermented sausages by the bacteriocinogenic
Lactobacillus sake CTC494. Journal of Applied Bacteriology, 79,
322–330.
Hugas, M., Garriga, M., & Monfort, J. M. (2002). New mild technologies
in meat processing: High pressure as a model technology. Meat
Science, 62(3), 359–371.
Hugas, M., Neumeyer, B., Pagès, F., Garriga, M., & Hammes, W. P.
(1996). Antimicrobial activity of bacteriocin-producing cultures in
meat products. Comparison of the antilisterial potential of bacteriocin-
producing lactobacilli in fermented sausages. Fleischwirtschaft, 76(6),
649–652.
Kemp, J. D., Abidoye, D. F. O., Langlois, B. E., Franklin, J. B., & Fox, J.
D. (1980). Effect of curing ingredients, skinning, and boning on yield,
quality, and microflora of country hams. Journal of Food Science, 45,
174–177.
Kemp, J. D., & Fox, J. D. (1985). Effect of needle tenderization on salt
absorption, yields, composition and palatability of dry-cured hams
produced from packer style and skinned green hams. Journal of Food
Science, 50, 295–299.
Kuraishi, C., Sakamoto, J., Yamakazi, K., Susa, Y., Kuhara, C., & Soeda,
T. (1997). Production of restructured meat using microbial transglu-
taminase without salt or cooking. Journal of Food Science, 62,
488–490, 515.
Leistner, L. (1985). Empfehlungen für sichere Produkte. In Mikrobiologie
und Qualität für Rohwurst und Rohschinken (pp. 219–244). Kulmbach:
Institut für Mikrobiologie, Toxikologie und Histologie der Bundesan-
stalt für Fleischforschung.
Leistner, L. (1995). Principles and applications of hurdle technology. In G.
W. Gould (Ed.), New Methods of Food Preservation (pp. 1–21).
London: Blackie Academic & Professional.
Lu, J., & Townsend, W. E. (1973). Feasibility of adding freeze-dried meat
in the preparation of fermented dry sausage. Journal of Food Science,
45, 622–626.
Marcos, B., Aymerich, T., Guàrdia, M. D., & Garriga, M. (2007).
Assessment of high hydrostatic pressure and starter culture on the
quality properties of low-acid fermented sausages. Meat Science, 76(1),
46–53.
Marriott, N. G., Graham, P. P., & Claus, J. R. (1992). Accelerated dry
curing of pork legs (hams): a review. Journal of Muscle Foods, 3,
159–168.
Marriott, N. G., Graham, P. P., Shaffer, C. K., & Phelps, S. K. (1987a).
Accelerated production of dry cured hams. Meat Science, 19(1), 53–64.
Marriott, N. G., Phelps, S. K., Graham, P. P., & Shaffer, C. K. (1987b).
Effect of inoculation on quality of dry-cured hams. Journal of Food
Quality, 10, 351–358.
Marriott, N. G., Tracy, J. B., Kelly, R. F., & Graham, P. P. (1983).
Accelerated processing of boneless hams to dry-cured state. Journal of
Food Protection, 46, 717–721.
Møller, J. K. S., Adamsen, C. E., Catharino, R. R., Skibsted, L. H., &
Eberlin, M. N. (2007). Mass spectrometric evidence for a zinc-
porphyrin complex as the red pigment in dry-cured Iberian and Parma
ham. Meat Science, 75(2), 203–210.
Montgomery, R. E., Kemp, J. D., & Fox, J. D. (1976). Shrinkage,
palatability and chemical characteristics of dry-cured country ham as
affected by skinning procedure. Journal of Food Science, 41,
1110–1115.
Motoki, M., & Seguro, K. (1998). Transglutaminase and its use for food
processing. Trends in Food Science & Technology, 9, 204–210.
Naes, H., Holck, A. L., Axelsson, L., Andersen, H. J., & Blom, H. (1995).
Accelerated ripening of dry-fermented sausage by addition of a
Lactobacillus proteinase. International Journal of Food Science and
Technology, 29, 651–659.
Nissen, H., & Holck, A. (1998). Survival of Escherichia coli O157:H7,
Listeria monocytogenes and Salmonella kentucky in norwegian fer-
mented dry sausage. Food Microbiology, 15, 273–279.
Ockerman, H. W., & Organisciak, C. S. (1978). Diffusion of curing brine
in tumbled and non-tumbled porcine tissue. Journal of Food Protec-
tion, 41, 178–181.
Ordóñez, J. A., Hierro, E. M., Bruna, J. M., & De la Hoz, L. (1999).
Changes in the components of dry-fermented sausages during ripening.
Critical Reviews in Food Science and Nutrition, 39(4), 329–367.
Parolari, G., Virgili, R., & Schivazappa, C. (1994). Relationship between
cathepsin B activity and compositional parameters in dry-cured hams
of normal and defective texture. Meat Science, 38, 117–122.
Rogers, R. W., Kemp, J. D., & Varney, W. Y. (1965). Effects of enzyme
addition on palatability, tenderness and rancidity development in
quick aged country-style hams. Food Technology, 19, 1509–1513.
Ruiz, J., Muriel, E., & Ventanas, J. (2002). The flavour of Iberian ham. In
F. Toldrá (Ed.), Research Advances in the Quality of Meat and Meat
Products (pp. 289–309). Kerala, India: Research Signpost.
Ruiz-Ramı́rez, J., Arnau, J., Serra, X., & Gou, P. (2005a). Relationship
between water content, NaCl content, pH and texture parameters in
dry-cured muscles. Meat Science, 70, 579–587.
Ruiz-Ramı́rez, J., Arnau, J., Serra, X., & Gou, P. (2006). Effect of pH24,
NaCl content and proteolysis index on the relationship between water
content and texture parameters in biceps femoris and semimembranosus
muscles in dry-cured ham. Meat Science, 72, 185–194.
Ruiz-Ramı́rez, J., Serra, X., Gou, P., & Arnau, J. (2005b). Profiles of
water content, water activity and texture in crusted dry-cured loin and
in non-crusted dry-cured loin. Meat Science, 69, 519–525.
Sánchez-Molinero, F., & Arnau, J. (1998). The effect of inoculation of a
starter culture on sensory characteristics of dry-cured hams using two
different resting technologies. In Proceedings 44th international con-
gress of meat science and technology (Vol. 2) (pp. 1010–1011).
Barcelona, Spain.
Savic, Z., Sheng, Z. K., & Savic, I. (1988). Chinese-style sausages: a special
class of meat products. Fleischwirtschaft, 68(5), 612–617.
Serra, X., Gou, P., Guerrero, L., Guàrdia, M. D., Picouet, P., Grèbol, N.
et al. (2006). Effect of high pressure on the sensory properties of pasty
hams. Workshop on applications of novel technologies in food and
biotechnology, IFT-NDP/EFFoST, Cork, Ireland (p.125).
Serra, X., Gou, P., Fulladosa, E., Costa, A., & Arnau, J. (2007).
Aplicación de lactato potásico en jamón curado deshuesado con un
contenido reducido de NaCl. IV Congreso Mundial del Jamón,
Salamanca, Spain.
Serra, X., Ruiz-Ramı́rez, J., Arnau, J., & Gou, P. (2005). Texture
parameters of m. biceps femoris from dry-cured hams as a function of
water activity and water content. Meat Science, 69, 249–254.
Sirami, J., & Louthellier, P. (2002). Method for continuous production of
minced meat, dried and restored in thin slabs. Patent WO/2003/
007722.
Stiebing, A. (1999). Effect of proteins on the course of ripening at raw
sausages. Fleischwirtschaft International, 2, 30–34.
Toldrá, F. (1998). Proteolysis and lipolysis in flavour development of dry-
cured meat products. Meat Science, 49, S101–S110.
Toldrá, F., & Flores, M. (1998). The role of muscle proteases and lipases
in flavour development during the processing of dry-cured ham.
Critical Reviews in Food Science and Nutrition, 38, 331–352.
Townsend, W. E., Davis, C. E., Lyon, C. E., & Mescher, S. E. (1980).
Effect of pork quality on some chemical, physical and processing
properties of fermented dry sausage. Journal of Food Science, 45,
622–626.
Vandendriessche, F., Vandekerckhove, P., & Demeyer, D. (1980). The
influence of spices on the fermentation of a Belgian sausage. In
Proceedings of the 26th European Meeting of Meat Research Workers
(pp. 128–133), Colorado Springs, U.S.A.
 J. Arnau et al. / Meat Science 77 (2007) 81–89
Ventanas, J., Córdoba, J. J., Antequera, T., Garcı́a, C., López-Bote, C., &
Asensio, M. A. (1992). Hydrolysis and Maillard reactions during
ripening of Iberian Ham. Journal of Food Science, 57(4), 813–815.
Virgili, R., Parolari, G., Schivazappa, C., Bordini, C. S., & Borri, M.
(1995). Sensory and texture quality of dry-cured ham as affected by
endogenous cathepsin B activity and muscle composition. Journal of
Food Science, 60, 1183–1186.
Wakamatsu, J., Nishimura, T., & Hattori, A. (2004a). A Zn-porphyrin
complex contributes to bright red colour in Parma ham. Meat Science,
67, 95–100.
Wakamatsu, J., Okui, J., Ikeda, I., Nishimura, T., & Hattori, A. (2004b).
Establishment of a model experiment system to elucidate the mech-
anism by which ‘‘Zn-protoporphyrin IX’’ is formed in nitrite free dry-
cured ham. Meat Science, 68, 313–317.
Zaika, L. L., & Kissinger, J. C. (1984). Fermentation enhancement by
spices: Identification of active component. Journal of Food Science,
49(1), 5–9.
89
Zalacain, I., Zapelena, M. J., Astiasarán, I., & Bello, J. (1995). Dry-
fermented sausages elaborated with lipase from Candida cylindr-
acea. Comparison with traditional formulations. Meat Science, 40,
55–61.
Zalacain, I., Zapelena, M. J., Astiasarán, I., & Bello, J. (1996).
Addition of lipase from Candida cylindracea to a traditional
formulation of a dry-fermented sausage. Meat Science, 42,
155–163.
Zalacain, I., Zapelena, M. J., Paz de Peña, M., Astiasarán, I., & Bello, J.
(1997a). Use of lipase from Rhizomucor miehei in dry-fermented
sausages elaboration: microbial, chemical and sensory analisis. Meat
Science, 45, 99–105.
Zalacain, I., Zapelena, M. J., Paz de Peña, M., Astiasarán, I., & Bello, J.
(1997b). Application of Lipozyme 10,000 L (from Rhizomucor miehei)
in dry-fermented sausage technology: study in a pilot plant and at the
industrial level. Journal of Agricultural and Food Chemistry, 45,
1972–1976.
 MEAT
SCIENCE
Meat Science 77 (2007) 90–96
www.elsevier.com/locate/meatsci

Interventions to reduce/eliminate Escherichia coli O157:H7 in

ground beef
M. Koohmaraie *, T.M. Arthur, J.M. Bosilevac, D.M. Brichta-Harhay, N. Kalchayanand,
S.D. Shackelford, T.L. Wheeler
Roman L. Hruska US Meat Animal Research Center, ARS, USDA, Clay Center, P.O. Box 166, Spur 18-D, NE 68933-0166, USA

Received 9 March 2007; received in revised form 30 March 2007; accepted 2 April 2007

Abstract

The Escherichia coli O157:H7 (E. coli O157:H7) outbreak in the Northwestern United States ushered in an era that has dramatically
changed the way beef processors in the United States convert live cattle into meat. Unprecedented cooperation among the beef proces-
sors and massive investment in research by the US government and the beef industry have resulted in an acceptable level of control of
E. coli O157:H7 in ground beef. The evidence to support the progress in control of E. coli O157:H7 is the CDC data for reduction in
human illness as well as the dramatic reduction in the number of E. coli O157:H7-positive samples in USDA-FSIS ground beef moni-
toring. This manuscript highlights some of the recent findings from our laboratory on the control of E. coli O157:H7 in ground beef. We
have also summarized the key events/decisions/milestones that have contributed to the control of E. coli O157:H7 in ground beef in the
United States. While there is much to be done to bring E. coli O157:H7 under complete control in the beef sector of the food industry,
E. coli O157:H7 also is becoming a major issue in the fresh vegetable sector, as evidenced by the 2006 outbreaks in the United States. We
have discussed how the fresh vegetable industry can benefit from the beef industry’s experience to expedite the control of E. coli O157:H7
in their products.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Escherichia coli O157:H7; Non-O157 Shiga toxin-producing E. coli; Salmonella; Interventions; Test-and-hold

1. Introduction E. coli O157:H7-related illness per 100,000 populations.

The 1993 outbreak of Escherichia coli O157:H7 in the
Northwestern United States began an era of intense effort
to eliminate this pathogen from the red meat supply. The
US meat industry and government have invested millions
of dollars in research leading to control of E. coli
O157:H7. These efforts have been very successful, as
demonstrated by CDC reports of E. coli O157:H7-related
illnesses, which were 1.03, 0.9, and 1.06 per 100,000 pop-
ulations for 2003, 2004, and 2005, respectively (Anony-
mous, 2005). The Healthy People 2010 goal is 1.0
*
Corresponding author. Tel.: +1 402 762 4109; fax: +1 402 762 4111.
E-mail address: Mohammad.Koohmaraie@ARS.USDA.GOV (M.
Koohmaraie).
This progress is the result of the implementation of
research findings by the industry as well as a process
called test-and-hold (for review, Koohmaraie et al.,
2005).
In spite of all efforts and great vigilance by the beef pro-
cessing sector, there continued to be ground beef-related ill-
nesses and costly product recalls. Thus, in the late 1990s
some members of the industry began to implement the
test-and-hold process (Brabban, Nelsen, Kutter, Edring-
ton, & Callaway, 2004). To minimize the likelihood of
the finished product containing E. coli O157:H7, the raw
ground beef materials (beef trim) or finished ground beef
would be sampled, tested for presence of E. coli O157:H7
and if results were negative the product would be released
into commerce. If positive, the entire lot (typically 10,000
pounds for trim and 1 h of production or about 90,000

0309-1740/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.004
 M. Koohmaraie et al. / Meat Science 77 (2007) 90–96
pounds for ground beef) would be diverted to fully cooked
product, or rendered. Soon the rest of industry adopted
this costly process. The test-and-hold process costs the beef
processing sector millions of dollars annually.
Since the implementation of the test-and-hold process,
we at the US Meat Animal Research Center (USMARC)
have been conducting research to reduce the number of
positive samples identified by test-and-hold, thereby
increasing the safety of the food supply and decreasing
the economic burden on the beef processing sector. For
example, the knowledge that pathogens originate primarily
from hides and that they are transferred to carcasses during
the hide removal process (Barkocy-Gallagher et al., 2003;
Bosilevac et al., 2004; Bosilevac, Nou, Osborn, Allen, &
Koohmaraie, 2005; Bosilevac, Shackelford, Brichta, &
Koohmaraie, 2005; Nou et al., 2003) has greatly helped
the industry in reducing the number of positive samples
identified by test-and-hold. Some processors have imple-
mented the hide-on carcass wash intervention system, some
have spent a great deal of time training their employees to
properly remove hides so as to minimize transfer of patho-
gens from hides to carcasses, and others are attempting to
implement new interventions based on this knowledge.
This manuscript describes additional contributions from
our laboratory aimed at further reducing E. coli O157:H7
from the red meat supply.
2. Escherichia coli O157:H7
Escherichia coli serotype O157:H7 and its significance to
public health and commerce are well documented. This
bacterium is capable of producing large quantities of toxins
(Shiga toxins) that cause severe damage to the intestinal
lining and is recognized as the causative agent of outbreak
hemorrhagic diarrhea (Johnson, Lior, & Bezanson, 1983;
Riley et al., 1983). It has been established that E. coli
O157:H7 can be found in animals and is associated with
contaminated meat (Chapman et al., 1993; Hancock, Bes-
ser, Lejeune, Davis, & Rice, 2001). For a more thorough
discussion of all aspects of E. coli O157:H7, the reader is
referred to a number of review papers that have been writ-
ten on the subject, including Acheson (2000) and Koohma-
raie et al. (2005).
3. Development of methodology to allow routine
enumerations of E. coli O157:H7
Several years ago, we recognized the significance of enu-
meration data and we began to use the available enumera-
tion method at the time, most probable number (MPN), to
enumerate E. coli O157:H7 on hides and to quantify the
effect of interventions in reducing E. coli O157:H7. We
decided to develop alternative enumeration methodology
because: (1) the MPN methodology was very expensive,
costing about $100 per sample in 2001, thereby dramati-
cally reducing the number of samples that can be subjected
to this analysis; (2) MPN results were often inconsistent,
91
raising doubt about the validity of the results; and (3)
MPN assay is labor intensive. Therefore we began work
on the development of an enumeration method that does
not have the problems associated with MPN assay. We
now have developed and validated two high-throughput
methods for the enumeration of Salmonella and E. coli
O157:H7 from various sample types (feces, hides, carcass
sponges, ground beef, and beef trim) collected during beef
processing (Brichta-Harhay et al., 2007). These methods
can be performed in a fraction of the time and cost of other
culture-based enumeration methods such as MPN analy-
ses. Consequently, their use allows for routine testing and
quantification, thus providing useful information about
the effectiveness of intervention strategies. The availability
of accurate and inexpensive enumeration methods would
also aid those involved in food safety research, as evi-
denced by the example given below. The power of such
an assay became clear to us when we examined the efficacy
of hide washing to control E. coli O157:H7 levels on hides
(Arthur, Bosilevac, Brichta-Harhay, Kalchayanand et al.,
2007). We had clearly shown that hide wash interventions
were effective controls for E. coli O157:H7 (Bosilevac
et al., 2004; Bosilevac, Nou et al., 2005; Bosilevac, Shackel-
ford et al., 2005). The measure of effectiveness for these
hide interventions was based on reductions in prevalence.
Most current prevalence assays for E. coli O157:H7 are
quite sensitive for the target organism and will give positive
results even when only low levels of target cells (e.g., 10–
50 CFU) are present in the sample (Barkocy-Gallagher
et al., 2005). Therefore, hides harboring E. coli O157:H7
at 5 CFU/100 cm2 and at 50,000 CFU/100 cm2 will give
the same positive results.
The interventions cited above were shown to be effec-
tive in reducing E. coli O157:H7 hide prevalence of beef
cattle during processing, indicating that large reductions
in the E. coli O157:H7 load on cattle hides were occurring
(Bosilevac et al., 2004; Bosilevac, Nou et al., 2005; Bosil-
evac, Shackelford et al., 2005). These interventions, while
being quite thorough, may not be suitable for all beef pro-
cessing plants to implement, due to cost and space restric-
tions. Several years ago we tested the efficacy of washing
cattle hides in a cabinet, using 100 to 200 ppm chlorine.
Since we used the reduction in prevalence to determine
efficacy and did not observe a significant decrease in
E. coli O157:H7 prevalence, we erroneously concluded
that washing cattle hides with 100–200 ppm chlorine was
not a viable method of reducing E. coli O157:H7 on cattle
hides. With the advent of the new enumeration methods,
we re-examined the efficacy of this wash cabinet. This time
we saw dramatic reductions in E. coli O157:H7 on car-
casses in plants that were washing cattle hides with chlo-
rine. Having developed an accurate and inexpensive
method to enumerate E. coli O157:H7 and Salmonella,
we were able to demonstrate that although use of this hide
wash cabinet does not dramatically reduce hide prevalence
of E. coli O157:H7 and Salmonella, it does significantly
reduce the level of E. coli O157:H7 and Salmonella on
92 M. Koohmaraie et al. / Meat Science 77 (2007) 90–96
the hides of beef cattle during processing (Table 1, Arthur,
Bosilevac, Brichta-Harhay, Kalchayanand et al., 2007).
For example, of the 288 carcasses in the study, in 250 sam-
ples E. coli O157:H7 levels were less than the detection
limit (40 CFU/100 cm2), compared to only 187 samples
before the hides were washed with 100 ppm chlorine.
Before washing hides with chlorine, 51 samples had
>100 E. coli O157:H7/100 cm2 and after chlorine washing
only 14 samples had >100 E. coli O157:H7/100 cm2. These
results are extremely significant when one considers the
following common knowledge. We believe all the interven-
tions that are used in beef processing plants have the
capacity to control a given but unknown level of E. coli
O157:H7. As long as the load on the incoming cattle is
within the capacity of these interventions, we do not
expect the contamination of the resultant trim/ground
beef to be at a level that can be detected in the test-and-
hold process. The problem becomes very significant and
costly (due to the amount of product that is rejected by
the test-and-hold process) when the load on the incoming
cattle exceeds the capacity of the interventions. For this
reason we have always searched for interventions that
would bring the levels of the incoming load in line with
the capacity of the current intervention. Clearly, hide
washing with 100–200 ppm chlorine is such an interven-
tion. We have given all the details to a wash cabinet man-
ufacturer and fully expect wider implementation of this
very effective hide wash intervention to control E. coli
O157:H7 on the hide, and thereby the carcass, and ulti-
mately in ground beef.
We look forward to seeing the impact of widespread
use of our affordable enumeration methods and, more
importantly, the role they will play in further control of
E. coli O157:H7 in red meat, particularly by the small
and medium sized beef processing plants. We have also
shown that our E. coli O157:H7 and Salmonella enumera-
Table 1
E. coli O157:H7 level and prevalence (%) before and after hide wash
cabinet
na Total
288
Enumeration
Before cabinetb 35.1
After cabinet 13.2d
Prevalence
Before cabinetc 97.6
After cabinet 89.6d
a
Number of cattle hides sampled each trip.
b
Enumeration data are presented as the percentage of total samples that
were above the limit of detection for enumeration. Enumeration limit of
detection for hide samples was P40 CFU/100 cm2.
c a major source of contamination is that it removes the
Prevalence values given are the number of hide samples that were
positive divided by the total number of hides sampled and expressed as
percentage.
d
Value differs significantly (P 6 0.05) from value before hide wash
cabinet.
tion methods can be used to enumerate E. coli O157:H7 in
compost and aged manure (unpublished data) and Salmo-
nella in poultry carcass rinses (Brichta-Harhay et al.,
2007).
4. Transportation and lairage environment effects on
prevalence and levels of E. coli O157:H7 on hides and
carcasses of beef cattle at processing
It is our assessment that if the current knowledge about
the source of E. coli O157:H7 and how it is transferred to
carcasses and ultimately to ground beef is put into prac-
tice, there is not much more that can be done to reduce
the incidence of E. coli O157:H7 in ground beef during
harvest (from the time that cattle are stunned until pro-
cessing is completed). For a variety of reasons, we became
interested in determining the relative role of lairage at the
processing plant (from off loading of cattle until stunning)
in E. coli O157:H7 contamination of carcasses (Arthur,
Bosilevac, Brichta-Harhay, Guerini, Kalchayanand et al.,
2007). During the summer of 2004 and at the conclusion
of a 9-mo study of E. coli O157:H7 in a feedlot, we sam-
pled the hides of 286 cattle and transported them to a com-
mercial beef processing plant and sampled the hides again
at the plant (after stunning and exsanguination). The prev-
alence of E. coli O157:H7 on hides increased from 50.3%
to 94.4%, between loading onto tractor-trailers at the feed-
lot and before hide removal in the processing plant. Prior
to transport, nine animals were found to have E. coli
O157:H7 at high concentrations (>0.4 CFU/cm2) on their
hides. When sampled at the slaughter facility, the number
of animals with high hide concentrations of E. coli
O157:H7 had increased to 70. Overall, only 29% (221 of
764) of the E. coli O157:H7 isolates collected post-harvest
were found to match pulsed field gel electrophoresis types
collected prior to transport. The results suggested trans-
port to and lairage at processing plants can lead to
increases in the prevalence and levels of E. coli O157:H7
contamination on hides and the number of E. coli
O157:H7 pulsed field gel electrophoresis types associated
with the animals. To confirm these findings, the study
was repeated. This time we took cattle from the same feed-
lot (known PFGE patterns) to three different commercial
beef processing plants (Arthur et al., 2007b). We con-
firmed the results of the previous year and concluded that
cattle holding areas at the processing plants are a major
source of hide contamination and it is quite possible that
cattle hides could be free of E. coli O157:H7 when leaving
the feedlot and become highly contaminated with E. coli
O157:H7 at the plant.
There are a number of major outcomes from this study.
The first outcome of knowing that the plant holding area is
incentive for feedlot operators to use interventions at the
feedlot (and as of now there is no effective pre-harvest or
feedlot intervention) because the likelihood of their cattle
becoming re-contaminated at the plant is extremely high.
 M. Koohmaraie et al. / Meat Science 77 (2007) 90–96
The second outcome is that a given lot of cattle could
become the source of another pathogen such as multi-drug
resistant Salmonella. For example, in a plant that processes
cull dairy cows as well as steers and heifers, the steers and
heifers could be contaminated with MDR Salmonella,
which is more often associated with cull cows than fed beef.
Lastly is the realization that it is impossible to devise a
cost-effective intervention to prevent cross contamination
during holding of cattle at the plant. We are certain that
such an intervention could be implemented, but we are
equally certain that such an intervention would be cost pro-
hibitive. In a recently completed study (unpublished data),
hide intervention and/or sanitary hide removal were shown
to negate the lairage effect.
5. Efficacy of pre-harvest versus post-harvest interventions
We have stated that harvest is the most logical and effec-
tive step in the beef production system at which to maxi-
mally reduce E. coli O157:H7 (as well as other
pathogens) on cattle and, thereby, in ground beef (for
details see Koohmaraie et al., 2005). This is not to say that
we should not focus on pre-harvest controls. However, it is
our belief that after large investments in research to control
E. coli O157:H7 in live animals, to date we can not direct a
producer or a feedlot operator to a practice to allow its
control in the living animal. Some would use the highly
publicized E. coli O157:H7 outbreaks in the United States
in 2006, which were caused by E. coli O157:H7-contami-
nated spinach, and the yet-to-be-determined source(s) of
the outbreaks at Taco Bell and Taco John’s to make the
case for the importance of pre-harvest over post-harvest
control. In our assessment, the same case can be made
for the post-harvest control of E. coli O157:H7 in vegeta-
bles. Government and industry have invested heavily in
research that has led to the control of E. coli O157:H7 in
ground beef at the processing plant and the same can
and should be done to control E. coli O157:H7 in
vegetables.
One of the main reasons for our conclusion that post-
harvest is a far better control point than pre-harvest is
that pre-harvest controls will have to be pathogen specific,
for example, using vaccination or probiotics to control
E. coli O157:H7 in cattle. Focusing on one pathogen in
a pre-harvest setting may aid in control of that one path-
ogen, but there is far more than one pathogen that can
cause human disease. For a variety of reasons, E. coli
O157:H7 has been the focus. But there is no doubt that
as we become efficient in controlling E. coli O157:H7, if
our approach does not include the control of other patho-
gens, these other pathogens will then become the cause of
human illness. Pathogen-specific interventions are not via-
ble approaches. Other pathogens that have the potential
to become of concern are non-O157 Shiga toxin-produc-
ing E. coli (non-O157 STEC), MDR Salmonella, and Lis-
teria monocytogenes, to mention a few. This is contrasted
with the fact that most, if not all, post-harvest interven-
93
tions are effective against all pathogens and not just
E. coli O157:H7. What is needed to control E. coli
O157:H7 and other pathogens in fresh vegetables is a sys-
tematic study of their production and processing steps to
identify a monitoring and control system for all patho-
gens of concern. The reader is referred to Koohmaraie
et al. (2005) for more detailed reasoning of our preference
for post-harvest control.
6. Significant events/decisions/milestones that have
contributed to the control of E. coli O157:H7 in ground beef
in the United States
In Table 2, we listed key events that have played a piv-
otal role in the control of E. coli O157:H7 in the US meat
supply. Though close, these events are not listed in chrono-
logical order. Each of these events has contributed greatly
to the collective success of all segments of the beef indus-
try’s control of E. coli O157:H7 in our red meat supply
(Table 3), but in our assessment knowledge sharing has
played the most significant role (Table 2 items #5, 6, 8,
11, and 14). Perhaps the most significant of these items
with respect to knowledge sharing is the decision by the
presidents of the beef processing companies to allow shar-
ing of information and not to keep such knowledge as
trade secrets. As a result, organizations such as the
National Cattlemen’s Beef Association can take the lead
in forming the Beef Industry Food Safety Council (BIF-
SCo) or the Beef Safety Summits to bring the industry lead-
ers, people that operate beef processing plants, and the
scientists engaged in food safety research together to
address problems. The reader is s encouraged to go to
the BIFSCo website (www.BIFSCo.org). The American
Meat Institute Foundation holds ‘‘Best Practices Work-
shops’’ for similar purposes, although these are more direc-
ted to employees who are in charge of operating beef
processing plants.
The results presented in Fig. 1 were collected by
USMARC scientists and represent one of many such exam-
ples of knowledge-sharing within the industry. The decision
to conduct this study was made at a Beef Safety Summit
and was carried out at the request of the participating
plants and for the purpose of bench-marking. Bench-mark-
ing is routinely used to improve the process quality by all
plants. Results in Fig. 1 were collected from two different
beef processing plants. Since we had established that hide
is the major source of E. coli O157:H7 and that the pro-
cesses involved in the removal of the hide determine how
much E. coli O157:H7 is transferred to the carcass, we sam-
pled the hides (to determine the bacterial load as presented
for slaughter) and then sampled the carcasses right after
hide removal and before any intervention (to determine
the extent of bacterial transfer from the hide onto the car-
cass). Such data demonstrate which plant has the lowest
rate of transfer and by learning how they achieve the low
rate and sharing that information with the others, they
would all improve.
94 M. Koohmaraie et al. / Meat Science 77 (2007) 90–96

Table 2
List of significant events/decisions/milestones that have contributed to the control of E. coli O157:H7 in ground beef in United States
1 Jack-in-the-Box outbreak in Northwestern United States (1993).
2 Industry begins the very first screening for E. coli O157:H7 in raw materials and finished products on a very limited basis (1994).
3 FSIS begins rigorous enforcement of zero tolerance (1994).
4 Mike Taylor (FSIS Administrator) declares E. coli O157:H7 to be an adulterant in ground beef in 1994 and five years later includes beef trim as
well (1994).
5 Ben Nelson (Nebraska Governor) convenes the ‘‘Governor’s Conference.’’ (1995)
6 The National Livestock and Meat Board creates the ‘‘Blue Ribbon Task Force on O157:H7’’ (1994)’’.
7 USDA mandates the implementation of HACCP (1996).
8 Beef Industry Food Safety Council (BIFSCo) is formed to bring industry together to develop unified plans to control Escherichia coli O157:H7 in
red meat (1997).
9 Availability of rapid tests for Escherichia coli O157:H7 detection (begins in 1993; more widespread adoption in 1998).
10 Implementation of acid, hot water, steam vacuum and other efficacious interventions (beginning in 1994 and continuing to date).
11 The Presidents of the major companies of the beef processing segment agree that food safety is a non-competitive area and encourage collaboration
and knowledge sharing (1998).
12 Implementation of test-and-hold (beginning in 1995).
13 Industry searches for and implements additional interventions to reduce product loss due to test-and-hold. This process continues to date
(continuous).
14 BIFSCo organizes the first E. coli O157:H7 summit involving all segments of the beef industry (2003).
15 Development of ‘‘Best Practices’’ by BIFSCO, American Meat Institute and National Meat Association (1998).
16 The scientists at the US Meat Animal Research Center demonstrate that hide is the source of Escherichia coli O157:H7 in red meat and develop
hide wash intervention (2002).
17 Beef processors install hide-on carcass wash cabinets or intensive training of the employees to remove hide properly to minimize transfer of
pathogens from hide onto the carcass (2003).
18 Elsa Murano (USDA Under Secretary) announces policy that government would not challenge a company’s negative E. coli O157:H7 results, even
if there are positives for the same production day, provided the company conducts 100% testing of materials for raw comminuted products (2002).

Table 3 Hide and Pre-evisceration Daily Prevalence

US Department of Agriculture’s Food Safety and Inspection Service Hide Pre-evisceration
(USDA-FSIS)–Microbiological results of raw ground beef products 90.0%
analyzed for E. coli O157:H7 80.0%
Year No. of positives No. tested % Positive 70.0%
1994 0 891 0.0 60.0%
1995 3 5407 0.05
% Positive

50.0%
1996 4 5703 0.07
1997 4 6065 0.07 40.0%
1998 14a 8080 0.17
30.0%
1999 32b 7785 0.4
2000 55 6375 0.86 20.0%
2001 59 7010 0.84 10.0%
2002 55 7025 0.78
2003 20 6584 0.30 0.0%
Day 1 Day 2 Day 3 Day 1 Day 2 Day 3
2004 14 8010 0.17 Plant 1 Plant 2
2005 19 10976 0.17
2006 20 11779 0.17 Fig. 1. The prevalence of E. coli O157:H7 on hides and carcasses right
a after hide removal in two beef processing plants. Plants were sampled on
During October 1997, the amount analyzed was increased from a 25-g
three consecutive days. Total number of samples per plant was 300 each of
sample to a 325-g sample to provide increased detection sensitivity.
b hide and carcass samples.
On September 3, 1999, a new selection and detection method was
introduced to further increase test sensitivity.

7. Other pathogens of concern
We believe that it is in the best interests of food-related
industries to anticipate the emerging pathogens and do all
of the necessary work required for their control prior to
their becoming an issue. Simply put, this means industry
needs to be proactive rather than reactive. Salmonella in
general and MDR Salmonella specifically, non-O157
Shiga toxin-producing E. coli (Non-O157 STEC), Clos-
tridium difficile, and Mycobacterium avium Paratuberculo-
sis (MAP) are some of the organisms that are being
studied. The objective is to learn as much as possible
about these and other organisms that can cause human
illness and develop effective controls before they become
the cause of human illness and regulation is forced as a
result. Clearly, if those involved in the fresh produce
industry had taken this approach, they would not have
experienced the devastating economic losses and been
faced with the regulations that have/will follow these
events.
 M. Koohmaraie et al. / Meat Science 77 (2007) 90–96
8. Efficacy of post-harvest interventions against current and
emerging pathogens
As discussed above, we prefer post-harvest interventions
over pre-harvest interventions and have provided some of
the reasoning for such a preference. Another reason is
the universality of the post-harvest interventions as
opposed to specificity of pre-harvest interventions.
Although the notion stated in the previous sentence seems
logical (i.e., heat kills all bacteria but vaccination would be
bacterial specific), at the request of the industry we con-
ducted an experiment to confirm the concept (Arthur
et al., 2007a). Beef surface tissues were inoculated with var-
ious organisms, and after allowing an appropriate amount
of time for attachment to take place, as well as simulating
the length of time a bacterium is exposed to the carcass sur-
face in a commercial beef processing plant before any inter-
ventions, the beef surface tissue was subjected to a number
of interventions for a period of time equivalent to that used
in commercial beef processing plants. The organisms used
included generic Salmonella (Newport and Typhimurium),
MDRSalmonella (Newport and Typhimurium), and vari-
ous E. coli O157:H7 isolates with variations with respect
to the ability of each to cause human disease. We used a
number of interventions that are currently used in industry
(hot water, organic acids) and some that are not currently
used in industry (electrolyzed water, ozone, and Fresh FX).
The results indicated that MDR Salmonella is reduced as
effectively as E. coli O157:H7 when treated with antimicro-
bial interventions currently in use at most US beef process-
ing plants (Arthur et al., 2007a).
9. Summary and conclusions
The outbreak of E. coli O157:H7 associated with the con-
sumption of undercooked hamburgers in the Northwestern
United States in 1992 and 1993 was the event that forced the
government and the industry to control this pathogen in the
US ground beef supply. The CDC data indicates that, for
all practical purposes, as a nation we have met our objective
(Healthy People 2010) to have only one case of E. coli
O157:H7-related foodborne illness per 100,000 populations.
We have outlined the key events that led to this success and
have suggested that the principal reason for the control of
E. coli O157:H7 in ground beef has been the sharing of
knowledge obtained by massive investment in research by
the government and the meat industry. Such efforts con-
tinue. This collaborative effort should be used as a model
for other sectors of the food industry to control whatever
issue is facing that particular sector.
Acknowledgements
We are grateful to Dell Allen, Tim Biela, Dennis John-
son, Randy Huffman, and Ranzell ‘‘Nick’’ Nickelson, II,
for their assistance with Table 2 and Carol Grummert for
her editorial assistance.
95
References
Acheson, D. W. (2000). How does Escherichia coli O157:H7 testing in
meat compare with what we are seeing clinically? Journal of Food
Protection, 63, 819–821.
Anonymous. (2005). Preliminary FoodNet data on the incidence of infection
with pathogens transmitted commonly through food–10 sites (Vol. 54).
United States, 2004. Morbidity and Mortality Weekly Report. (pp. 352–
356).
Arthur, T. M., Bosilevac, J. M., Brichta-Harhay, D. M., Guerini, M. N.,
Kalchayanand, N., Shackelford, S. D., et al. (2007). Transportation
and lairage environment effects on prevalence and levels of Escherichia
coli O157:H7 on hides and carcasses of beef cattle at processing.
Journal of Food Protection, 70, 280–286.
Arthur, T.M., Bosilevac, J.M., Brichta-Harhay, D.M., Kalchayanand, N.,
Shackelford, S.D., Wheeler, T.L. et al. (2007). Effects of a minimal hide
wash cabinet on the levels and prevalence of E. coli O157:H7 and
Salmonella on the hides of beef cattle at slaughter. Journal of Food
Protection. (accepted for publication).
Arthur, T.M., Kalchayanand, N., Bosilevac, J.M., Guerini, M.N.,
Brichta-Harhay, D.M., Shackelford, S.D. et al., (2007a). Comparison
of effects of antimicrobial interventions on multi-drug resistant
Salmonella, non-multi-drug resistant Salmonella, and Escherichia coli
O157:H7 (in preparation).
Arthur, T.M., Kalchayanand, N., Bosilevac, J.M., Guerini, M.N.,
Brichta-Harhay, D.M., Shackelford, S.D. et al., (2007b). Source
tracking of Escherichia coli O157:H7 and Salmonella on the hides and
carcasses of beef cattle at processing (in preparation).
Barkocy-Gallagher, G. A., Arthur, T. M., Rivera-Betancourt, M.,
Nou, X., Shackelford, S. D., Wheeler, T. L., et al. (2003).
Seasonal prevalence of Shiga toxin-producing Escherichia coli,
including O157:H7 and non-O157 serotypes, and Salmonella in
commercial beef processing plants. Journal of Food Protection, 66,
1978–1986.
Barkocy-Gallagher, G. A., Edwards, K. K., Nou, X., Bosilevac, J. M.,
Arthur, T. M., Shackelford, S. D., et al. (2005). Methods for
recovering Escherichia coli O157:H7 from cattle fecal, hide, and
carcass samples: Sensitivity and improvements. Journal of Food
Protection, 68, 2264–2268.
Bosilevac, J. M., Arthur, T. M., Wheeler, T. L., Shackelford, S. D.,
Rossman, M., Reagan, J. O., et al. (2004). Prevalence of Escherichia
coli O157 and levels of aerobic bacteria and Enterobacteriaceae are
reduced when hides are washed and treated with cetylpyridinium
chloride at a commercial beef processing plant. Journal of Food
Protection, 67, 646–650.
Bosilevac, J. M., Nou, X., Osborn, M. S., Allen, D. M., & Koohmaraie,
M. (2005). Development and evaluation of an on-line hide decontam-
ination procedure for use in a commercial beef processing plant.
Journal of Food Protection, 68, 265–272.
Bosilevac, J. M., Shackelford, S. D., Brichta, D. M., & Koohmaraie, M.
(2005). Efficacy of ozonated and electrolyzed oxidative waters to
decontaminate hides of cattle before slaughter. Journal of Food
Protection, 68, 1393–1398.
Brabban, A. D., Nelsen, D. A., Kutter, E., Edrington, T. S., & Callaway,
T. R. (2004). Approaches to controlling Escherichia coli O157:H7, a
foodborne pathogen and an emerging environmental hazard. Environ-
mental Practice, 6, 208–229.
Brichta-Harhay, D.M., Arthur, T.M., Bosilevac, J.M., Guerini,
M.N., Kalchayanand, N., & Koohmaraie, M. (2007). Enumer-
ation of Salmonella and Escherichia coli O157:H7 in ground
beef, cattle carcass, hide and fecal samples using direct plating
methods. Journal of Applied Microbiology. (accepted for
publication).
Chapman, P. A., Siddons, C. A., Wright, D. J., Norman, P., Fox, J., &
Crick, E. (1993). Cattle as a possible source of verocytotoxin-
producing Escherichia coli O157 infections in man. Epidemiology and
Infection, 111, 439–447.
96 M. Koohmaraie et al. / Meat Science 77 (2007) 90–96
Hancock, D., Besser, T., Lejeune, J., Davis, M., & Rice, D. (2001). The
control of VTEC in the animal reservoir. International Journal of Food
Microbiology, 66, 71–78.
Johnson, W. M., Lior, H., & Bezanson, G. S. (1983). Cytotoxic
Escherichia coli O157:H7 associated with haemorrhagic colitis in
Canada. Lancet, 76, 8314–8315.
Koohmaraie, M., Arthur, T. M., Bosilevac, J. M., Guerini, M., Shackel-
ford, S. D., & Wheeler, T. L. (2005). Post-harvest interventions to
reduce/eliminate pathogens in beef. Meat Science, 71, 79–91.
Nou, X., Rivera-Betancourt, M., Bosilevac, J. M., Wheeler, T. L.,
Shackelford, S. D., Gwartney, B. L., et al. (2003). Effect of chemical
dehairing on the prevalence of Escherichia coli O157:H7 and the levels
of aerobic bacteria and Enterobacteriaceae on carcasses in a commer-
cial beef processing plant. Journal of Food Protection, 66, 2005–2009.
Riley, L. W., Remis, R. S., Helgerson, S. D., McGee, H. B., Wells, J. G.,
Davis, B. R., et al. (1983). Hemorrhagic colitis associated with a rare
Escherichia coli serotype. New England Journal of Medicine, 308,
681–685.
 Meat Science 77 (2007) 97–104
www.elsevier.com/locate/meatsci

Application of proteomics to understand the molecular mechanisms

behind meat quality
a,¤
Kristin Hollung , Eva Veiseth a, Xiaohong Jia a,b, Ellen Mosleth Færgestad a,
Kjell Ivar Hildrum a
a
Matforsk AS, Osloveien 1, N-1430 Ås, Norway
b
Department of Chemistry, Biotechnology and Food Science, Norwegian University of Life Sciences, P.O. Box 5003, N-1432 Ås, Norway

Received 12 March 2007; received in revised form 21 March 2007; accepted 21 March 2007

Abstract

The proteome is expressed from the genome, inXuenced by environmental and processing conditions, and can be seen as the molecular
link between the genome and the functional quality characteristics of the meat. In contrast to traditional biochemical methods where one
protein is studied at a time, several hundred proteins can be studied simultaneously. Proteomics is a promising and powerful tool in meat
science and this is reXected by the increasing number of studies emerging in the literature using proteomics as the key tool to unleash the
molecular mechanisms behind diVerent genetic backgrounds or processing techniques of meat. Thus understanding the variations and
diVerent components of the proteome with regard to a certain meat quality or process parameter will lead to knowledge that can be used
in optimising the conversion of muscles to meat. At present, there has been focus on development of techniques and mapping of proteo-
mes according to genotypes and muscle types. In the future, focus should be more towards understanding and Wnding markers for meat
quality traits. This review will focus on the methods used in the published proteome analyses of meat, with emphasis on the challenges
related to statistical analysis of proteome data, and on the diVerent topics of meat science that are investigated.
© 2007 Elsevier Ltd. All rights reserved.

Keywords: Proteomics; Proteome analysis; Two-dimensional electrophoresis; 2-DE; Mass spectrometry; Muscle proteome; Protein fractionation

1. Introduction lated by a large number of factors such as environmental

In a situation with increased competition with other
foods, a predictable meat quality is essential to the meat
industry, with tenderness and juiciness being the most
important quality attributes as deWned by the consumers
(Hildrum & Tornberg, 1998; Miller, HuVman, Gilbert,
Hamman, & Ramsey, 1995). These traits are inXuenced
both by genetics, environmental factors and processing
conditions. However, the underlying molecular mecha-
nisms are far from understood. While the genes remain con-
stant during the lifetime of the animal, the expression of the
genes to mRNA and proteins is very dynamic and is regu-
*
Corresponding author. Tel.: +47 64 97 01 42; fax: +47 64 97 03 33.
E-mail address: kristin.hollung@matforsk.no (K. Hollung).
and processing conditions. The proteins expressed from the
genome may thus be viewed as the mirror image of the gene
activity.
The proteome is the protein complement of the genome
and consists of the total amount of proteins expressed at a
certain time point (Wilkins et al., 1996). In contrast to the
genome, the proteome is continuously changing according
to factors inXuencing on either protein synthesis or degra-
dation. Thus, analysing the proteome can be viewed as
analysing snap-shots into a system in constant change. In
this regard, the proteome can be seen as the molecular link
between the genome and the functional quality characteris-
tics of the meat, being expressed from the genome under the
given environmental/processing conditions. While the
genome contains the information on which genes are

0309-1740/$ - see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.03.018
98 K. Hollung et al. / Meat Science 77 (2007) 97–104

available, the proteome contains information on which
genes are actually being expressed. Thus understanding the
variations and diVerent components of the proteome with
regard to a certain quality or processing parameter will lead
to knowledge that can be used in optimising the conversion
of muscles to meat (Bendixen, 2005; Bendixen et al., 2005).
2. Tools for proteomics
Proteomics are the tools used to analyse the proteomes.
Over the last decade there have been signiWcant improve-
ments of methods in this Weld, largely driven by medical sci-
ence, as a demand for understanding the functions of the
genomes following the large genome sequencing projects.
While the analytical tools for characterising the genomes
have reached a high satisfaction, the corresponding tech-
niques for proteome analyses still need much reWnements
and developments. Most of the proteomics tools are based
on protein separation in at least two dimensions, using
either chromatographic methods or electrophoresis, and is
usually followed by the use of mass spectrometry (MS).
Detailed descriptions of these methods are beyond the scope
of this paper and are reviewed elsewhere (Aebersold, 2003;
Aebersold & Mann, 2003; Bendixen, 2005; Gorg, Weiss, &
Dunn, 2004; Rabilloud, 2002; Reinders, Lewandrowski,
Moebius, Wagner, & Sickmann, 2004; Righetti, Castagna,
Antonioli, & Boschetti, 2005; Stasyk & Huber, 2004; Steen
& Mann, 2004). However, we will focus on the methods for
two-dimensional gel electrophoresis (2-DE) as this has so
far been the method most widely used in meat science, with
special emphasis on the sample preparation and analysis of
2-DE data. A schematic diagram of the work-Xow of a pro-
teome analysis is shown in Fig. 1. Usually, the Wrst step is to
set up an experiment with diVerent animals, treatments or
muscles with diVerent quality traits e.g. colour or tenderness
(Step 1). This is followed by protein extraction, 2-DE, image
analysis and statistical analysis (Step 2–6). Evaluation of the
data and selection of signiWcantly changed proteins are criti-
cal steps in the experiments (Step 7–8). Furthermore, identi-
Wcation of the proteins by MS and interpretation of the
results (Step 9–10) often lead to new hypotheses and new
cycles of proteome analyses to be performed.

Fig. 1. Schematic drawing of the diVerent steps in the work-Xow in proteome analysis using 2-DE and mass spectrometry. (1) Animal or sample chosen for
analysis, (2) sample extraction, (3) isoelectric focusing (IEF), (4) SDS-PAGE, 2-dimensional electrophoresis, (5) alignments and comparisons of 2-DE
images, (6) data analysis, (7) data interpretation and selection of signiWcantly changed proteins, (8) extraction of signiWcantly changed protein spots, (9)
identiWcation of protein spots by MALDI-TOF MS, (10) interpretation of the results.
 K. Hollung et al. / Meat Science 77 (2007) 97–104
2.1. Muscle protein fractionation
Several authors have tried to estimate the total number
of diVerent proteins expressed from a genome including
splice variants and post-translational modiWcations in a
eukaryotic cell, with numbers ranging from 100–500,000
(Righetti et al., 2005; Stasyk & Huber, 2004). These proteins
are localised in diVerent compartments. Some are part of
large protein aggregates, like the myoWbrillar proteins,
some are localised in membranes and others are enzymes
localised in the cytoplasma. With the help of pre-fraction-
ation of the proteins based on solubilisation and extraction
procedures, the enrichment of soluble sarcoplasmic or myo-
Wbrillar proteins can be achieved. These proteins will have
very diVerent chemical properties which can be used to iso-
late the proteins in diVerent fractions. DiVerent strategies
for fractionation of proteins are reviewed in Righetti et al.
(2005). Several issues should be considered before develop-
ing a sample preparation strategy. It is advisable to keep
sample preparation as simple as possible to avoid protein
losses. However, if only a subset of the proteins in the tissue
or cells is of interest, pre-fractionation can be employed
during sample preparation. Depending on the project and
hypothesis diVerent strategies for extraction and pre-frac-
tionation of muscle proteins should be considered. Typi-
cally, approximately 1000 diVerent proteins, or 10% of the
total number of proteins, can be analysed on one 2-DE
image.
The presence of high-abundance proteins in a tissue or
cell often masks low-abundance proteins and thus generally
prevents their detection and identiWcation in proteome
studies. The use of pre-fractionation methods can assist in
the identiWcation and detection of low-abundance proteins
that may ultimately prove to be informative biomarkers.
For comprehensive proteome analysis by 2-DE, pre-frac-
tionation is essential for the following reasons. First, by
partitioning the proteome into compartments, the complex-
ity of each compartment is dramatically reduced facilitating
spot identiWcation and quantitative analysis. Secondly,
there is a pronounced bias inherent in 2-DE towards abun-
dant proteins. This has the eVect of masking low-abun-
dance proteins. Pre-fractionation enriches low-abundance
proteins. Thirdly, the amount of any given protein that can
be resolved on a 2-DE is limited. Pre-fractionation allows
the proteins present in a particular fraction to be loaded at
high levels, further increasing the representation of low-
abundance proteins. Finally, relative to whole cell prepara-
tions, the number of proteins that are solubilised during the
diVerential extraction procedures is greatly increased yield-
ing a more comprehensive representation of the proteome.
Several established protein and peptide fractionation
techniques include stepwise extractions of proteins, immun-
odepletion, reverse phase or ion-exchange chromatography
and gel Wltration (Righetti et al., 2005). The choice of tech-
nique is greatly dependent on which subset of proteins that
is of interest. In muscle cells, structure proteins, such as
actin and tubulin are high abundant proteins and will dom-
99
inate in 2-DE. Fig. 2 shows the representative 2-DE gels of
a bovine LD sample following a sequential fractionation:
water-soluble proteins, salt-soluble proteins and remaining
proteins. For comparison, a total protein extract from the
same animal is also included in the Wgure.
2.2. Separation by 2-DE
Separation of proteins by 2-DE has been done for
decades. However, major technical improvements such as
the introduction of immobilized pH gradients have been
important for the reproducibility of the method, for a
review see (Gorg et al., 2000; Gorg et al., 2004). The infor-
mation that can be extracted from 2-DE is very informative
and provides information on which proteins that are pres-
ent as well as information on modiWcations. In short, the
proteins are Wrst separated by charge using isoelectric
focusing, and then the focused proteins are separated by
mass using SDS-PAGE. After separation, diVerent tech-
niques are used to visualise the proteins, resulting in a large
number of spots on a 2-DE image of the gel which is subse-
quently digitalized. It is important to remember that both
choice of protein extraction method, pH gradient in the iso-
electric focusing step, and staining technology will deter-
mine which proteins can be observed and analysed.
2.3. Handling of 2-DE images and statistical analysis of
proteome data from 2-DE
Although being very informative, the gel images are
complex consisting of a very high number of more or less
overlapping protein spots, where the position of the protein
spots may vary from one gel image to another. Further-
more, the staining intensity and background may be vari-
able throughout the gels and from one gel to another. Thus,
the process of analysing the gel images to search for the
information revealed by the proteins is a critical and com-
plex step of the process.
Several commercially available softwares are designed to
align and analyse the images from 2-DE experiments. How-
ever, due to the noisy appearance of the images this is not
an easy task, and improvements are still needed to get an
optimal analysis. In principle, there are two diVerent
approaches for matching data from one gel to another. One
approach is to detect spots in each gel and map the result-
ing spots from one gel to another. This approach is both
time consuming and it faces a number of challenges result-
ing in a signiWcant number of missing values in the data
analysis (Grove, Hollung, Uhlen, Martens, & Faergestad,
2006). An alternative approach is based on alignment of the
images instead of matching protein spots. When the images
are aligned, spots can be detected across all gels simulta-
neously using common boundaries around the spots for all
gel images (Luhn, Berth, Hecker, & Bernhardt, 2003).
Thereafter protein spot tables can be generated without
missing values. However, overlapping protein spots and
saturated protein spots are still major challenges for the
100 K. Hollung et al. / Meat Science 77 (2007) 97–104

Fig. 2. 2-DE images of proteins from bovine LD muscle separated by pH 4–7 and 12.5% SDS-PAGE. Fifty g protein was loaded on each gel; (a) total
protein extract in urea-buVer (7 M Urea, 2 M Thiourea, 2% CHAPS, 1% DTT, 0.5% IPG 3–10 buVer); (b) proteins soluble in TES-buVer (10 mM Tris, pH
7.6, 1 mM EDTA, 0.25 M Sucrose); (c) remaining proteins from (b) soluble in NaCl-buVer (10 mM Tris, pH 7.6, 0.5 M NaCl); (d) remaining proteins from
(c) extracted in urea-buVer, as in (a).

data analysis. A useful approach to overcome these chal-
lenges is to analyse the aligned gel images pixel by pixel.
Variation from one gel to another even in strongly overlap-
ping protein spots and in saturated protein spots can then
be detected.
Proteome data need special attention in analysis and sta-
tistical validation of the outcome. In contrast to traditional
experiments with few measurements/variables in many
samples/animals, the opposite is true for proteomics data.
Here the number of samples is usually limiting because of
work-load, but the amount of data that is collected from
each animal exceeds several hundred observations. This is
also a challenge in other –omics data, such as transcripto-
mics and must be handled with care during the statistical
analysis.
Several statistical approaches have been used to analyse
proteomics data. Multivariate analyses such as principal
component analysis (PCA) (Martens & Martens, 2000;
Næs, Isaksson, Fearn, & Davies, 2002) is now included in
several softwares for analysis of 2-DE experiments. Multi-
variate approaches have also been used for selection of sig-
niWcant changes in the 2-DE data according to the design
parameters (Jessen, Lametsch, Bendixen, Kjaersgard, &
Jorgensen, 2002; Jia et al., 2006; Kjaersgard, Norrelykke, &
Jessen, 2006). Assessment of hierarchical clustering meth-
odologies, commonly used in transcriptomics studies, has
also been performed on proteomics data (Meunier et al.,
2007). The diVerent statistical methods will shed light on
diVerent aspects of the proteomics data as has been dis-
cussed in several papers (Jacobsen et al., 2007; Maurer,
Feldmann, Bromme, & Kalenka, 2005).
2.4. Limitations of proteomics
As with all advanced methods it is necessary to ensure
the experimental design is made such that it is possible to
analyze the results. Choice of extraction method for the
proteins is determining which proteins that can be studied,
and proteins that are not extracted will thus not be consid-
ered. Very hydrophobic proteins, membrane proteins and
high MW proteins are often diYcult to solubilise and to
analyse by 2-DE (Fey & Larsen, 2001; Gorg et al., 2004). It
is also important to keep in mind that whatever method is
chosen for proteome analysis there is no protocol yet pro-
viding an analysis of the complete proteome in one run.
Several groups have tried to compare diVerent strategies on
the same samples ending up with partly overlapping results
(Bodenmiller, Mueller, Mueller, Domon, & Aebersold,
2007; Frohlich et al., 2006; McDonald et al., 2006). Usually,
a few hundred to several thousand proteins can be analysed
in one experimental setup, but still this is only a small part
of the entire proteome. Thus it is important to draw conclu-
sions based on the proteins that are actually under investi-
gation and not extrapolate the results to the proteins that
 K. Hollung et al. / Meat Science 77 (2007) 97–104
failed to be analysed. As discussed in the previous section
analysis of the data is not straightforward, and careful con-
sideration should be made to choose the most optimal
strategy.
3. Applications of proteomics in meat science
In meat science proteomics is a fairly new tool. However,
over the last years several studies are published where pro-
teomics shed light on diVerent aspects of meat, both during
the aging process and also in response to diVerent process-
ing conditions. Several species have been studied, such as
cattle, pork, lamb and chicken. In most of these studies 2-
DE has been used in combination with MS identiWcation of
speciWc proteins as the proteomics tool.
3.1. Proteome mapping
Several proteome studies have been performed aiming at
describing the diVerent proteins that are present in a meat
sample. This approach can be described as proteome map-
ping. A proteome mapping study of bovine semitendinosus
(ST) muscle using a combination of 2-DE and mass spec-
trometry allowed the detection of roughly 500 reproducible
protein spots (Bouley, Chambon, & Picard, 2004). Of these,
a total of 129 protein spots corresponding to 75 diVerent
gene products were identiWed by matrix-assisted laser
desorption ionisation-time of Xight (MALDI-TOF) MS.
Approximately 25% of the identiWed proteins were involved
in metabolic pathways, while 17%, 16%, and 14.5% were
involved in cell structure, cell defence, and the contractile
apparatus, respectively. However, how these proteins are
related to meat quality was not investigated in this study.
A mapping strategy has also been used to compare the
proteome expression of sarcoplasmic proteins between four
diVerent ovine muscles (Hamelin et al., 2006). Muscle sam-
ples were taken within 30 min after slaughter from longissi-
mus dorsi (LD), vastus medialis (VM), semimembranosus
(SM), and tensor fasciae latae (TL). After 2-DE and data
analysis, 77 protein spots were found to be diVerentially
expressed between the muscles, and 47 of these were identi-
Wed by MALDI-TOF MS. The VM muscle had a diVerent
protein expression than the other three muscles. This mus-
cle had the highest proportion of slow-twitch oxidative
Wbres and the highest abundance of enzymes involved in
oxidative metabolism and oxidative stress-related proteins.
The TL, on the contrary, had a reduced expression of most
of these proteins, which reXects that this muscle had the
highest rate of fast-twitch glycolytic Wbres.
3.2. Proteome changes due to genetic variations
Genetic variations may cause phenotypic diVerences that
can be studied using proteomics. Changes in expression of
bovine skeletal muscle proteins induced by hypertrophy
were studied by proteome analyses (Bouley et al., 2005).
The muscle hypertrophy was caused by a deletion in the
101
myostatin gene, and samples from the ST muscle were taken
from bulls that were heterozygote and homozygote for the
myostatin deletion and from control bulls. The 2-DE analy-
sis allowed the detection of 400 common protein spots, and
statistical analyses revealed 28 protein spots that diVered
between the control and the homozygote bulls, while only
one protein diVered between the control and the heterozyg-
otic bulls. Changes of proteins in the contractile apparatus
and metabolic enzymes indicate that the myostatin deletion
results in a shift towards a fast-twitch glycolytic muscle
type, which is in accordance with the results from the mus-
cle Wbre type analyses performed on the same animals. This
demonstrates that proteomics reXects the shifts in Wbre type
in the muscles.
In another study the eVect of a quantitative trait loci
(QTL) for muscle hypertrophy on sarcoplasmic proteins
expression in four ovine muscles was investigated (Hamelin
et al., 2006). Thirty minutes after slaughter, samples from
LM, VM, SM, and TL were taken for analyses by 2-DE.
Only three of the muscles (LM, SM, and TL) were hyper-
trophied due to the QTL, while the VM was normal. By use
of MALDI-TOF MS, 63 spots that were expressed diVeren-
tially between the genotypes were identiWed. These changes
included increased levels of enzymes involved in the glyco-
lytic metabolism, which indicates a switch towards a more
glycolytic muscle type. Moreover, several chaperon pro-
teins were also increased in abundance, which may indicate
a requirement for more building blocks to promote myoW-
bril assembly.
3.3. Post-mortem changes
Studying post-mortem changes in the proteome will lead
to an increased understanding of the biochemical mecha-
nisms behind meat quality traits, such as tenderness. A
number of studies have described how post-mortem degra-
dation of myoWbrillar protein may be involved in meat ten-
derness, and in particular, the degradation patterns of
contractile proteins have been described in detail (Geesink
& Koohmaraie, 1999; Hopkins & Taylor, 2002; Melody
et al., 2004; Taylor, Geesink, Thompson, Koohmaraie, &
Goll, 1995).
Lately, proteomics has been used to study changes
occurring in muscle during post-mortem storage. Total pro-
tein extracts from pork LD samples collected at 0, 4, 8, 24
and 48 h after slaughter revealed that 15 proteins were
changed, some increasing and some decreasing in abun-
dance after slaughter (Lametsch & Bendixen, 2001; Lame-
tsch, RoepstorV, & Bendixen, 2002). Several of these
proteins were identiWed as fragments of structural proteins
such as actin, myosin heavy chain and troponin T. Earlier
studies using one-dimensional SDS-PAGE have concluded
that actin is not degraded post-mortem (Bandman &
Zdanis, 1988; HuZonergan, Parrish, & Robson, 1995;
Koohmaraie, 1994). However, using 2-DE the resolution
allows for a better separation of the diVerent actin frag-
ments and demonstrates the potential of 2-DE. Some of the
102 K. Hollung et al. / Meat Science 77 (2007) 97–104
actin fragments observed in post-mortem pork muscle was
also signiWcantly correlated with tenderness. In addition,
several metabolic enzymes involved in energy metabolism
were altered in the pork muscle.
During post-mortem storage the calpain system is
believed to be important for degradation of myoWbrillar
proteins and development of tenderness (Goll, Taylor,
Christiansen, & Thompson, 1991; Koohmaraie, 1994,
1996). In a proteome study in pork LD muscle several of
the myoWbrillar substrates for -calpain were identiWed
(Lametsch, RoepstorV, Moller, & Bendixen, 2004). Among
the substrates that were degraded by in vitro incubation
with -calpain were desmin, actin, myosin heavy chain,
myosin light chain I, troponin T, tropomyosin 1, tropomy-
osin 4, thioredoxin and CapZ.
Changes in metabolic protein composition in biopsies
from live animals to post-mortem samples collected
shortly after slaughter in the cattle LD muscle revealed
that 24 protein spots were changed (Jia et al., 2006). This
reXects the contribution of several factors such as trans-
portation, lairage, stunning, exsanguination and dehiding
on the LD muscle proteome. IdentiWcation of the pro-
teins by MALDI-TOF/TOF MS revealed that a wide
range of metabolic enzymes and stress proteins increased
in abundance after slaughter. Several of these proteins
were glycolytic enzymes such as enolase, aldehyde dehy-
drogenase, phosphoglycerate kinase or enzymes involved
in oxidative metabolism such as ATP-speciWc succinyl-
CoA synthetase and isocitrate dehydrogenase. This sup-
ports an expected shift in energy metabolism in the mus-
cle post-mortem via the glycolytic pathway, and also an
increase in aerobic energy metabolism the Wrst hour after
slaughter.
Changes in the muscle proteome between slaughter and
24 h storage in bovine LD and ST muscles was observed
both between the sampling times and between the LD and
ST muscles (Jia, Hollung, Therkildsen, Hildrum, & Bend-
ixen, 2006). In this study 5 proteins were changed in both
muscles, namely coWlin, lactoylglutathione lyase, substrate
protein of mitochondrial ATP-dependent proteinase SP-22,
HSP27KDa and HSP20KDa. However, 15 proteins were
changed in either LD or ST muscles. These diVerences reX-
ect distinct metabolic and physiological functions of the
diVerent muscles.
3.4. Proteome changes due to pre-slaughter conditions
During the last few years, proteomics has also been
applied to investigate proteome changes induced by diVer-
ent pre-slaughter conditions. The Wrst example involves a
study of compensatory growth in pigs, which has been asso-
ciated with more tender meat (Kristensen, Therkildsen,
Aaslyng, Oksbjerg, & Ertbjerg, 2004). In a proteome study
of pigs at slaughter, seven proteins were changed according
to compensatory growth. Among them several stress pro-
teins and glycolytic proteins were decreased in abundance
(Lametsch et al., 2006). In the same study eight proteins
were aVected 48 h after slaughter including both structural
proteins and enzymes.
In another study, the eVect of post-mortem storage time (0,
12, and 72 h) and diVerent pre-slaughter treatments (trans-
ported immediately before slaughter vs. transported 12 h
before slaughter) of pigs on the LD muscle proteome were
analysed (Morzel et al., 2004). Most diVerences were found as
a result of post-mortem storage, and 37 spots varied with
storage time. Ten of these increased in abundance during the
storage period, such as fragments of actin, troponin T, and -
crystallin resulting from proteolysis. The pre-slaughter treat-
ments aVected 8 protein spots, while they had no eVect on the
proteolytic events taking place during the storage period.
3.5. Proteome changes associated with meat quality traits
In a study of proteome changes related to tenderness
(Warner Bratzler shear force) in pork LD muscle 6 proteins
were aVected (Lametsch et al., 2003). A great increase of
fragments of actin during post-mortem storage was
observed. Three actin fragments as well as a myosin heavy
chain fragment were correlated to shear force. Moreover,
myosin light chain II and the glycolytic enzyme triose phos-
phate isomerase were correlated to tenderness. This has also
been conWrmed by Hwang (Hwang, Park, Kim, Cho, & Lee,
2005). In another study, proteome analysis of SM muscle
from normal hams and from PSE-zones of defective hams
demonstrated a reduced proteolysis of troponin T, MLC 1,
and -crystallin in the defect muscles (Laville et al., 2005).
Recently, 2-DE and MS have also been used to investi-
gate the biochemical mechanisms behind variation in meat
colour (Hwang et al., 2005; Sayd et al., 2005). Comparison
of the sarcoplasmic proteome in pig SM muscle from two
groups of animals having a light or dark meat colour
revealed that 22 protein spots were diVerentially expressed
(Sayd et al., 2006). The animals, 12 animals in each group,
were selected from samples of 1000 pigs based on extreme
L¤-values. While the dark muscles had an increased abun-
dance of mitochondrial proteins, indicating a more oxida-
tive metabolism, the light muscles had an increased
abundance of cytosolic proteins involved in glycolysis.
Proteome analyses have also been used to study protein
changes in dry-cured hams. In one study, the myoWbrillar
proteins of raw meat and dry-cured hams after 6, 10 and 14
months of ripening were analysed by 2-DE (Di Luccia
et al., 2005). Both actin, tropomyosin and myosin light
chain disappeared during the ripening period, and were
almost completely hydrolysed after 12 months. In a pilot
study of Norwegian dry-cured hams we have earlier
observed a variation in the protein degradation pattern
between hams ripened for 6 months from diVerent produc-
ers (Sidhu, Hollung, & Berg, 2005).
3.6. MS-analysis of peptides
The majority of proteomics studies in meat science are
based on a combination of 2-DE and MS. However, recently
 K. Hollung et al. / Meat Science 77 (2007) 97–104
the occurrence of low-molecular weight peptides in bovine
pectoralis profundus muscle during post-mortem storage and
cooking were analysed directly by MS (Bauchart et al.,
2006). Samples were taken approximately 30 min after
slaughter (T0) and after 14 days of ageing. On day 14 sam-
ples were taken from both cooked (T14c) and uncooked
(T14) muscle. These extracts were subsequently analysed for
amino acid composition, and a combination of MALDI-
TOF MS and nano-LC-ion trap MS/MS analyses were used
to determine peptide composition and identiWcation, respec-
tively. The peak patterns from the MALDI-TOF MS analy-
sis were very reproducible between animals, and 7, 18 and 92
peptides were detected at T0, T14, and T14c, respectively.
None of the peptides found at T0 and T14 were identiWed,
however seven peptides corresponding to 5 diVerent pro-
teins were identiWed from the T14c samples. Three of these
proteins are known targets of post-mortem proteolysis (tro-
ponin T, nebulin, cypher protein) (Geesink & Koohmaraie,
1999; Morzel et al., 2004), while the other two proteins were
the connective tissue proteins procollagen types I and IV.
These are known to be very stable during meat ageing, how-
ever prolonged heating above 70 °C can promote collagen
peptide bond breakage and explain the occurrence of these
peptides in the cooked samples.
4. Future prospects
Proteomics has turned out to be a promising and power-
ful tool in meat science over the last years yielding informa-
tion on diVerences between muscles in a carcass.
Furthermore, the combined approach of 2-DE and MS has
demonstrated a great strength in studying metabolic altera-
tions, post-mortem proteolysis, and changes induced by
environmental and processing conditions. As a result an
increasing number of studies are emerging in the literature
using proteomics as the key tool to unleash the molecular
mechanisms behind diVerent genetic backgrounds or pro-
cessing techniques of meat. Although the genetic back-
ground is important, the major contribution to meat
quality is caused by processing and environmental condi-
tions. Thus proteomics could be the tool to reXect the
important mechanisms and contributions to development
of a satisfactory meat quality. In contrast to traditional
methods where only one or a few proteins are studied at a
time, several hundred proteins can be studied simulta-
neously by proteome analysis. By manipulations with the
protein extraction, diVerent proteins are enriched and this
should be carefully considered before designing the experi-
ments. By proteomics it is possible to investigate proteins
that originally were not included in the hypothesis of the
experiment and may thereby provide answers leading to
new hypotheses according to a certain treatment or quality.
At present, there has been focus on development of tech-
niques and mapping of proteomes according to genotypes
and muscle types. In the future, focus should be directed
towards understanding variation and Wnding markers for
meat quality traits.
103
While proteomics yields important information by itself,
the potential in linking information generated by this tech-
nique with other –omics techniques is vast. An integrated
functional genomics approach can be used to monitor
quantitative and qualitative diVerences in the transcrip-
tome, proteome, and metabolome, creating a powerful tool
to study gene function and cellular responses to external
stimuli. In meat science, this approach will be of great
importance, since meat quality is highly inXuenced by exter-
nal stimuli such as environmental and processing condi-
tions. However, linking and extracting information from
such large data matrices is a formidable task, and will
require signiWcant research attention in the coming years.
This implies diligent coordination between scientists that
masters the techniques and analyses in the diVerent Welds. If
we are successful in this endeavour, the future for proteo-
mics in meat research is bright. Proteomics has the poten-
tial to shift the understanding of molecular mechanisms
underlying meat quality a great leap forward.
Acknowledgement
This work has been supported by the Fund for the
Research Levy on Agricultural Products in Norway.
References
Aebersold, R. (2003). A mass spectrometric journey into protein and pro-
teome research. Journal of the American Society for Mass Spectrome-
try, 14, 685–695.
Aebersold, R., & Mann, M. (2003). Mass spectrometry-based proteomics.
Nature, 422, 198–207.
Bandman, E., & Zdanis, D. (1988). An immunological method to assess
protein-degradation in post-mortem muscle. Meat Science, 22, 1–19.
Bauchart, C., Remond, D., Chambon, C., Mirand, P. P., Savary-Auzeloux,
I., Reynes, C., et al. (2006). Small peptides (<5 kDa) found in ready-to-
eat beef meat. Meat Science, 74, 658–666.
Bendixen, E. (2005). The use of proteomics in meat science. Meat Science,
71, 138–149.
Bendixen, E., Taylor, R., Hollung, K., Hildrum, K. I., Picard, B., & Bouley,
J. (2005). Proteomics, an approach towards understanding the biology
of meat quality. In J. F. Hocquette & S. Gigli (Eds.), Indicators of milk
and beef quality (Vol. 112, pp. 81–94). Wageningen: Wageningen Aca-
demic Publishers.
Bodenmiller, B., Mueller, L. N., Mueller, M., Domon, B., & Aebersold, R.
(2007). Reproducible isolation of distinct, overlapping segments of the
phosphoproteome. Nature Methods, 4, 231–237.
Bouley, J., Chambon, C., & Picard, B. (2004). Mapping of bovine skeletal
muscle proteins using two-dimensional gel electrophoresis and mass
spectrometry. Proteomics, 4, 1811–1824.
Bouley, J., Meunier, B., Chambon, C., De Smet, S., Hocquette, J. F., &
Picard, B. (2005). Proteomic analysis of bovine skeletal muscle hyper-
trophy. Proteomics, 5, 490–500.
Di Luccia, A., Picariello, G., Cacace, G., Scaloni, A., Faccia, M., Liuzzi, V.,
et al. (2005). Proteomic analysis of water soluble and myoWbrillar pro-
tein changes occurring in dry-cured hams. Meat Science, 69, 479–491.
Fey, S. J., & Larsen, P. M. (2001). 2D or not 2D. Two-dimensional gel elec-
trophoresis. Current Opinion in Chemical Biology, 5, 26–33.
Frohlich, T., Helmstetter, D., Zobawa, M., Crecelius, A. C., Arzberger, T.,
Kretzschmar, H. A., et al. (2006). Analysis of the HUPO Brain Prote-
ome reference samples using 2-D DIGE and 2-D LC–MS/MS. Proteo-
mics, 6, 4950–4966.
104 K. Hollung et al. / Meat Science 77 (2007) 97–104
Geesink, G. H., & Koohmaraie, M. (1999). Postmortem proteolysis and
calpain/calpastatin activity in callipyge and normal lamb biceps femo-
ris during extended postmortem storage. Journal of Animal Science, 77,
1490–1501.
Goll, D. E., Taylor, R. G., Christiansen, J. A., & Thompson, V. F. (1991).
Role of proteinases and protein turnover in muscle growth and meat
quality. Proceedings of the Reciprocal Meat Conference, 44, 25–33.
Gorg, A., Obermaier, C., Boguth, G., Harder, A., Scheibe, B., Wildgruber,
R., et al. (2000). The current state of two-dimensional electrophoresis
with immobilized pH gradients. Electrophoresis, 21, 1037–1053.
Gorg, A., Weiss, W., & Dunn, M. J. (2004). Current two-dimensional elec-
trophoresis technology for proteomics. Proteomics, 4, 3665–3685.
Grove, H., Hollung, K., Uhlen, A. K., Martens, H., & Faergestad, E. M.
(2006). Challenges related to analysis of protein spot volumes from
two-dimensional gel electrophoresis as revealed by replicate gels. Jour-
nal of Proteome Research, 5, 3399–3410.
Hamelin, M., Sayd, T., Chambon, C., Bouix, J., Bibe, B., Milenkovic, D., et
al. (2006). Proteomic analysis of ovine muscle hypertrophy. Journal of
Animal Science, 84, 3266–3276.
Hildrum, K. I., & Tornberg, E. (1998). Nordfood: Managing the meat ten-
derization process: A Nordic cooperation project between Norwegian and
Swedish meat industry and research institutes. Oslo: Nordic Industrial
Fund.
Hopkins, D. L. & Taylor, R. G. (2002). Post-mortem muscle proteolysis
and meat tenderness.
HuZonergan, E., Parrish, F. C., & Robson, R. M. (1995). EVects of post-
mortem aging time, animal age, and sex on degradation of titin and
nebulin in bovine longissimus muscle. Journal of Animal Science, 73,
1064–1073.
Hwang, I. H., Park, B. Y., Kim, J. H., Cho, S. H., & Lee, J. M. (2005).
Assessment of postmortem proteolysis by gel-based proteome analysis
and its relationship to meat quality traits in pig longissimus. Meat Sci-
ence, 69, 79–91.
Jacobsen, S., Grove, H., Jensen, K. N., Sørensen, H. A., Jessen, F., Hollung,
K., et al. (2007). Multivariate analysis of two-dimensional gel electro-
phoresis protein patterns - Practical approaches. Electrophoresis, 28,
1289–1299.
Jessen, F., Lametsch, R., Bendixen, E., Kjaersgard, I. V., & Jorgensen, B.
M. (2002). Extracting information from two-dimensional electrophore-
sis gels by partial least squares regression. Proteomics, 2, 32–35.
Jia, X., Hildrum, K. I., Westad, F., Kummen, E., Aass, L., & Hollung, K.
(2006). Changes in enzymes associated with energy metabolism during
the early post-mortem period in longissimus thoracis bovine muscle
analyzed by proteomics. Journal of Proteome Research, 5, 1763–1769.
Jia, X., Hollung, K., Therkildsen, M., Hildrum, K. I., & Bendixen, E.
(2006). Proteome analysis of early post-mortem changes in two bovine
muscle types: M. longissimus dorsi and M. semitendinosis. Proteomics,
6, 936–944.
Kjaersgard, I. V., Norrelykke, M. R., & Jessen, F. (2006). Changes in cod
muscle proteins during frozen storage revealed by proteome analysis
and multivariate data analysis. Proteomics, 6, 1606–1618.
Koohmaraie, M. (1994). Muscle proteinases and meat aging. Meat Sci-
ence, 36, 93–104.
Koohmaraie, M. (1996). Biochemical factors regulating the toughening
and tenderization processes of meat. Meat Science, 43, S193–S201.
Kristensen, L., Therkildsen, M., Aaslyng, M. D., Oksbjerg, N., & Ertbjerg,
P. (2004). Compensatory growth improves meat tenderness in gilts but
not in barrows. Journal of Animal Science, 82, 3617–3624.
Lametsch, R., & Bendixen, E. (2001). Proteome analysis applied to meat
science: Characterizing postmortem changes in porcine muscle. Journal
of Agricultural and Food Chemistry, 49, 4531–4537.
Lametsch, R., Karlsson, A., Rosenvold, K., Andersen, H. J., RoepstorV, P.,
& Bendixen, E. (2003). Postmortem proteome changes of porcine mus-
cle related to tenderness. Journal of Agricultural and Food Chemistry,
51, 6992–6997.
Lametsch, R., Kristensen, L., Larsen, M. R., Therkildsen, M., Oksbjerg, N.,
& Ertbjerg, P. (2006). Changes in the muscle proteome after compensa-
tory growth in pigs. Journal of Animal Science, 84, 918–924.
Lametsch, R., RoepstorV, P., & Bendixen, E. (2002). IdentiWcation of pro-
tein degradation during post-mortem storage of pig meat. Journal of
Agricultural and Food Chemistry, 50, 5508–5512.
Lametsch, R., RoepstorV, P., Moller, H. S., & Bendixen, E. (2004). IdentiW-
cation of myoWbrillar substrates for mu-calpain. Meat Science, 68,
515–521.
Laville, E., Sayd, T., Sante-Lhoutellier, V., Morzel, M., Labas, R., Franck,
M., et al. (2005). Characterisation of PSE zones in semimembranosus
pig muscle. Meat Science, 70, 167–172.
Luhn, S., Berth, M., Hecker, M., & Bernhardt, J. (2003). Using standard
positions and image fusion to create proteome maps from collections
of two-dimensional gel electrophoresis images. Proteomics, 3, 1117–
1127.
Martens, H., & Martens, M. (2000). ModiWed Jack.knife estimation of
parameter uncertainty in bilinear modelling by partial least squares
regression (PLSR). Food Quality and Preference, 11, 5–16.
Maurer, M. H., Feldmann, R. E., Jr., Bromme, J. O., & Kalenka, A. (2005).
Comparison of statistical approaches for the analysis of proteome
expression data of diVerentiating neural stem cells. Journal of Proteome
Research, 4, 96–100.
McDonald, T., Sheng, S., Stanley, B., Chen, D., Ko, Y., Cole, R. N., et al.
(2006). Expanding the subproteome of the inner mitochondria using
protein separation technologies: one- and two-dimensional liquid chro-
matography and two-dimensional gel electrophoresis. Molecular and
Cellular Proteomics, 5, 2392–2411.
Melody, J. L., Lonergan, S. M., Rowe, L. J., Huiatt, T. W., Mayes, M. S., &
HuV-Lonergan, E. (2004). Early postmortem biochemical factors inXu-
ence tenderness and water-holding capacity of three porcine muscles.
Journal of Animal Science, 82, 1195–1205.
Meunier, B., Dumas, E., Piec, I., Bechet, D., Hebraud, M., & Hocquette, J.
F. (2007). Assessment of hierarchical clustering methodologies for pro-
teomic data mining. Journal of Proteome Research, 6, 358–366.
Miller, M. F., HuVman, K. L., Gilbert, S. Y., Hamman, L. L., & Ramsey, C.
B. (1995). Retail consumer acceptance of beef tenderized with calcium
chloride. Journal of Animal Science, 73, 2308–2314.
Morzel, M., Chambon, C., Hamelin, M., Sante-Lhoutellier, V., Sayd, T., &
Monin, G. (2004). Proteome changes during pork meat ageing follow-
ing use of two diVerent pre-slaughter handling procedures. Meat Sci-
ence, 67, 689–696.
Næs, T., Isaksson, T., Fearn, T., & Davies, T. (2002). A user-friendly guide to
multivariate calibration and classiWcation. Chichester: NIR Publications.
Rabilloud, T. (2002). Two-dimensional gel electrophoresis in proteomics:
old, old fashioned, but it still climbs up the mountains. Proteomics, 2,
3–10.
Reinders, J., Lewandrowski, U., Moebius, J., Wagner, Y., & Sickmann, A.
(2004). Challenges in mass spectrometry-based proteomics. Proteomics,
4, 3686–3703.
Righetti, P. G., Castagna, A., Antonioli, P., & Boschetti, E. (2005). Prefrac-
tionation techniques in proteome analysis: the mining tools of the third
millennium. Electrophoresis, 26, 297–319.
Sayd, T., Morzel, M., Chambon, C., Franck, M., Figwer, P., Larzul, C., et
al. (2006). Proteome analysis of the sarcoplasmic fraction of pig semi-
membranosus muscle: Implications on meat color development. Jour-
nal of Agricultural and Food Chemistry, 54, 2732–2737.
Sidhu, M. S., Hollung, K., & Berg, P. (2005). Proteolysis in Norwegian dry-
cured hams; preliminary results. In 51th ICOMST – the international
congress of meat science and technology, Baltimore, USA.
Stasyk, T., & Huber, L. A. (2004). Zooming in: Fractionation strategies in
proteomics. Proteomics, 4, 3704–3716.
Steen, H., & Mann, M. (2004). The ABC’s (and XYZ’s) of peptide
sequencing. Nature Reviews Molecular Cell Biology, 5, 699–711.
Taylor, R. G., Geesink, G. H., Thompson, V. F., Koohmaraie, M., & Goll,
D. E. (1995). Is Z-disk degradation responsible for postmortem tender-
ization. Journal of Animal Science, 73, 1351–1367.
Wilkins, M. R., Pasquali, C., Appel, R. D., Ou, K., Golaz, O., Sanchez, J. C.,
et al. (1996). From proteins to proteomes: large scale protein identiWca-
tion by two-dimensional electrophoresis and amino acid analysis. Bio-
technology (NY), 14, 61–65.
 MEAT
SCIENCE
Meat Science 77 (2007) 105–113
www.elsevier.com/locate/meatsci

Tenderness and oxidative stability of post-mortem muscles from

mature cows of various ages
Youling L. Xiong a,b,*, Oliver E. Mullins a, John F. Stika c, Jie Chen b,
Sue P. Blanchard a, William G. Moody a
a
Department of Animal and Food Sciences, University of Kentucky, Lexington, KY 40546, USA
b
College of Food Science, Southern Yangtze University, Wuxi, Jiangsu 214036, China
c
Certified Angus Beef Program, Wooster, OH 44691, USA

Received 8 March 2007; received in revised form 6 April 2007; accepted 11 April 2007

Abstract

Two experiments were conducted to investigate the influence of age of mature cows (2–4 yr, 6–8 yr, and 10–12 yr cows; n = 6 in each)
on beef quality. In Experiment 1, Longissimus dorsi (LD) steaks were stored at 3 C for 0–10 d. Steaks from more mature cows had an
increased (P < 0.05) Warner-Bratzler shear force (WBSF) and a slower troponin-T post-mortem degradation. Storage reduced WBSF in
all steak samples regardless of animal age. In Experiment 2, Semitendinosus (ST) and Semimembranosus (SM) patties were stored at 3 C
for 0–7 d simulating retail display. The rate of lipid oxidation during storage increased with animal age (P < 0.05) and was greater in ST
than in SM patties. However, myoglobin oxidation was minimally affected by animal age. Thus, advanced maturation not only inten-
sified cow meat toughness but also lowered its oxidative stability.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Cow; Beef; Proteolysis; Tenderness; Lipid oxidation; Myoglobin

1. Introduction ure of connective tissue collagen (i.e., type and degree of

Mature cows of advanced ages represent a significant
source or meat for the beef industry. In 2005, mature cattle
(mostly cows) accounted for 14.7% of the total beef cattle
slaughtered in the US (USDA, 2007). Meat from mature
cows is used mainly as ground beef. In order to enhance
its economical value, efforts are needed to understand
intrinsic factors that influence the quality attributes of beef
from this particular group of cattle.
One such intrinsic factor is proteolysis, which produces
muscle fiber fragmentation. Beef from physiologically
mature carcasses is generally perceived to be less tender
than that from young animals, for which the chemical nat-
*
Corresponding author. Address: Department of Animal and Food
Sciences, University of Kentucky, Lexington, KY 40546, USA. Tel.: +1
859 257 3822; fax: +1 859 257 5318.
E-mail address: ylxiong@uky.edu (Y.L. Xiong).
cross-linking) has been implicated (Shorthose & Harris,
1990; Lawrie, 1998). Furthermore, oxidative stability of
meat may be affected by animal age. Lipid oxidation that
takes place in post-mortem meat can be a main impediment
to the successful marketing of cow meat. Likewise, discol-
oration resulting from myoglobin oxidation could nega-
tively impact the utility of cow meat. Boccard et al.
(1979) reported that increased chronological age in cattle
resulted in a darker color of lean meat, but the post-mor-
tem stability of the muscle pigments was not clear.
While little is known about the possible age influence on
oxidative stability of beef from mature cows, or differences
between their muscle types, biochemical studies on small
animals or humans have demonstrated an increased
susceptibility of muscle cells to oxidizing agents as age pro-
gresses (Stadtman, 2006). It may be speculated that age-
related loss in the redox potential may occur in mature
cows, and this loss could predispose post-mortem muscles

0309-1740/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.012
106 Y.L. Xiong et al. / Meat Science 77 (2007) 105–113
from cows to a higher rate of oxidation. On the other hand,
a limited number of investigations have been conducted to
examine the possible effect of animal age on the post-
mortem progression of proteolysis in relation to meat
tenderness. Parrish, Selvig, Culler, and Zeece (1981),
Koohmaraie, Kennick, Elgasim, and Anglemier (1984)
and Huff-Lonergan, Parrish, and Robson (1995) observed
a reduced rate in post-mortem myofibrillar protein degra-
dation in muscle from physiologically mature (C50–E
maturity) cattle when compared to muscle from young (A
maturity) carcasses. The difference was implicated in the
variation of beef tenderness from cattle of different physio-
logical ages.
The lack of literature report on the potential differences
in the oxidative stability and protein degradation in post-
mortem muscles from mature cows of various ages, cou-
pled with the abundant supply of mature cows as a meat
source in the market, indicates a need to further explore
factors that may be responsible for the overall quality of
post-mortem meat. Thus, the objective of the present study
was to investigate the influence of cow age on the above
chemical and biochemical changes during post-mortem
storage of meat.
2. Materials and methods
2.1. Experimental design and meat sample preparation
A total of 18 cows composed predominantly of Angus ·
Simmental genetics and absent of Bos indicus influence,
purchased from local producers, were used in this study.
These cattle represented three age groups (six in 2–4 yr;
six in 6–8 yr; and six in 10–12 yr). Cows were placed on
the same fescue pasture for a minimum of 2 months prior
to slaughter to obtain a similar nutritional background.
In Experiment 1, nine cows with three from each of
the age groups were humanely slaughtered at the
USDA-inspected University of Kentucky abattoir using
standard industry procedures. Each animal was used in
an independent trial. After evisceration, carcasses were
electrically stimulated with 550 V, 2.2 A, using a Boss
electrical stunner (Model No. 1004D, Cincinnati Butcher
Supply, Cincinnati, OH) as described elsewhere (Schaake
et al., 1993). Carcasses were immediately split and chilled
in a 3 C cooler for 24 h. Carcass maturity, based on
both lean color (Longissimus muscle between the 12th
and 13th rib) and bone ossification, was determined using
an A00–E100 maturity scale to establish physiological
ages.
At 24 h post-mortem, the LD (13th rib through the 5th
lumbar vertebra) from both sides of carcasses were
removed, cut into 2.54-cm thick steaks, individually vac-
uum-packaged, and stored in a 3 C cooler. Warner-Brat-
zler shear (WBSF) force and sodium dodecyl sulfate–
polyacrylamide gel electrophoresis (SDS–PAGE) analyses
were performed with randomly selected steaks at 1, 3,
and 10 d post-mortem.
In Experiment 2, another nine cows with three from
each of the age groups were slaughtered and electrically
stimulated as described in Experiment 1. Each animal
was used in an independent trial. After chilling for 24 h,
Semitendinosus (ST) and Semimembranosus (SM) muscles
from both sides of carcasses were removed and ground
through a 4.5-cm orifice plate. Ground meat was formed
into patties of approximately 100 g in weight and 2 cm in
thickness. Patties were placed on Styrofoam plates and
over-wrapped with oxygen permeable plastic film (Ziploc
Hand-Wrap, Dow Chem. Co., Indianapolis, IN). The
plates were placed on a 3 C shelf under fluorescent lighting
that simulated retail display. Patties were analyzed for lipid
and myoglobin oxidation daily for a total of 7 d as
described below.
2.2. Detection of proteolysis
The post-mortem proteolytic changes in cow muscle
(LD) was monitored on myofibrils isolated from 1 d, 3 d,
and 10 d most-mortem steaks. Myofibrils were extracted
according to Goll, Young, and Stromer (1974) using a rigor
buffer consisting of 0.1 M NaCl, 0.05 M Na2HPO4, 5 mM
EDTA, and 1 mM NaN3 (pH 7.0). To identify protein deg-
radation, SDS–PAGE was performed using the procedure
of Laemmli (1970), which was modified to adapt to a smal-
ler electrophoretic apparatus, i.e., the SE 250 Mighty Small
II slab gel electrophoresis unit (Hoefer Scientific Instru-
ments, San Francisco, CA). A 3% acrylamide stacking gel
and a 10% acrylamide resolving gel were used. An equal
amount of protein (15 lg) was loaded in each lane.
Protein bands were stained with Coomassie brilliant
blue and identified by comparing the electrophoretic pat-
tern with published literature (Porzio & Pearson, 1977).
Molecular weights (MW) of unknown proteins were esti-
mated from the regression line generated by plotting log
MW of protein standards vs. their migration distance.
The protein MW standard was a cocktail of eight individ-
ual marker proteins (MW 14–200 kDa) (Sigma Chemical
Co., St. Louis, MO). Images of destained gels were cap-
tured with a digital camera, and the individual protein
bands were quantitatively analyzed using the UN-SCAN-
IT Gel digitizing software (Ver. 6.1, Silk Scientific Corp.,
Orem, UT). The number of pixels in each whole band were
recorded and used to calculate the percent weight of the
corresponding protein in the myofibril sample, assuming
that the amount of pixels was equivalent to the mass of
the protein.
2.3. Measurement of shear force
LD Steaks were cooked to an internal temperature of
70 C on a Farberware Open Hearth electric broiler (Far-
berware, Inc., Bronx, NY). After cooling to 22 C, six
1.27-cm-diameter cores from each cooked steak were
removed parallel to the muscle fibers. Cores were sheared
with a Warner-Bratzler shear device attached to an Instron
 Y.L. Xiong et al. / Meat Science 77 (2007) 105–113
4301 universal testing machine (Instron Corp., Canton,
MA) with a crosshead speed of 20 cm/min. The mean of
the WBSF values from different sample cores was deter-
mined and used for statistical analysis.
2.4. Measurement of lipid oxidation
Lipid oxidation in ST and SM patties after specific stor-
age times was measured before and after cooking using the
thiobarbituric acid (TBA)-reaction method, and was
expressed as TBA-reactive substances or TBARS (Sinnhu-
ber & Yu, 1977; Wang & Xiong, 2005). For cooked sam-
ples, patties were cooked to an internal temperature of
70 C on a Farberware Open Hearth electric broiler (Far-
berware, Inc., Bronx, NY) and then cooled to 22 C prior
to analysis. The TBARS values were calculated as mg of
malonaldehyde/kg of muscle sample.
2.5. Measurement of surface color
The surface color of raw ST and SM patties stored for
different days was measured with a Hunterlab colorimeter
(Model D25-2; Hunter Associates Laboratories, Inc., Fair-
fax, VA). The instrument, with a type DZA halogen lamp
light source and a 3.5-cm aperture, was calibrated using a
Hunterlab calibration plate no. C2-13717 (L* = 68.6,
a* = 23.5, and b* = 12.8) every time before use. Values
obtained from triplicate patties were averaged to obtain
the mean for each color parameter.
2.6. Determination of pigments
Total myoglobin and metmyoglobin (MMb) in ST and
SM raw patties were determined according to Krzywiki
(1982) using a 40 mM phosphate extraction buffer (pH
6.8). The absorbance of the pigment solution was read at
572, 565, 545, and 525 nm in a UV–visible spectrophotom-
eter (Model UV-160, Shimadzu Co. Ltd., Columbia, MD).
Total myoglobin and percent MMb were calculated based
on the following equations (Krzywiki, 1982):
Total myoglobinðmmol=LÞ
¼ ð0:166R1 þ 0:086R2 þ 0:088R3 þ 0:099Þ  A525
MMbð%Þ ¼ ð2:514R1 þ 0:777R2 þ 0:800R3 þ 1:098Þ
 100%
where R1, R2, and R3 = absorbance ratios of A572/A525,
A565/A525, and A545/A525, respectively. For the total myo-
globin content, the values were converted to mg/kg meat
based on the dilution and the molecular weight
(16,950 Da) of myoglobin (Dobberstein & Schroeder,
1993).
2.7. Statistical analysis
Each animal within the same age group was treated as a
replicate. Data from the three independent (replicated) tri-
107
als were analyzed using the General Linear Procedure of
the Statistix 7.0 software package (Analytical Software,
Inc., Tallahassee, FL) for microcomputers. Two-way anal-
ysis of variance (ANOVA) was performed to determine the
significance of the effect of animal age and muscle post-
mortem storage time. The ANOVA tables obtained were
further analyzed for the comparison of means by Least Sig-
nificant Difference procedures.
3. Results and discussion
3.1. Tenderness and proteolysis
The LD steaks from 10 yr to 12 yr cows were less tender,
i.e., having a higher mean WBSF (P < 0.05), than steaks
from the other two age groups (Fig. 1). The 6–8 yr group
steaks also tended to be less tender than the 2–4 yr group
steaks. The result agreed with previous findings that LD
muscle from E maturity carcasses (old) was tougher com-
pared to either C or D maturity carcasses (younger) (Hilton
et al., 1998; Smith et al., 1982). When the mean WBSF val-
ues of steaks across the storage time were compared, it was
clear that tenderness progressively improved (P < 0.01)
during post-mortem storage (Fig. 1), a physicochemical
process that occurs in meat during post-mortem storage
at refrigerator temperatures regardless of the animal age
(Geesink & Koohmaraie, 1999; Goll et al., 1983; Kong,
Diao, & Xiong, 2006; Rowe, Maddock, Lonergan, &
Huff-Lonergan, 2004).
Because proteases are widely implicated in the tenderiza-
tion of meat during post-mortem storage, SDS–PAGE was
performed to identify different proteolytic patterns that
might explain the age-related variations in meat tenderness
as well as the post-mortem storage effect. Several discern-
able proteolytic events occurred in all LD muscle samples
60
A
55
a
50
Shear Force (N)
b
B
b
45 C
40
35
30
2-4 6-8 10-12 0 1 3 10
Animal Age (yr) Post—mortem Time (d)
Fig. 1. Effects of animal age and post-mortem storage (3 C) time on the
WBSF of Longissimus muscle from cows of different ages. The standard
errors of treatment means are indicated. Means in the same treatment
group (age or storage time) with different letters differ significantly
(P < 0.05).
108 Y.L. Xiong et al. / Meat Science 77 (2007) 105–113

during storage, with the progressive degradation of tropo- X Y

nin-T and the concomitant appearance of a 33-kDa new
polypeptide (‘‘band Y’’) being the most noticeable changes
(Fig. 2). The progression of proteolysis into 10 d post-mor-
tem also revealed a new protein band (designated as ‘‘band
X’’) between troponin-T and tropomyosin, presumably
originating from troponin-T. All these proteolytic changes
were consistent with previous literature reports (Huff-
Lonergan et al., 1995; Koohmaraie et al., 1984; Xiong,
Moody, Blanchard, Liu, & Burris, 1996). The results sug-
gest that the improved tenderness of LD muscle during
post-mortem storage (Fig. 1) was a result of proteolytic 2–4 yr
fragmentation of myofibrils.
Despite the similar proteolytic patterns exhibited by all
the age groups, the proteolysis appeared to have proceeded
at a faster rate and to a slightly greater extent in the 2–4 yr 6–8 yr
age group samples, which was revealed by SDS–PAGE
(Fig. 2) and indicated by digital image analysis (Fig. 3).
For example, by day 3, the troponin-T band was less
intense but the 33 kDa band was more noticeable for the 10–12 yr
2–4 yr age group steaks compared to samples from 6 to 8
Fig. 3. Densitometric scan of an SDS–PAGE gel portion (22–45 kDa) of
and 10 to 12 yr age groups. By day 10, both the band
myofibrils isolated from 10-d post-mortem Longissimus muscle.
‘‘X’’ polypeptide and the 33 kDa component (band ‘‘Y’’)
became rather salient in the 2–4 yr age group samples
(Fig. 3). The total digital pixels for protein ‘‘X’’ band in (1995) postulated that LD samples from old cattle proba-
the 2–4 yr, 6–8 yr, and 10–12 yr age samples (10 d) were, bly contained higher calpastatin activity that inhibited cal-
respectively, 10,450, 7733, and 5748, which represented a pain-mediated protein degradation in post-mortem beef.
100:74:55 ratio. The more pronounced proteolysis in the Recent studies on aging have demonstrated that protein
2–4 yr age group muscle would result in more rapid and oxidation was an integral part of the physiological aging
extensive disruption of the myofibrils, thus, explaining process (Rizvi, Jha, & Maurya, 2006; Stadtman, 2006),
the age effect on LD muscle tenderness observed in the and in muscle cells, calpain was one of the most susceptible
present study. enzymes affected by oxidation (Guttmann, Elce, Bell,
Several muscle endogenous enzyme systems have been Isbell, & Johnson, 1997). It has been shown that when
implicated in post-mortem tenderization of meat, including exposed to an oxidizing environment, m-calpain was read-
cathepsins, calpains, and proteosomes (Dutaud et al., 2006; ily inactivated, resulting in a reduced tenderization of post-
Goll et al., 1983; Rowe et al., 2004). A large body of evi- mortem beef muscle (Rowe et al., 2004). When all these
dence links improved beef tenderization during post-mor- factors are put together, it can be hypothesized that muscle
tem storage to calpain-mediated degradation of key from older cows probably had reduced calpain activity due
myofibrillar proteins responsible for the integrity of the to greater oxidative stress and therefore, exhibited
overall muscle cell cytoskeleton (Christensen, Young, Law- increased toughness, when compared with younger cattle.
son, Larsen, & Purslow, 2004; Geesink & Koohmaraie, On the other hand, a comparative study showed that
1999; Rhee, Ryu, & Kim, 2006). Huff-Lonergan et al. meat toughness was better correlated with, thus, could be

Day 1 Day 3 Day 10

45 kDa Actin
Troponin T
X Tropomyosin
Y

22 kDa Troponin I

2-4 6-8 10-12 2-4 6-8 10-12 2-4 6-8 10-12

Age of Animals (yr)

Fig. 2. SDS–PAGE of myofibrils isolated from post-mortem Longissimus muscle of cows from different age groups. Muscle samples were stored at 3 C
for various days. Shown in the graph is the portion of the gel (22–45 kDa) that exhibited most proteolytic changes. An equal amount of protein (15 lg) was
loaded to each lane.
 Y.L. Xiong et al. / Meat Science 77 (2007) 105–113 109

more reliably predicted by, muscle serine peptidase inhib-

Average
itors than by calpastatin or cysteine peptidase inhibitors

37.74x
18.29y
11.08x
32.92y
20.56x
10.08y
(Zamora et al., 2005). Moreover, using fragmentation
structural analysis, Dutaud et al. (2006) found that 20S
proteasome could reproduce the type of sequential, ultra-

33.16ab (1.68)
37.89a (1.96)
14.96e (1.19)
10.61c (0.47)

9.44c (0.50)
17.90f (0.94)
structural changes normally occurring in post-mortem
myofibrils, suggesting that this protease superfamily
may also be intimately involved in the aging process of

7
meat. Further research is needed to identify the exact
enzyme systems involved in the tenderization of mature

33.20ab (1.71)
37.77a (1.96)
15.20e (1.09)
10.37c (0.50)

9.32c (0.43)
17.83f (0.86)
cow meat seen in the present study.

Means in the same row without a common superscript letter differ significantly (P > 0.05). Values in parentheses are standard errors of the respective means.
Hunter calorimetric values of raw ground beef patties from Semitendinosus and Semimembranosus muscles of cull cows during simulated retail display at 3 C
3.2. Oxidative stability

6
3.2.1. Color and pigments

18.86ef (1.11)
32.62b (1.82)
38.06a (2.18)
15.57e (1.04)
10.38c (0.33)

9.52c (0.51)
There were no significant differences between the 2–4,
6–8, and 10–12 yr age groups for raw ST or raw SM pat-
ties before storage or during storage for all the surface

Indicate significant difference (P > 0.05) between the two muscle types for the same color parameters shown in the ‘Average’ column.
color parameters (L*, a*, and b*), although numerically,

5
the L*-values of raw patties of the 10–12 yr age group

19.69de (1.21)
17.22d (1.15)

32.64b (1.89)
37.62a (2.19)
tended to be slightly lower (but nonsignificant) and

10.37c (0.50)

9.47c (0.42)
appeared darker when compared with the younger age
groups. Hence, the respective colorimetric values were
pooled to obtain the means. Overall, there was no consis-

4
tent (P > 0.05) change in the L*-value during storage, but
a*- and b*-values decreased (P > 0.05) after 7 d for both

20.38cd (1.25)
18.11d (0.88)

32.27b (2.22)
37.70a (1.84)

10.60c (0.33)

9.72c (0.39)
ST and SM samples (Table 1). The most noticeable
decrease in the a*-value was seen between day 0 and
day 1, indicating a significant loss of red color in the early 3
stage of storage.
Comparison of ST and SM patties showed that the for-
11.19bc (0.40)

10.13bc (0.32)
32.24b (2.11)
37.15a (1.74)
19.32c (0.98)

21.28c (1.22)
mer had consistently greater lightness (L*) but lower red-
ness (a*) (P > 0.05) than the latter. It can be suggested
that the color difference was a result of unequal distribu-
2

tions of specific fiber types. Kirchofer, Calkins, and

Gwartney (2002) reported 24.3%, 26.0%, and 49.7%,
21.24b (1.11)
11.65b (0.29)
32.49b (1.85)
22.51b (1.29)
10.64b (0.41)
36.87a (1.55)

respectively, of b-red, a-red, and a-white fibers in beef

ST, compared to 26.3%, 28.6%, and 45.1% in beef SM.
However, total myoglobin content (3.15 ± 0.55 mg/g
1

muscle) was similar (P > 0.05) between ST and SM sam-

ples, and it was not influenced (P > 0.05) by animal age
Storage time

38.87a (1.45)
24.66a (1.42)
13.48a (0.33)
34.76a (1.35)
26.00a (1.80)
12.40a (0.60)

(results not shown). Hence, the slightly less red (a*-value)

in ST muscle must be due to a lower oxygenation of myo-
globin when compared with SM muscle.
0

Most of the pigments in fresh, raw patties were deoxy-

myoglobin (interior) and oxymyoglobin (surface) that
Color parameter

imparted the bright red color with a relatively high

a*-value. During the simulated retail display under fluo-
L*-value

L*-value
a*-value
b*-value

a*-value
b*-value

rescent light, a large proportion of the pigments was

converted to metmyoglobin, with the most rapid conver-
sion occurring during the first day (Fig. 4). By day 7, met-
myoglobin accounted for almost 70% of the total
Semimembranosus

pigment. Animal age was not a factor in the metmyoglo-

Semitendinosus

bin production except for the ST patties that contained a

slightly greater (P > 0.05) amount of metmyoglobin in the
Table 1

Sample

10–12 yr age group on day 5 than in the 2–4 yr and 4–6 yr

a–f

groups. The results followed the trend of decreasing a*-

x,y
110 Y.L. Xiong et al. / Meat Science 77 (2007) 105–113

100 ST 100 SM
3-4 yr 3-4 yr
6-8 yr 6-8 yr
80 10-12 yr 80 10-12 yr

% Metmyoglobin

% Metmyoglobin
60 60

40 40

20 20

0 0

0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8

Time (day) Time (day)

Fig. 4. Metmyoglobin content in raw beef patties stored at 3 C for various days. Patties were prepared from Semitendinosus (ST) and Semimembranosus
(SM) muscles of cows from different age groups.

values during storage (Table 1). Demos and Mandigo 3.2.2. Lipid oxidation
(1996) studied the discoloration of ground beef stored at The analysis of TBARS in patties indicated pronounced
refrigerator temperature in an oxygen-permeable packag- differences in the rate and extent of lipid oxidation between
ing system and showed that by day 7, approximately 65% the three age groups of cows as well as between ST and SM
of myoglobin was changed to metmyoglobin. Similar muscles during refrigerated retail display. Raw patties of
results were reported by van den Oord and Wesdorp both ST and SM from 10 to 12 yr cows were most suscep-
(1971), who observed that percent oxymyoglobin dropped tible to lipid oxidation, followed by the 6–8 yr age group,
to about 50% by day 7 during storage (5 C), and the and the 2–4 yr group was least susceptible (Fig. 5). On
reduction was positively correlated with the undesirable average (ST and SM from all ages combined), the TBARS
color ratings by a sensory panel. values of raw samples increased from 3.87 mg/kg on day 0
It was noteworthy that although both the a*-value to 11.42 mg/kg after 7 d (P > 0.05).
reduction and the metmyoglobin formation occurred most Proximate analysis in a preliminary experiment showed
rapidly during the first 24 h, the latter was more dramatic. no significant difference in total lipid between all samples,
This can be explained because samples for pigment extrac- i.e., 5.59%, 5.44%, and 5.03% for ST, and 5.15%, 5.57%,
tion were whole meat patties (blended) while only the sur- and 5.37% for SM from 2 to 4 yr, 6 to 8 yr, and 10 to
face color of the patties was measured by the colorimeter. 12 yr cows, respectively. Hence, the animal age-associated
The color of the patties, and thus the pigment composition, variation in lipid oxidation as well as the difference between
were not consistent throughout the whole patties, i.e., red muscle types cannot be attributed to the lipid content. It
on the surface (up to 3–4 mm deep) and purplish/brown has been shown that in humans, the total plasma antioxi-
in the interior after 1 d (preliminary observation). The dants decreased with age (Rizvi et al., 2006). While there
reduced partial pressure of oxygen inside the meat patties is no literature report on age-related changes in redox
facilitated myoglobin oxidation. potential of post-mortem meat, it may be suggested that

6 ST 6 SM
2-4 yr 2-4 yr
6-8 yr 6-8 yr
5 5
TBARS (mg/kg meat)

TBARS (mg/kg meat)

10-12 yr 10-12 yr

4 4

3 3

2 2

1 1

0 0
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Time (day) Time (day)

Fig. 5. TBARS content in raw beef patties stored at 3 C for various days. Patties were prepared from Semitendinosus (ST) and Semimembranosus (SM)
muscles of cows from different age groups.
 Y.L. Xiong et al. / Meat Science 77 (2007) 105–113 111

the endogenous antioxidants could be lower in the 10–12 yr of myoglobin. These myoglobin molecules were readily oxi-
cow muscle than in muscle from younger groups, thus, dized and were responsible for the rapid production of met-
enhancing the oxidative susceptibility of lipids. The rate myoglobin observed in the first 24 h of retail display.
and extent of TBARS generation were greater (P > 0.05) Subsequent oxidation was slow because the remaining
for ST than for SM. It is plausible that the ST muscle myoglobin molecules, which would still be confined in the
was deficient in endogenous antioxidants when compared cell, were relatively inaccessible to molecular oxygen for
with the SM muscle. Cooking slightly increased the all muscle samples. On the other hand, lipid oxidation
TBARS production. The amount of TBARS in these occurred mostly outside the cell or at the membrane inter-
stored patties after cooking increased by an average of face, in which case the concentration and type of oxidants
about 10% when compared with uncooked patties or antioxidants, which may vary among animal age groups,
(Fig. 6). Membrane damage, the release of prooxidative became crucial factors.
heme, and the heat input would all promote lipid oxidation Indeed, within the same muscle type or animal age
(Grunwald & Richards, 2006; Monahan, Crackel, Gray, group, lipid oxidation (TBARS) was significantly corre-
Bukley, & Morrissey, 1993). lated with metmyoglobin production (Fig. 7), suggesting
The significant animal age effects on muscle lipid oxida- that the same oxidation initiators may be involved. Mona-
tion but not on metmyoglobin formation, and the notable han, Asghar, Gray, Buckley, and Morrissey (1994) noted
difference between muscle types in their lipid stability but that oxidation of myoglobin preceded oxidation of muscle
not in their color stability, suggest that lipid and pigment lipids in pork chops stored at 4 C. However, the opposite
oxidation in these muscle samples were probably controlled was observed by the same research lab for beef muscle
by different mechanisms, one of which was physical barri- homogenates that were treated with oxidizing agents
ers. With coarse grinding, some of the cells were disrupted, (FeCl3/ascorbate) (O’Grady, Monahan, & Brunton,
irrespective of origin of the muscles, resulting in the release 2001). The conflicting results reported by these researchers

6 ST 6
SM
2-4 yr 2-4 yr
5 6-8 yr 5 6-8 yr
TBARS (mg/kg meat)

TBARS (mg/kg meat)

10-12 yr 10-12 yr

4 4

3 3

2 2

1 1

0 0
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Time (day) Time (day)

Fig. 6. TBARS content in cooked beef patties prepared from Semitendinosus (ST) and Semimembranosus (SM) muscles of cows from different age groups.
The raw patties were stored at 3 C for various days before cooking.

ST SM
6 6
2-4 yr (r = 0.92) 2-4 yr (r = 0.96)
6-8 yr (r = 0.97) 6-8 yr (r = 0.83)
5 5
TBARS (mg/kg meat)

TBARS (mg/kg meat)

10-12 yr (r = 0.98) 10-12 yr (r = 0.94)

4 4

3 3

2 2

1 1

0 0
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
% Metmyoglobin % Metmyoglobin

Fig. 7. Linear regression plots of TBARS versus % metmyoglobin for raw patties prepared from Semitendinosus (ST) and Semimembranosus (SM) muscles
of cows from different age groups. The correlation coefficients (r) are indicated.
112 Y.L. Xiong et al. / Meat Science 77 (2007) 105–113
with our present findings (in which beef sample oxidation
occurred under fluorescent light) indicated the specificity
of lipid and myoglobin oxidation to the type of oxidation
initiators.
4. Conclusions
The findings validated the hypothesis that oxidative
stability of lipids and tenderness in post-mortem beef
muscle decreased with the age of mature cows, and
substantiated the general notion that the overall quality
of beef is reduced with the age of cattle. The results also
demonstrated that muscle type was a significant factor
affecting the overall oxidative stability of beef from
mature cows. Thus, meat processors should take into con-
sideration the specific ages of cows when utilizing meat as
fresh steaks or for further processing. For this particular
group of beef cattle, while toughness is expected to
become a lesser issue in ground cow meat, antioxidative
strategies are needed in order to improve its oxidative sta-
bility and flavor.
References
Boccard, R. L., Naude, R. T., Cronje, D. E., Smit, M. C., Venter, H. J., &
Rossouw, E. J. (1979). The influence of age, sex and breed of cattle on
their muscle characteristics. Meat Science, 3, 261–280.
Christensen, M., Young, R. D., Lawson, M. A., Larsen, L. M., &
Purslow, P. P. (2004). Effect of added l-calpain and post-mortem
storage on the mechanical properties of bovine single muscle fibers
extended to fracture. Meat Science, 66, 105–112.
Demos, B. P., & Mandigo, R. W. (1996). Color of fresh, frozen and
cooked ground beef patties manufactured from mechanically recov-
ered neck bone lean. Meat Science, 42, 415–429.
Dobberstein, P., & Schroeder, E. (1993). Accurate mass determination of a
high-molecular-weight protein using electrospray-ionization with a
magnetic-sector instrument. Rapid Communication in Mass Spectrom-
etry, 7, 861–864.
Dutaud, D., Aubry, L., Guignot, F., Vignon, X., Monin, G., & Ouali, A.
(2006). Bovine muscle 20S proteasome. II: Contribution of the 20S
proteasome to meat tenderization as revealed by an ultrastructural
approach. Meat Science, 74, 337–344.
Geesink, G. H., & Koohmaraie, M. (1999). Effect of calpastatin on
degradation of myofibrillar proteins by micro-calpain under post-
mortem conditions. Journal of Animal Science, 77, 2685–2692.
Goll, D. E., Otsuka, Y., Nagainis, P. A., Shannon, J. D., Sathe, S. K., &
Muguruma, M. (1983). Role of muscle proteases in maintenance of
muscle integrity and mass. Journal of Food Biochemistry, 7, 137–177.
Goll, D. E., Young, R. B., & Stromer, M. H. (1974). Separation of
subcellular organelles by differential and density gradient configura-
tion. In Proceedings of reciprocal meat conference (Vol. 27, pp. 250–
297).
Grunwald, E. W., & Richards, M. P. (2006). Studies with myoglobin
variants indicate that released hemin is the primary promoter of lipid
oxidation in washed fish muscle. Journal of Agricultural and Food
Chemistry, 54, 4452–4460.
Guttmann, R. P., Elce, J. S., Bell, P. D., Isbell, J. C., & Johnson, G. V.
(1997). Oxidation inhibits substrate proteolysis by calpain I but not
autolysis. Journal of Biological Chemistry, 272, 2005–2012.
Hilton, G. G., Tatum, J. D., Williams, S. E., Belk, K. E., Williams, F. L.,
Wise, J. W., et al. (1998). An evaluation of current and alternative
systems for quality grading carcasses of mature slaughter cows. Journal
of Animal Science, 76, 2094–2103.
Huff-Lonergan, E., Parrish, F. C., Jr., & Robson, R. M. (1995). Effects of
postmortem aging time, animal age, and sex on degradation of titin
and nebulin in bovine longissimus muscle. Journal of Animal Science,
73, 1064–1073.
Kirchofer, K. S., Calkins, C. R., & Gwartney, B. L. (2002). Fiber-type
composition of muscles of the beef chuck and round. Journal of Animal
Science, 80, 2872–2878.
Kong, B. H., Diao, X. P., & Xiong, Y. L. (2006). Postmortem calcium
chloride injection alters ultrastructure and improves tenderness of
mature Chinese yellow cattle Longissimus muscle. Journal of Food
Science, 71, C124–C129.
Koohmaraie, M., Kennick, W. H., Elgasim, E. A., & Anglemier, A. F.
(1984). Effects of postmortem storage on muscle protein degradation:
Analysis by SDS–polyacrylamide gel electrophoresis. Journal of Food
Science, 49, 292–293.
Krzywiki, K. (1982). The determination of heme pigments in meat. Meat
Science, 7, 29–36.
Laemmli, U. K. (1970). Cleavage of structural proteins during the
assembly of the head of bacteriophage T. Nature, 227, 680–685.
Lawrie, R. A. (1998). Meat science (6th ed.). Cambridge, UK: Woodhead
Publishing, Ltd., p. 336.
Monahan, F. J., Asghar, A., Gray, J. I., Buckley, J. D., & Morrissey, P. A.
(1994). Effect of oxidized dietary lipid and vitamin E on the colour
stability of pork chops. Meat Science, 37, 205–215.
Monahan, F. J., Crackel, R. L., Gray, J. I., Bukley, D. L., & Morrissey, P.
A. (1993). Catalysis of lipid oxidation in muscle model systems by
haem and inorganic iron. Meat Science, 34, 95–106.
O’Grady, M. N., Monahan, F. J., & Brunton, N. P. (2001). Oxymyoglobin
oxidation and lipid oxidation in bovine muscle – Mechanistic studies.
Journal of Food Science, 66, 386–392.
Parrish, F. C., Jr., Selvig, C. J., Culler, R. D., & Zeece, M. G. (1981). CAF
activity, calcium concentration, and the 30,000-Da component of
tough and tender bovine longissimus muscle. Journal of Food Science,
46, 308–309, 311.
Porzio, M. A., & Pearson, A. M. (1977). Improved resolution of
myofibrillar proteins with sodium dodecyl sulfate–polyacrylamide gel
electrophoresis. Biochimistry and Biophysics Acta, 490, 27–34.
Rhee, M. S., Ryu, Y. C., & Kim, B. C. (2006). Post-mortem metabolic rate
and calpain system activities on beef longissimus tenderness classifi-
cations. Bioscience, Biotechnology and Biochemistry, 70, 1166–1172.
Rizvi, S. I., Jha, R., & Maurya, P. K. (2006). Erythrocyte plasma
membrana redox system in human aging. Rejuvenation Research, 9,
470–474.
Rowe, L. J., Maddock, K. R., Lonergan, S. M., & Huff-Lonergan, E.
(2004). Oxidative environments decrease tenderization of beef steaks
through inactivation of m-calpain. Journal of Animal Science, 82,
3254–3266.
Schaake, S. L., Skelley, G. C., Halpin, E., Grimes, L. W., Brown, R. B.,
Cross, D. L., et al. (1993). Carcass and meat sensory traits for steers
finished on fescue and clover, summer forage, or for different periods
in drylot. Journal of Animal Science, 71, 3199–3205.
Shorthose, W. R., & Harris, P. V. (1990). Effect of animal age on the
tenderness of selected beef muscles. Journal of Food Science, 55,
1–8.
Sinnhuber, R. O., & Yu, T. C. (1977). The 2-thiobarbituric acid
reaction, an objective measure of the oxidative deterioration
occurring in fats and oils. Journal of Japanese Oil Chemist Society,
26, 259–267.
Smith, G. C., Cross, H. R., Carpenter, Z. L., Murphey, C. E., Savell, J.
W., Abraham, H. C., et al. (1982). Relationship of USDA maturity
group to palatability of cooked beef. Journal of Food Science, 47,
1100–1107.
Stadtman, E. R. (2006). Protein oxidation and aging. Free Radical
Research, 40, 1250–1258.
USDA (2007). USDA-NASS agricultural statistics. <http://www.nass.us-
da.gov/index.asp>.
van den Oord, A. H. A., & Wesdorp, J. J. (1971). Color rating and
pigment composition of beef. Journal of Food Technology, 6, 15–20.
 Y.L. Xiong et al. / Meat Science 77 (2007) 105–113 113

Wang, L. L., & Xiong, Y. L. (2005). Inhibition of lipid oxidation in
cooked beef patties by hydrolyzed potato protein is related to its
reducing and radical scavenging ability. Journal of Agricultural and
Food Chemistry, 53, 9186–9192.
Xiong, Y. L., Moody, W. G., Blanchard, S. P., Liu, G., & Burris, W. R.
(1996). Postmortem proteolytic and organoleptic changes in hot-boned
muscle from grass- and grain-fed and zeranol-implanted cattle. Food
Research International, 29, 27–34.
Zamora, F., Aubry, L., Sayd, T., Lepetit, J., Lebert, A., Sentandreu, M.
A., et al. (2005). Serine peptidase inhibitors, the best predictor of beef
ageing amongst a large set of quantitative variables. Meat Science, 71,
730–742.
 MEAT
SCIENCE
Meat Science 77 (2007) 114–120
www.elsevier.com/locate/meatsci

Biochemical changes during processing of traditional Jinhua ham

a,* b
G.H. Zhou , G.M. Zhao
a
Key Laboratory of Meat Processing and Quality Control EDU, College of Food Science and Technology, Nanjing Agricultural University,
Nanjing 210095, PR China
b
Food Science and Technology, Henan Agricultural University, Zhengzhou 450002, PR China

Received 7 March 2007; received in revised form 28 March 2007; accepted 29 March 2007

Abstract

Jinhua ham is the most famous traditional meat product of China and one of the most famed dry-cured hams in the world. Its pro-
cessing consists of six stages: green ham preparation, salting, washing and sun-drying and shaping, ripening, and post-ripening. Intense
proteolysis and lipolysis occur during processing period. As a result, the content of free amino acids in final ham products is 14–16 times
that of green ham, and 191 volatile compounds have been identified during processing, which make a major contribution to the flavor of
Jinhua ham.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Jinhua ham; Processing; Proteolysis; Lipolysis; Flavor development

1. Introduction
China has a long, glorious history and a splendid dietary
culture. Many traditional meat products have been devel-
oped in China, among which Jinhua ham is the most
famous. Jinhua ham has an attractive color, unique flavor
and bamboo leaf-like shape. Its rose-like muscle, golden
yellow skin and pure white fat make Jinhua ham one of
the most preferred items in Chinese cuisine. The processing
technology of Jinhua ham was introduced to European
countries by Marco Polo during the 13th to 14th century
and had an important impact on the development of dry-
cured ham processing technology outside China.
Jinhua ham is formed and produced in Jinhua District,
Zhejiang Province in China. The earliest legend regarding
the method of processing Jinhua ham may be traced back
to the Tang Dynasty (A.D. 618–907); however, the name
‘‘Jinhua ham’’ was formally bestowed by the first emperor
*
Corresponding author. Tel.: +86 025 84395376; fax: +86 025
84432420.
E-mail addresses: ghzhou@njau.edu.cn (G.H. Zhou), gmzhao126@
yahoo.com.cn (G.M. Zhao).
of the South Song Dynasty about 800 years ago (Wu, Sun,
& Sun, 1959). Typical Jinhua ham processing generally
takes 8–10 months, starting in winter and finishing in
autumn of the following year. During a long ripening pro-
cess, muscle protein and fat are hydrolyzed to some extent
by internal enzymes, and many small peptides, free amino
acids, free fatty acids and volatiles are produced, which
eventually contribute to the unique flavor of Jinhua ham.
This paper outlines the processing technology of tradi-
tional Jinhua ham, and describes the flavor- and taste-
related compounds that are produced and the enzymes that
are involved during the course of processing.
2. Processing of traditional Jinhua ham
Jinhua ham is traditionally processed under natural con-
ditions where air temperature and relative humidity depend
on climate and weather condition. Its unique quality is not
only due to the elaborate processing technology, but also
related to the unique local geographic terrain and climate.
About 70% of the Jinhua District is mountainous with four
distinct seasons. Air temperature fluctuates regularly with-
out extremes, which is desirable for processing dry-cured

0309-1740/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.03.028
 G.H. Zhou, G.M. Zhao / Meat Science 77 (2007) 114–120
meat products. Usually, air temperature in winter is
between 0 and 10 °C, favorable for ham salting. In spring,
air temperature goes up to 20 °C or higher, suitable for
sun-drying. Summer in the Jinhua region is relatively hot.
The temperature can reach 40 °C, ideal for ham ripening.
Post-ripening begins in autumn when the temperature
falls. The entire processing of Jinhua ham, beginning in
winter and finishing in the following autumn, takes 8–10
months.
Processing of Jinhua ham consists of six stages: green
ham preparation, salting, washing, sun-drying and shap-
ing, ripening, and post-ripening. In all, there are more than
90 steps within the six stages.
2.1. Green ham preparation
Traditionally, only hind legs from the Jinhua ‘Lian-
gtouwu’ pig or its cross offspring could be used for produc-
ing Jinhua ham. Desirable legs should be fresh, with a thin
skin and slim shank bone, well-developed muscle, and a
thin layer of white fat. Broken bone should be particularly
avoided. The exposed part of the bones as well as the fat,
tendon and muscle membrane on the meat surface of
selected legs are removed and the hams then trimmed into
a shape like a bamboo leaf. The remaining blood should be
squeezed out. A leg weight of between 5.5 and 7.5 kg after
trimming is preferred (Zhao & Zhou, 2003).
2.2. Salting
Salting is a critical stage in Jinhua ham processing and
inappropriate salting may cause spoilage. Ambient temper-
ature and humidity have great effects on the salting process.
With regard to temperature, it’s difficult for salt to pene-
trate meat when it is below 0 °C, while fast growth of
microbes will occur when temperature is above 15 °C. With
regard to humidity, when ambient humidity is below 70%
RH, undesirable water loss will occur, which also causes
insufficient salt penetration. When ambient humidity is
above 90%, salt will flow away in the form of brine, causing
pastiness on the surface of the ham. Therefore, the desir-
able ambient temperature and humidity for salting is 5–
10 °C and 75–85% RH, respectively (Gong, 1987; Zhao &
Zhou, 2003; Zhao et al., 2004). Average duration of salting
is about 30 d, varying from 25 d for small hams (<5 kg) to
35 d for large hams (more than 8 kg), during which time
each ham is salted 5–7 times. Usually only dry salt is used
during salting, however, nitrate may also be used during
the first two times of salting when irregular weather condi-
tion is encountered.
2.3. Soaking and washing
The purpose of this procedure is to remove excess salt
and wash off any dirty substances on the surface of the
meat. Usually, hams are initially soaked in water for 4–
6 h and then washed with bamboo brushes. Water is chan-
115
ged after initial washing, and hams are again soaked in
water for another 16–18 h.
2.4. Sun-drying and shaping
Achieving appropriate dehydration is the objective of
sun-drying as insufficient dehydration may cause spoilage.
For balance, a pair of hams of similar weight are tied with
a rope and hung on a rack. Sufficient ventilation and expo-
sure to sunshine are considered when positioning the hams.
When hams are hung, hoofs are removed, water and dirt
on skin is razed off with blade and the brand is sealed on
the skin. Then hams can be removed from the racks and
shaped into a bamboo leaf-like shape. Sun-drying can be
terminated when hams start to drop liquidified fat, which
generally requires about 7 sunny days.
2.5. Ripening
Ripening is the key process for generating ham flavor
substances. During this period, muscle proteins and fat
are hydrolyzed mainly by endogenous enzymes, which
results increased amounts of peptides, free amino acids
and free fatty acids. These products constitute the main
part of ham flavor substances and may continue to react
with one other or be hydrolyzed to produce volatiles that
contribute to the unique aroma of Jinhua ham.
Ham pairs are fastened to a centipede rack (named
because of its centipede-like shape) with the meat surface
toward the windows in the ripening room. Ham quality is
susceptible to the microclimate in the ripening room. High
temperature with low humidity stimulates weight loss and
fat oxidation, while high humidity may result in ham spoil-
age. On the other hand, low temperature, especially in the
later phase of ripening, slows aroma formation in hams.
The ripening room should be well ventilated. Room tem-
perature should increase from 15 °C to 37 °C gradually
and humidity is controlled to within 55–75% during the
6–8 months ripening.
During ripening, skin and muscle shrink to some extent
because of moisture loss and bones around joints protrude
out of the surface of the meat. Therefore, hams are usually
removed from the centipede rack and retrimmed in mid-
April. This is normally the last shaping and it requires cut-
ting off protruding bones, superfluous skin and fat. After
reshaping, hams are retied to the centipede rack for com-
pletion of ripening. Usually ripening process terminates
in mid-August when temperatures begin to drop.
2.6. Post-ripening
After ripening, the surface of the meat becomes very dry
and is often covered with a thin layer of mould spores and
dust. For this reason, hams are first brushed clean and a
thin layer of vegetable oil applied to soften muscle and pre-
vent excessive fat oxidation, then stacked with skin side up
for post-ripening. The post-ripening stage is a process
116 G.H. Zhou, G.M. Zhao / Meat Science 77 (2007) 114–120

designed to stabilize and intensify ham flavor; it is carried 30 Total soluble

out in a warehouse and usually takes two months. During nitrogen %
accouting for
post-ripening, the ham piles are turned over from time to 25
total nitrogen%
prevent unexpected fermentation, which affects ham
quality. 20 Non protein
nitrogen
% accouting for
2.7. Grading and storage 15 total nitrogen%

Soluble protein
It’s common practice to grade Jinhua ham into different 10 nitrogen
accouting for
categories according to quality, depending primarily on total nitrogen%
aroma intensity. A grader appointed by the government 5
1 2 3 4 5 6
assesses ham’s aroma by inserting a bamboo probe into a Stage
ham and smelling the probe when removing it. There are Fig. 1. Changes of NPN, soluble protein nitrogen and total soluble
three fixed locations: ‘up’, ‘middle’ and ‘lower’ position nitrogen in total nitrogen during Jinhua ham processing. Note: Stage 1,
on a ham for this special purpose. Jinhua ham can be before salting; 2, end of salting; 3, end of sun-drying; 4, middle of ripening;
stored for years and peak flavor is reached when stored 5, end of ripening; and 6, end of post-ripening.
for around 12 months.
80 Non protein
3. Proteolysis and lipolysis nitrogen
70 accouting for
total soluble
Proteins and lipids constitute the major chemical com- nitrogen %
ponents of ham muscles and their proportions in muscle 60
increase gradually with the loss of muscle moisture during Soluble protein
% 50 nitrogen
processing. Intense proteolysis and lipolysis occur over the accouting for
course of Jinhua ham processing. As a result, the content of total soluble
40 nitrogen %
free amino acids in final Jinhua ham products is 14–16
times that of green ham, and 191 volatile compounds have
30
been identified, which make major contribution to the
aroma and flavor of Jinhua ham. 20
1 2 3 4 5 6
Stage
3.1. Effects of muscle enzymes in the processing of Jinhua
ham Fig. 2. Changes of NPN and soluble protein nitrogen in total nitrogen
during Jinhua ham processing. Note: Stage 1, before salting; 2, end of
3.1.1. Proteolysis salting; 3, end of sun-drying; 4, middle of ripening; 5, end of ripening; and
6, end of post-ripening.
Muscle proteins are hydrolyzed continuously during
processing of Jinhua ham and small molecular products
such as peptides and free amino acids (FAA) are generated dases and aminopeptidases have been the focus of further
(Zhao et al., 2005b). At the end of processing, the proteo- studies.
lytic index (non-protein nitrogen accounting for total nitro-
gen, %) of Jinhua ham reaches 14–20. During proteolysis, 3.1.2. Cathepsins
about 10% of insoluble proteins are hydrolyzed to soluble In vitro experiments show that cathepsin B and L retain
proteins which are then actively hydrolyzed to small pep- 17.55% and 22.92% of their initial activity at the end of the
tides and FAA. In the soluble fraction, the proportion of ripening stage and 9.31% and 13.66% at the end of the
soluble proteins drops from 71.63% before the beginning post-ripening stage, respectively. Their activities, however,
of the salting stage to 27.57% at the end of post-ripening are greatly affected by temperature, salt content and pH
period, most of the increase of non-protein nitrogen values that change continuously during Jinhua ham pro-
(NPN) occurring at the same time is free amino acids cessing. The activity observed under defined in vitro condi-
and small peptides of molecular weight <1 kDa, account- tions may not reflect that which occurs during ham
ing for more than 95% of the total NPN at the end of pro- processing because of the unfavorable environment in
cessing (see Figs. 1 and 2). This change is likely be the result ham. Two quadratic regression equation models were built
of endogenous enzyme activity since low numbers of to express the effects of temperature, salt content and pH
micro-organisms are found inside dry-cured ham (Toldrá value on the activity of cathepsin B and L individually.
& Etherington, 1988; Zhen, He, Li, Zhou, & Zhang, According to the models, cathepsin B and L always remain
2004) and molds growing on the ham surface do not affect active under production conditions of Jinhua ham, though
Jinhua ham flavor development (Lin et al., 1992). Thus, <5% of the activity in an in vitro environment was observed
endogenous enzymes such as cathepsins, dipeptidyl pepti- during Jinhua ham processing (see Fig. 3). Cathepsin B and
 G.H. Zhou, G.M. Zhao / Meat Science 77 (2007) 114–120 117

12000 In vitro cath. 1000 Practical cath. 30000 25000 Practical DPPI
B activity activity
B activity In vitro DPPI
Practical cath. activity Practical DPPIV
10000 Invitro cath. L activity 25000 activity
800 20000
L activity In vitro DPPIV
activity
8000 20000
U·g–1

U·g–1

U·g–1

U·g–1
600 15000

6000 15000
400 10000
4000 10000

200 5000
2000 5000

0 0 0
0 1 2 3 4 5 6
1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6
Stage Stage
Stage Stage

Fig. 3. Changes of cathepsin B and L activities in in vitro (left) and Fig. 4. Changes of dipeptidase I and IV (DPPI and IV) activities in in vitro
production environments (right) during Jinhua ham processing. Note: (left) and production environments (right) during Jinhua ham processing.
Stage 1, before salting; 2, end of salting; 3, end of sun-drying; 4, middle of Note: Stage 1, before salting; 2, end of salting; 3, end of sun-drying; 4,
ripening; 5, end of ripening; and 6, end of post-ripening. middle of ripening; 5, end of ripening; and 6, end of post-ripening.

maintain strong activity during Jinhua ham processing,

L may play key roles in hydrolyzing insoluble proteins to
although only 7.98%, 5.70% and 9.76% of the initial activ-
soluble proteins, and soluble proteins to peptides as a
ity of AAP, RAP and LAP are, respectively, retained at the
results of in vitro conditions and the long duration of Jin-
end of ripening and 3.05%, 1.90% and 6.02%, respectively,
hua ham processing (Zhao et al., 2005c).
at the end of post-ripening. They are all affected by changes
of temperature, salt content and pH value. The activities of
3.1.3. Dipeptidyl peptidases (DPP)
AAP, RAP and LAP under practical Jinhua ham process-
Experiments in an in vitro environment show that both
ing conditions are estimated at 1.27–4.13%, 1.14–2.21%
dipeptidyl peptidase I (DPPI) and dipeptidyl peptidase IV
and 3.99–17.54%, respectively, according to the mathemat-
(DPPIV) possess considerable activity during Jinhua ham
ical models that include factors that affect the three amino-
processing. The activity of muscle DPPI decreased gradu-
peptidases (see Fig. 5). LAP and AAP exhibit strong
ally before the beginning of ripening and then increased
activity throughout processing, while RAP shows consider-
gradually until the end of post-ripening while the activity
able activity only during the first three processing stages.
of muscle DPPIV decreased continuously. At the end of
The data shows that LAP and AAP are the most important
processing, 141.71% of initial DPPI activity and 11.19%
aminopeptidases in Jinhua ham processing, while RAP
of initial DPPIV activity remained (Zhao et al., 2005a).
may be more effective at the early stages of processing
The change in DPPIV activity contradicts research on
(Zhao et al., 2006; Zhao et al., 2005b).
European dry-cured hams (Sentandreu & Toldrá, 2001c).
This may be due to the differences of micro-organisms
3.2. Lipolysis and oxidation
growing on the ham surface because some micro-organisms
produce an isozyme of DPPI that may penetrate into inner
Lipolysis and lipid oxidation are a major source of ham
muscle and play a role in DPPI activity (Zhao et al.,
flavor compounds. During Jinhua ham processing, intense
2005a).
degradation and oxidation occur in both muscle lipid and
As in the case of cathepsins, the activities of DPPI and
IV are also significantly influenced by temperature, salt
content and pH value. In the mathematical models estab-
lished to illustrate the effects of temperature, salt content 200000
In vitro AAP
9000
Practical AAP
activity activity
and pH value on the activities of DPPI and IV, DPPI 180000
In vitro RAP 8000 Practical RAP
160000
always retains very strong activity under practical Jinhua 140000
activity 7000 activity
6000 Practical LAP
ham processing conditions, whereas the activity of DPPIV In vitro LAP
U·g–1

U·g–1

120000 activity
activity 5000
is always very weak despite its strong activity shown in a 100000
4000
80000
defined in vitro environment (see Fig. 4). Therefore, muscle 60000 3000
2000
DPPI may be a key enzyme responsible for the generation 40000
20000 1000
of dipeptides in Jinhua ham processing (Zhao et al., 2005a). 0
1 2 3 4 5 6 1 2 3 4 5 6
Stage Stage
3.1.4. Aminopeptidases
Toldrá, Flores, and Sanz (1997) have shown that the Fig. 5. Changes of alanyl, arginyl and leucyl aminopeptidase activities in
in vitro (left) and production environments (right) during Jinhua ham
free amino acids in dry-cured hams are primarily generated
processing. Note: Stage 1, before salting; 2, end of salting; 3, end of sun-
from muscle proteins and peptides by the actions of amino- drying; 4, middle of ripening; 5, end of ripening; and 6, end of post-
peptidases. Experiments in vitro demonstrated that alanyl, ripening. AAP, RAP and LAP, respectively, stand for alanyl, arginyl and
arginyl and leucyl aminopeptidase(AAP, RAP and LAP ) leucyl aminopeptidases.
118 G.H. Zhou, G.M. Zhao / Meat Science 77 (2007) 114–120

subcutaneous fat tissues (Huan, 2005; She & Tong, 2005;
Yan, Li, & Jiang, 2005). The lipolysis and, in part, the oxi-
dation are attributed to activities of muscle lipases, includ-
ing phospholipase and lipoxygenase (Huan, 2005; Huan,
Zhou, Zhao, Xu, & Peng, 2005a).
3.2.1. Lipolysis
The results of research on Jinhua ham in the past few
years show that, with the gradual loss of moisture, the pro-
portion of lipid in ham increases from 3.65% at the green
stage to 8.10% at the end of post-ripening (Huan, 2005).
Muscle lipid and subcutaneous fat experience different deg-
radation processes. In muscle, the proportions of both tri-
acylglycerol and free fatty acids increase during Jinhua
ham processing, while phospholipids decrease dramatically
with about 66.67% of muscle phospholipids hydrolyzed
during processing (see Fig. 6). On the other hand, both
the proportions of phospholipids and triacylglycerol in
subcutaneous fat decrease during Jinhua ham processing,
while only the proportion of free fatty acids increases dra-
matically (see Fig. 7). It seems that the phospholipid degra-
dation in muscle and both phospholipid and triacylglycerol
degradation in subcutaneous fat are the main reactions of
ham lipolysis and it appears that phospholipids are a main
100
90
Triacylglycerol, %
source of precursors of ham flavor substances (Huan,
2005).
3.2.2. Lipid oxidation
Lipid oxidation occurs throughout Jinhua ham process-
ing with the most intense lipid oxidation occurring at the
sun-drying stage. Free fatty acids are the main substrates
of the lipid oxidation process. The proportions of saturated
and monounsaturated fatty acids in both muscle and sub-
cutaneous fat tissue increase during Jinhua ham process-
ing. The proportion of monounsaturated fatty acids
increases most dramatically, while the proportion of poly-
unsaturated fatty acids decreases sharply. This indicates
that intense oxidation of polyunsaturated fatty acids has
occurred; especially of linoleic acid which is the fatty acid
with the biggest decrease both in muscle and subcutaneous
tissues during processing (Huan, 2005).
3.2.3. Lipases and phospholipases
According to the optimum pH of different lipases, at
least three types of lipases can be categorized: acid lipases,
neutral lipases and basic lipases. Considering that the pH
value of ham is about 6, basic lipases may not be effective
and are not considered in our study. In vitro experiments
show that lipases and phospholipases are not evenly dis-
25
tributed in muscle and subcutaneous fat tissues. Neutral
lipase demonstrates the highest activity in both muscle
20 and subcutaneous tissues, while acid lipase and phospho-
lipases the lowest. The activity of the three enzymes both

Phospholipids and
free fatty acids, %

80 15 in muscle and subcutaneous fat tissues decreases sharply

during Jinhua ham processing (see Fig. 8). No activity of
70 10 muscle lipases or phospholipases can be detected after
post-ripening. At the end of ripening, 8.31%, 4.18% and
Triacylglycerol
60 5 7.56% of the initial activities of muscle acid lipases, neutral
Phospholipids
Free fatty acids
lipases and phospholipases, respectively, are retained, while
50 0 only 8.26%, 2.94% and 5.36% of the respective initial activ-
1 2 3 4 5 6
ity remained at the end of post-ripening. In subcutaneous
Stage
fat tissues, about 16.16%, 23.05% and 14.77% of the initial
Fig. 6. Changes of muscle lipid composition during Jinhua ham process- activity of acid lipases, neutral lipases and phospholipases,
ing. Note: Stage 1, before salting; 2, end of salting; 3, end of sun-drying; 4, respectively, occur at the end of ripening stage and no
middle of ripening; 5, end of ripening; and 6, end of post-ripening.

60 Acid lipase Acid lipase activity

activity neutral lipase activity
Triacylglycerol Phospholipase activity
neutral lipase
100 Phospholipids 10 50 200
activity
U·h.g protein, × 10 -1

Free fatty acids

-1

Phospholipase
U·h.g protein, × 10

96 8 40 activity
150
Triacylglycerol, %

Phospholipids and
free fatty acids, %

30
92 6
100
20
88 4
50
10
84 2
0 0
1 2 3 4 5 6 1 2 3 4 5 6
80 0 Stage
1 2 3 4 5 6 Stage
Stage
Fig. 8. Changes of in vitro activities of lipases and phospholipase in
Fig. 7. Changes of lipid composition in subcutaneous fat during Jinhua muscle (left) and subcutaneous fat tissues (right) during Jinhua ham
ham processing. Note: Stage 1, before salting; 2, end of salting; 3, end of processing. Note: Stage 1, before salting; 2, end of salting; 3, end of sun-
sun-drying; 4, middle of ripening; 5, end of ripening; and 6, end of post- drying; 4, middle of ripening; 5, end of ripening; and 6, end of post-
ripening. ripening.
 G.H. Zhou, G.M. Zhao / Meat Science 77 (2007) 114–120
0.7 Lipase activity
0.6
Pospholipase activity
U·h.g protein
0.5
0.4
0.3
0.2
0.1
0 make clear contribution to the characteristic aroma
1 2 3 4 5 6
Stage
Fig. 9. Changes of the activities of muscle lipases and phospholipase in
practical Jinhua ham processing. Note: Stage 1, before salting; 2, end of
salting; 3, end of sun-drying; 4, middle of ripening; 5, end of ripening; and
6, end of post-ripening.
lipase or phospholipase activity can be detected in the fol-
lowing stages (Huan, 2005; Huan, Zhou, & Xu, 2005b).
Response surface experiments show that the activities of
muscle lipases and phospholipases are affected by tempera-
ture, salt concentration and the interaction effect of the two
factors. Well-established mathematical models show that
the muscle environment during Jinhua ham processing is
quite favorable for muscle lipases and phospholipases, with
very high activity during most of the processing (see
Fig. 9). This explains the concentration changes of free
fatty acids in muscle during the processing of Jinhua ham
(Huan, 2005; Huan et al., 2005b).
3.2.4. Lipoxygenase
Lipoxygenase is a major enzyme responsible for lipid
oxidation. In vitro experiments show that subcutaneous
fat does not possess lipoxygenase activity. On the other
hand, muscle lipoxygenase displays very strong activity
during Jinhua ham processing. This enzyme increases to
its maximum activity at the end of the salting stage, then
decreases during the following stages with about 66% of
its initial activity retained at the end of the post-ripening
stage. Response surface experiments demonstrate that the
activity of lipoxygenase is influenced by temperature, salt
and nitrate concentration. Interaction effects also exist
between temperature and salt content as well as tempera-
ture and nitrate content. However, the established equation
model indicates that the practical muscle lipoxygenase
activity demonstrated during Jinhua ham processing
explains only about 20% of the oxidation effects. The result
suggests that auto-oxidation, not enzyme catalyzed oxida-
tion, is the major cause of muscle lipid oxidation during
Jinhua ham processing (Huan, 2005).
4. Flavor compounds identification and their changes along
processing stages
4.1. Identification of flavor compounds
More than 330 individual peaks are found in muscle
samples of Jinhua ham during processing and 191 of them
119
have been identified using headspace-solid phase microex-
traction (SPME) and GC/MS techniques (Huan et al.,
2005a; Zhao, 2004). The compounds identified belong to
alkanes and alkenes, aromatic hydrocarbons, alcohols,
aldehydes, ketones, carboxylic acids, esters, oxygenous het-
erocycle compounds, nitrogenous compounds, sulphur
compounds, chloride compounds, amides, and terpenes
(Huan, 2005; Huan et al., 2005a; Zhang, Wang, Liu,
Zhu, & Zhou, 2006; Zhao, 2004). At least 22 compounds
according to artificial sniffing (Tian, Wang, & Xu, 2004).
The compounds include eight aldehydes (3-methyl-butanal,
2-methyl-propanal, hexanal, (Z)-2-heptenal, 2-methyl-
propanal, octanal, phenyl acetaldehyde and nonanal), four
sulphur compounds (dimetyl disulfide, methanethiol, 3-
methylthiol-propanal and dimethyl trisulfide), three
ketones (acetone, 2,3-pentanedione and 2-heptanone),
three heterocycle compounds (methylpyrazine, 2,6-dimeth-
ylpyrazine and 2-pentyl-furan), one alcohol (1-octen-3-ol)
and one ester (ethyl acetate) (Tian et al., 2004).
4.2. Changes of flavor compounds during processing
During Jinhua ham processing, all the flavor com-
pounds increase in total peak area; however, their relative
percentage changes. The percentages of alcohol and ketone
areas drops sharply while aldehydes, acids, lactones, pyra-
zines, sulphur compounds and oxygen heterocyclic com-
pounds increase in area percentage during processing
(Huan, 2005; Huan et al., 2005a; Tian et al., 2004; Tian,
Wang, & Xu, 2005; Tian, Wang, & Xu, 2006; Zhao,
2004) (see Fig. 10).
Flavor is extremely important for dry-cured ham. The
development of flavor in dry-cured ham is a very complex
process and flavor compounds mostly originate from enzy-
matic action and/or chemical reactions such as lipid oxida-
tion, Maillard reactions and Strecker degradations of
muscle protein and fat. Proteolysis and lipolysis constitute
the main biochemical reactions in the generation of flavor
or flavor precursors (Toldrá, 1998) and are mainly attrib-
alcohols aldehydes 2.5 sulphur compounds
lactones
45 ketones acids pyrazines
oxygen heterocyclic compounds
40 2
35
30 1.5
25
%
20 1
15
10 0.5
5
0
1 2 3 4 5 6
0
1 2 3 4 5 6
Stage Stage
Fig. 10. Changes of some of the flavor compounds during Jinhua ham
processing. Note: Stage 1, before salting; 2, end of salting; 3, end of sun-
drying; 4, middle of ripening; 5, end of ripening; and 6, end of post-
ripening.
120 G.H. Zhou, G.M. Zhao / Meat Science 77 (2007) 114–120
uted to the endogenous enzymatic systems in view of the
low microbial counts found inside the hams, the conditions
that are unfavorable for microbial growth (Toldrá & Ethe-
rington, 1988; Zhen et al., 2004) and low microbial enzyme
activity levels (Molina & Toldrá, 1992). Therefore, the
quality and flavor characteristics of Jinhua ham depend
on its raw meat properties and processing technology
(Zhang et al., 2006) and the factors influencing enzyme
activity during the course of processing.
5. Conclusion
Traditional Jinhua ham processing consists of six stages:
green ham fabricating, salting, washing, sun-drying and
shaping, ripening, and post-ripening. Intense proteolysis,
lipolysis and oxidation reactions occur over the course of
processing and, as a result, many characteristic volatile fla-
vor compounds are produced. The content of free amino
acids in final Jinhua ham products is 14–16 times that of
green ham, and 191 volatile compounds have been identi-
fied during processing. Endogenous enzymes may play an
important role in the development of the characteristic fla-
vor of Jinhua ham.
References
Gong, Y. L. (1987). Jinhua ham processing technology. Beijing: Popular
Science Press (China).
Huan, Y. J. (2005). Studying on the changes of lipid and flavor
compounds during processing of Jinhua ham. Doctorial Paper of
NanJing Agricultural University Nanjing, China.
Huan, Y. J., Zhou, G. H., & Xu, X. L. (2005b). Study on time-related
changes of phospholipase during processing of Jinhua ham using
response surface method. Food and Fermentation Industries (China),
31(2), 120–123.
Huan, Y. J., Zhou, G. H., Zhao, G. M., Xu, X. L., & Peng, Z. Q. (2005a).
Changes in flavor compounds of dry-cured Chinese Jinhua ham during
processing. Meat Science, 71, 291–299.
Lin, K. Z., Yang, Y. H., Zhu, S. W., Wang, X. Y., Zhang, S. H., Bu, X. P.,
et al. (1992). Study on relationship of the quality and color, flavor
development of Jinhua ham with mold growth. Meat Research
(China)(2), 10–16.
Molina, I., & Toldrá, F. (1992). Detection of proteolytic activity in
microorganisms isolated from dry-cured ham. Journal of Food Science,
57, 1308–1310.
Sentandreu, M. A., & Toldrá, F. (2001c). Dipeptidyl peptidase activities
along the processing of Serrano dry-cured ham. European Food
Research and Technology, 213, 83–87.
She, X. J., & Tong, Q. Y. (2005). Changes of fatty acid in intramuscular
lipids during processing of Jinhua ham. Food and Fermentation
Industries (China), 31(1), 139–142.
Tian, H. X., Wang, Z., & Xu, S. Y. (2004). Characterization of odor-active
compounds in Jinhua ham by GC-olfactometry. Food and Fermenta-
tion Industries (China), 30(12), 117–123.
Tian, H. X., Wang, Z., & Xu, S. Y. (2005). Research on volatile flavors of
Jinhua ham. Journal of Wuxi University of Light Industry (China),
24(1), 69–73, 83.
Tian, H. X., Wang, Z., & Xu, S. Y. (2006). Separation and identification
of volatile flavors of Jinhua ham by gas chromatography–mass
spectrometry coupled with head space solid phase microextraction.
Chinese Journal of Chromatography, 24(2), 176–180.
Toldrá, F. (1998). Proteolysis and lipolysis in flavour development of dry-
cured meat products. Meat Science, 49(Suppl. 1), S101–S110.
Toldrá, F., & Etherington, D. J. (1988). Examination of cathepsins B, D,
H and L activities in dry-cures hams. Meat Science, 59, 531–538.
Toldrá, F., Flores, M., & Sanz, Y. (1997). Dry-cured ham flavour:
Enzymatic generation and process influence. Food Chemistry, 59(4),
523–530.
Wu, A. F., Sun, C. Y., & Sun, G. Q. (1959). Jinhua ham. Beijing: Light
Industry Press (China).
Yan, W. J., Li, X. M., & Jiang, Y. X. (2005). Fatty acids in intramuscular
and subcutaneous fat of Jinhua ham. Food and Fermentation Indus-
tries, 31(2), 124–126.
Zhang, J. H., Wang, L., Liu, Y., Zhu, J. H., & Zhou, G. H. (2006).
Changes in the volatile flavour components of Jinhua ham during the
traditional ageing process. International Journal of Food Science and
Technology, 41, 1033–1049.
Zhao G. M. (2004). Studies on the effects of muscle proteolytic enzymes in
the processing of Jinhua ham. Doctorial Paper of NanJing Agricul-
tural University Nanjing, China.
Zhao, G. M., & Zhou, G. H. (2003). The properties and traditional
processing technology of ‘Jinhua Huotui’. Food Science and Technol-
ogy (China), 12(z1), 322–326.
Zhao, G. M., Wang, Y. L., Tian, W., Zhou, G. H., Xu, X. L., & Liu, Y. X.
(2006). Changes of arginyl and leucyl aminopeptidase activities in
biceps femoris along Jinhua ham processing. Meat Science, 74(3),
450–458.
Zhao, G. M., Zhou, G. H., Tian, W., Xu, X. L., Wang, Y. L., & Luo, X.
(2005b). Changes of alanyl aminopeptidase activity and free amino
acid contents in biceps femoris during processing of Jinhua ham. Meat
Science, 71(4), 612–619.
Zhao, G. M., Zhou, G. H., Wang, Y. L., Xu, X. L., Huan, Y. J., & Wu, J.
Q. (2005c). Time-related changes in cathepsin B and L activities during
processing of Jinhua ham as a function of pH, salt and temperature.
Meat Science, 70(2), 381–388.
Zhao, G. M., Zhou, G. H., Xu, X. L., Jin, Z. M., Huan, Y. J., Chen, M.
W., et al. (2004). Optimization of traditional processing technology
of Jinhua ham and correlation analysis on main technological
parameters. Food and Fermentation Industries (China), 30(5),
119–123.
Zhao, G. M., Zhou, G. H., Xu, X. L., Peng, Z. Q., Huan, Y. J., Jin, Z. M.,
et al. (2005a). Studies on time-related changes of dipeptidyl peptidase
during processing of Jinhua ham using response surface methodology.
Meat Science, 69(1), 165–174.
Zhen, Z. Y., He, Z. F., Li, H. J., Zhou, G. H., & Zhang, J. H. (2004).
Prospect of study on the Jinhua ham maturating and microorganism
fermenting. Sichuan Food and Fermentation (China), 40(3), 1–3.
 MEAT
SCIENCE
Meat Science 77 (2007) 121–135
www.elsevier.com/locate/meatsci

Review

Effects of metabolic modifiers on carcass traits and meat quality

M.E. Dikeman
Department of Animal Sciences and Industry, Weber Hall 226, Kansas State University, Manhattan, KS 66506, USA

Received 7 March 2007; received in revised form 3 April 2007; accepted 5 April 2007

Abstract

Much research has been conducted and published about metabolic modifiers that increase growth rate, improve feed efficiency,
increase carcass leanness, and decrease carcass fatness. Most of these metabolic modifiers have been developed to improve efficiency
and profitability of livestock production and to improve carcass composition, with fewer of them developed and researched specifically
to improve meat quality. Some of the metabolic modifiers can have negative effects on visual and sensory meat quality, especially when
not used as recommended. This review evaluates the various kinds of metabolic modifiers that have been researched for their effects on
production efficiency, carcass composition, and meat quality. Nutritional composition of meat generally is improved from use of most of
the metabolic modifiers, visual quality is improved by others, but some can have a negative effect on marbling and tenderness. Anabolic
steroid implants are very cost effective and practical for beef cattle production but aggressive implants used within 70 days of slaughter or
too frequent use of them will reduce tenderness and marbling. Somatatropin and approved b-agonists are very effective in improving
growth performance and carcass leanness in pigs, and b-agonists are effective in cattle, but improper use of them can have negative effects
on marbling and tenderness. Feeding supplemental levels of vitamin E is quite beneficial for improving meat color and shelf-life of beef,
lamb, and pork, whereas not supplementing diets with vitamin A has potential for improving marbling in cattle. Immunocastration
shows promise for capitalizing on the efficiency of muscle growth of young boars up to a few weeks before slaughter, at which time boar
taint is prevented and marbling is improved by immunocastration. Potential exists for improving the fatty acid profile of lipids and
increasing conjugated linoleic acid content in beef through dietary manipulation. Supplementing swine diets with conjugated linoleic acid
can improve carcass composition of swine, but is not yet cost effective to use. Dietary inclusion of magnesium, manganese, or chromium
in diets of pigs and sheep has potential to improve meat color and water-holding capacity. Although, not all of these metabolic modifiers
are approved in all countries, proper use of the ones that are approved offers opportunities for economically improving production effi-
ciency and carcass leanness while maintaining acceptable marbling and tenderness, while some provide opportunities to enhance meat
color and quality.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Metabolic modifiers; Growth performance; Meat quality; Dietary manipulation; Carcass traits

Contents

1. Metabolic modifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

2. Anabolic steroids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
3. Vitamin D3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
4. Vitamin A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5. Vitamin E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
6. Somatotropin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
7. b-agonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

E-mail address: mdikeman@ksu.edu

0309-1740/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.04.011
122 M.E. Dikeman / Meat Science 77 (2007) 121–135

8. ‘Designer’ lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

9. Chromium, carnatine, magnesium, niacin, manganese and betaine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
10. Immunocastration . . . . . . . . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
11. Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

1. Metabolic modifiers agonists, (4) vitamins or vitamin-like compounds fed in

Metabolic modifiers are defined as compounds that are
either fed, injected, or implanted in animals to improve rate
of gain, improve feed efficiency, increase dressing percent,
increase carcass meat yield percentage, improve visual meat
quality, extend shelf-life, improve meat’s nutritional pro-
file, or improve meat palatability. Most metabolic modifi-
ers have been developed and researched to improve
growth performance and carcass composition, with fewer
of them developed or researched specifically to improve
meat quality. However, research activity on dietary manip-
ulation of livestock to improve or enhance meat quality has
been extensive in the past few years, particularly in swine.
Several in-depth review papers could be written on specific
categories or types of metabolic modifiers. This paper will
present a general review of the effects of metabolic modifi-
ers on carcass composition and meat quality. Little data
are presented in Tables because the review would be exces-
sively long, even when only the most relevant data were
presented for all metabolic modifiers. Ellis and McKeith
(1999) reviewed the effects of nutrition on the quality of
meat from non-ruminants and Owens and Gardner
(1999) reviewed the effects of ruminant nutrition on meat
quality at the 1999 Reciprocal Meat Conference. There-
fore, the effects of type of feed, energy source and related
topics on beef and pork meat quality will not be discussed
extensively in this review. Because there is considerably less
research on the effects of metabolic modifiers in sheep and
goat production or other minor meat species, the review
will focus mainly on cattle and pigs, with some inclusion
of lamb research.
Emphasis of this review is on those metabolic modifi-
ers that are, or likely will be approved for use in cattle
and pigs in the US and other developed countries. Dis-
cussion of meat quality will include factors that affect
visual quality, such as color, marbling, firmness, and
maturity; factors that affect processing or packaged dis-
play quality, such as pH, color, water-holding capacity,
and antioxidant potential; sensory traits of tenderness,
juiciness and flavor; safety; or characteristics of meat
that affect human nutritional quality. Meat quality will
be emphasized more than carcass composition or meat
yield percentage.
Metabolic modifiers will be categorized as: (1) anabolic
steroids, (2) somatotropin, (3) phenethanolamines or beta
supra-nutritional levels, (5) ‘designer’ lipids and (6) other
modifiers.
2. Anabolic steroids
Implants containing various anabolic steroids are used
widely by the beef cattle industry in the US in growing
and finishing cattle because of economic incentives to
increase growth rate and improve efficiency of feed utiliza-
tion (Dikeman, 1997). Heifers and steers show the greatest
response to steroid implants, whereas bulls show minimal
response. In fact, bulls may deposit more, rather than less
fat when implanted. Implanted steers often achieve a
growth rate and feed efficiency similar to bulls (Fisher,
Wood, & Whelehan, 1986). In general, estrogenic implants
are more effective in steers and androgenic implants more
effective in heifers. Combination implants generally pro-
duce an additive affect in both steers and heifers compared
to implants containing only an estrogenic or androgenic
compound.
A review of the literature shows that anabolic steroid
implants increase growth rate 10–20% when cattle are
slaughtered at the same age or time on feed. Preston
(1999) stated that growth rates can be improved by as
much as 30% and feed efficiency by as much as 15% when
slaughtered at the same live weight. In addition, carcass
leanness can be improved up to 8% when slaughtered at
the same live weight, but most implanted cattle are slaugh-
tered at heavier weights than non-implanted cattle. In gen-
eral, breeds or types of cattle that have the greatest
potential for muscle growth show the greatest response to
implants. Greater responses often are observed during the
first few weeks after implantation, suggesting a peak and
then a decline in circulating concentrations of the hor-
mones (Hayden, Bergen, & Merkel, 1992).
Dikeman (2003) published an extensive review of
research on the effects of anabolic steroid implants in cat-
tle production and summarized that the effects of
implants on meat quality show that some ‘aggressive’
implants and implant strategies have the potential to
increase the proportion of dark cutters, decrease mar-
bling and the percentage of Choice carcasses and(or)
decrease tenderness. The first two effects have direct
and immediate negative economic consequences, whereas
the latter affect likely results in decreased consumer con-
 M.E. Dikeman / Meat Science 77 (2007) 121–135 123

fidence and demand for beef, which has a significant Dikeman (2003) stated that cattle absolutely should not
long-term negative economic consequence. Dikeman be re-implanted with aggressive (combination implants
(2003) concluded that there are no major research reports containing trenbolone acetate) or moderately aggressive
demonstrating that anabolic steroid implants improve implants within 70 days of slaughter and special care
visual meat quality or meat palatability. The majority should be used when handling implanted cattle during
of research shows some reduction in marbling and(or) hot weather. Manufacturer recommendations and warn-
tenderness when compared to non-implanted controls, ings should be followed very closely to minimize the
although not all reductions are statistically significant. chances for negative effects of implants on meat quality.
Obviously, the type of implant or re-implant strategy, Implants are much too effective in improving efficiency of
the type of cattle, and the length of time cattle are on production to not use them, but their use needs to be man-
feed can have significant effects on the results. Morgan aged extremely well.
(1997) reported that the ‘aggressive’ use of anabolic In more recent research, Smith, Duckett, Asain, Sonon,
implants commonly compromises beef carcass quality and Pringle (2007) compared cattle fed a high-concentrate
grades and increases the incidence of dark cutting car- diet with no implant and those receiving two Synovex-
casses. Furthermore, Foutz, Gill, Dolezal, Gardner, and PlusTM (combination trenbolone acetate and estrogen)
Botts (1990), Samber et al. (1996), Morgan (1997) and implants. Implanting did not have a direct effect on intra-
Roeber et al. (1999) indicated that some aggressive muscular lipid deposition. These authors did find that
implanting strategies have been implicated as possibly implanting reduced tenderness after 14 days of aging but
causing reduced meat palatability, specifically tenderness. shear values were still in an acceptable range.
In general, both trained and consumer sensory panels Watson (submitted for publication, personal communi-
rate steaks from non-implanted controls as being more cation) conducted a meta-analysis of studies regarding
tender than those from implanted steers and heifers. the effects of hormone growth promotant use on beef pal-
Warner–Bratzler shear force (WBSF) values generally atability measured by objective testing and by consumer
are lower (more tender) for steaks from non-implanted studies. These authors stated that a meta-analysis is a more
controls than for those from implanted cattle (Morgan, rigorous alternative to the ‘casual, narrative discussions’ of
1997; Table 1). However, because at least 90% of all research results. This analysis included 32 studies with 22
fed cattle in the US are implanted, it is not very mean- treatment-control comparisons for sensory tenderness eval-
ingful to evaluate the effects of various implants and uation; and 18 studies with 24 treatment-control compari-
implant strategies on marbling and palatability in com- sons for shear force. Their hypothesis included the
parison to non-implanted controls. It is more meaningful assumption that there is little difference among the different
to evaluate whether or not there are differences among hormone growth promotants in their effect on palatability.
implants or implant strategies in their effects on the inci- The abstract from this analysis stated ‘‘we present evidence
dence of dark cutters, reduced marbling, or decreased that suggests strongly that growth promotant implants
meat palatability. have a negative effect on beef palatability.’’ Results from
the meta-analysis showed that shear force was increased
2.55 N and sensory tenderness score was decreased by 5
Table 1 units on a 100 point scale. In my opinion, however, the
Warner–Bratzler shear force value change stratified by implant strength rather modest decrease in tenderness shown from this
and typea meta-analysis is not overly concerning. The review by Mor-
First implant Second implant Third implant WBSFb (N) gan (1997) presented in Table 1 shows that there are differ-
Non-implanted ences among hormone growth promotants in their effect on
MEc – – +0.49 tenderness; some have a distinct negative effect, whereas
ME ME ME +4.11 others have a rather modest effect.
A – – +5.78 In a study by Thompson, Polkinghorne, Porter, and
ME/A ME/A – +6.96 Hunter (in press), 509 Brahman and F1 Brahman crossbred
SE – – +4.12
steers were either not implanted or implanted with 20 mg
SE SE – +4.31
SE/A – – +4.80 estradiol-17b, approximately every 100 days. Half were fin-
SE/A SE/A – +6.47 ished in a feedlot and half were finished on grass and
MC – – +0.98 slaughtered at one of three endpoints based on weight:
MC MC – +7.54 220 kg (Australian domestic), 280 kg (Korean), or 320 kg
SC – – +7.54
(Japanese). Small blocks of meat were cooked in a water
SC SC – +5.78
a
bath for 1 h at 70 C. Sensory tenderness scores for the
Source: OSU Implant data base (From Morgan, 1997).
b domestic endpoint were lowered 7.0 (100 point scale)
WBSF: Warner–Bratzler shear force value, N change in WBSF com-
pared to non-implanted controls. for feedlot and 12.7 for grass-fed cattle when compared
c
Implant classification: ME, SE, A, MC and SC are mild estrogen, to non-implanted cattle, whereas scores for the Korean
strong estrogen, androgen, mild combination and strong combination, and Japanese endpoints were only lowered from 1.3 to
respectively; – = not implanted. 4.5. In my opinion, the reduction in sensory tenderness
124 M.E. Dikeman / Meat Science 77 (2007) 121–135
scores for the two heavier slaughter groups is not concern-
ing because there can be numerous things pre- and post-
mortem to decrease tenderness to a greater extent than in
this study.
In a large study involving 2,748 Bos indicus crossbred
cattle, Barham et al. (2003) reported that implanting with
Synovex-S twice or Synovex-S followed by Revalor-S
increased shear force values compared to non-implanted
controls after only 14 days of aging, but not after 21 days
of aging. These authors determined that consumers failed
to detect any differences in steak samples related to implant
treatment after 7 or 14 days of aging.
Limited research was found on the effects of anabolic
implants in poultry, sheep, and pigs. Anabolic steroids
are not approved for growth regulation in pigs in the US
and numerous other countries. Even so, Lee et al. (2002)
studied the effects of implanting castrate pigs weighing
59 kg with Revalor-H (trenbolone acetate plus estradiol-
17b) on performance, carcass composition and meat qual-
ity. Overall, pork quality was not affected by implantation,
but backfat thickness was reduced with implantation.
3. Vitamin D3
Feeding vitamin D3 has received attention for its poten-
tial to improve meat tenderness. It has been shown to
increase blood and muscle calcium levels. Dikeman
(2003) summarized research on feeding vitamin D3 and
concluded that it may increase tenderness in beef early
postmortem, and improve firmness and color and decrease
drip loss in pork. However, feeding vitamin D3 likely will
not be adapted by the industry because it reduces feed
intake and performance. Moreover, Puls (1994) suggested
that supplementing cattle with at least 2 · 106 IU d 1 vita-
min D3 could result in cattle toxicity. In addition, there are
important questions that need to be addressed about its
potential toxicity in humans. Consequently, feeding vita-
min D3 as a metabolic modifier to improve meat quality
may not be adapted by the industry until more research
is conducted and it is proven consistently beneficial without
any human health risks.
4. Vitamin A
Recently, the role of vitamin A in adipocyte differentia-
tion and marbling development has been investigated.
Steers finished on diets containing low levels of vitamin
A have produced carcasses with increased marbling scores
(Gorocica-Buenfil, Fluharty, & Loerch, 2005; Kruk et al.,
2004; Nade, Hirabara, Okumura, & Fujita, 2003; Oka,
Maruo, Miki, Yamasaki, & Saito, 1998). Marbling scores
have been negatively correlated with concentrations of
vitamin A in blood (Adachi et al., 1999; Nakai, Kita, Hase-
gawa, & Hiramitsu, 1992; Torri, Matsui, & Yano, 1996)
and liver (Oka et al., 1998). Most of these studies have uti-
lized Japanese Black or Wagyu cattle that have a very high
propensity for marbling deposition. Furthermore, the
effects of high levels of vitamin A seem to be dependent
on age of cattle. In the published research by Oka et al.
(1998), these authors stated that high vitamin A decreased
marbling in 15-mo old cattle, but not in 21- or 23-mo old
cattle. Serum vitamin A diverged more than 30 lg dl 1
over a 16-mo period in the 15-mo old steers, which was
the only group with differences in marbling. Circulating
vitamin A diverged only 19.6 and 23.2 lg dl 1 in treatment
periods of only 10 and 6 mo for the 23- and 25-mo old cat-
tle, respectively. This difference in circulating vitamin A
was only observed in the last 30 days before slaughter in
the 25-mo old cattle. This suggests that duration and(or)
extent of vitamin A depletion or supplementation may be
important factors that affect marbling development in
addition to cattle age. Gorocica-Buenfil et al. (2005) dem-
onstrated that marbling score and percentage of USDA
Choice carcasses were 10% higher in steers fed diets with
no supplemental vitamin A. Retinoic acid, a derivative of
vitamin A, regulates adipose cell differentiation and, possi-
bly proliferation (Pairault, Quignard-Boulange, Dugail, &
Lasnier, 1988). This suggests that high vitamin A might
negatively affect differentiation of adipocytes and, conse-
quently, marbling.
Pigs fed a diet with no supplemental vitamin A had
higher percentages of longissimus intramuscular fat without
having detrimental effects on growth performance
(D’Souza, Pethick, Dunshea, & Mullan, 2003). In contrast
to the published results for cattle and pigs, Arnett, Dik-
eman, Spaeth, Johnson, and Hildabrand (2006) reported
that lambs fed diets with 1,365 IU vitamin A kg 1 diet dur-
ing the growing and finishing phases to slaughter had
higher percentages of intramuscular fat in the longissimus
muscle than those fed diets with no supplemental vitamin
A. Lambs fed no supplemental vitamin A during the 56-
day growing phase and finished on high vitamin A, and
lambs fed diets with high A during the growing phase
and finished with no supplemental vitamin A were interme-
diate in longissimus intramuscular fat. Lambs fed no sup-
plemental A also had the greatest increase in serum fatty
acids. There were no differences in circulating retinol levels
from day 0 to day 56, but then became higher at 84 days for
lambs on the high vitamin A diet. These results suggest that
there may be a species interaction on the effects of supple-
mental vitamin A on intramuscular fat deposition. Seibert,
Pitchford, Kuchel, Kruk, and Bottema (2000) reported that
adipose tissue from lambs fed a diet containing low b-car-
otene had decreased C18:0 (stearate) and increased C18:1
n-9 cis (oleate), resulting in 14.9 percentage points reduc-
tion in saturated fat. In addition, there was a 10 C lower
melting point for fat from lambs on the diet with low b-car-
otene than those on the high b-carotene diet. Seibert et al.
(2000) stated that this contradiction with results from cattle
might simply be that cattle do not metabolize b-carotene as
well as sheep.
In all of these studies, there have been no significant
effects of vitamin A supplemental level on growth rate, feed
intake, or carcass composition. However, retinoic acid, a
 M.E. Dikeman / Meat Science 77 (2007) 121–135 125

form of vitamin A, regulates growth hormone gene expres- Arnold et al. (1992) reported that color stability during
sion (Bedo, Santisteban, & Aranda, 1989). Dalke et al. retail display of longissimus lumborum steaks from Holstein
(1992) demonstrated reduced marbling scores in steers trea- steer calves supplemented with vitamin E was extended by
ted with growth hormone. Therefore, vitamin A may be 2.5–4.8 days (Fig. 3); color stability for gluteus medius
affecting marbling development indirectly by regulating steaks was extended 1.6–3.8 days. Liu, Scheller, Arp,
growth hormone secretion. Color shelf-life of beef could Schaefer, and Frigg (1996) fed Holstein steers 0, 250, 500
be negatively affected by not supplementing diets with vita-
min A. Muramoto, Nakanishi, Shibata, and Aikawa (2003)
reported that 7500 mg hd 1 d 1 of b-carotene supplemen-
tation for 28 days extended the color shelf-life of semi-
membranosus and longissimus lumborum muscles by 1.5
and 3 days, respectively, based on a threshold value of
20% metmyoglobin.

5. Vitamin E

Numerous studies have shown that feeding supra-nutri-

tional levels of vitamin E improves meat color and can
extend shelf-life of beef, lamb, and pork. There have been
no reports of negative effects on feed intake or performance
of cattle or pigs fed vitamin E, unlike supplemental feeding
Fig. 2. TBA numbers of fresh ground sirloin from control and vitamin E-
of vitamin D3. Therefore, this metabolic modifier is practi- supplemented Holstein steers during storage at 4 C. N = 11 for each
cal and effective, but feeding high levels of vitamin E group; standard error bars are indicated. (From Faustman et al., 1989).
should not increase costs of production to the point of
not being cost effective.
Dikeman (2003) and Dunshea, D’Souza, Pethic, Harper,
and Warner (2005) published reviews on effects of feeding
high levels of vitamin E on meat quality. Dikeman (2003)
summarized the research showing that dietary vitamin E
supplementation in cattle results in accumulation of a-
tocopherol in muscle tissue, which acts as an antioxidant
to delay lipid and myoglobin oxidation. Consequently,
color stability and retail shelf life of beef are prolonged
(Faustman et al., 1989; Figs. 1 and 2). Asghar et al.
(1991) and Monahan, Asghar, Gray, Buckley, and Morris-
sey (1992, 1994) published data showing that vitamin E
decreases lipid oxidation and drip loss while improving
the color of pork cuts.

Fig. 3. Effect of level of vitamin E supplementation on area of

Fig. 1. Metmyoglobin accumulation during storage at 4 C for fresh
ground sirloin patties from control and vitamin E-supplemented Holstein
steers. N = 11 for each group; standard error bars are indicated. (From
Faustman et al., 1989).
discoloration (a) and color score (b) of longissimus lumborum (loin) steaks
(n = 34 steers). Color scores were associated with the following descriptive
terms: 3 = moderately light cherry red, 4 = cherry red, 5 = slightly dark
red, and 6 = moderately dark red. E0 = 0 IU d 1 of supplemental vitamin
E; E500 = 300 IU d 1 of actual supplemental vitamin E. (From Arnold
et al., 1992).
126 M.E. Dikeman / Meat Science 77 (2007) 121–135
Table 2
Dose and duration effects of supplemental vitamin E on color display life for three bovine muscles held in vacuum storage for 14 days
Vitamin E (mg d 1) Longissimus lumborum
42 days 126 days
0 3.3 4.7
250 5.7 6.7
500 6.0 7.7
2,000 8.7 10.0
Duration 5.9a 7.2b
Muscle 6.6d
a–c
Means across durations within muscle or among doses lacking a common superscript letter differ (P < 0.01).
d–f
Means across muscles lacking a common superscript letter differ (P < 0.01). (From Liu et al., 1996).
or 2000 IU d 1 vitamin E for durations of 42 or 126 days
and reported that color display of fresh beef was extended
0.9–1.8 days from the lowest to highest feeding level (Table
2). Liu et al. (1996) determined that color display life across
the longissimus lumborum, semimembranosus and gluteus
medius muscles stored under vacuum until 14 days and then
displayed under simulated retail conditions was extended
2.0 days (E250) to 5.0 days (E2000). Collectively, supple-
mentation of 500 mg of a-tocopherol acetate per steer daily
improved the mean color display life of the three muscles
by 2.3 days, or essentially 100%. Lynch, Kerry, Buckley,
Faustman, and Morrissey (1998) reported that supplement-
ing Friesan steers with 2000 IU a-tocopherol acetate kg 1
feed d 1 for 50 days resulted in greater color and lipid oxi-
dative stability for longissimus lumborum, gluteus medius
and psoas major muscles than steaks from steers fed the
basal diet after 7 days of retail display.
Supplementation of lamb diets containing at least
1000 lg g 1 a-tocopherol has also been proven effective
for improving lamb meat color and shelf-life stability (Gui-
dera, Kerry, Buckley, Lynch, & Morissey, 1997; Wulf
et al., 1995). Supplementing diets for pigs with at least
200 IU kg 1 of feed reduced lipid oxidation (Jensen et al.,
1997), reduced drip loss (Cheah, Cheah, & Krausgril,
1995), and enhanced color stability (Lanari, Schaefer, &
Scheller, 1995), whereas drip loss and color stability in
other studies were not affected (Cannon et al., 1995; Jensen
et al., 1997). Hoving-Bolink, Eikelenboom, van Diepen,
Jongbloed, and Houben (1998) reported that color stability
was improved in the longissimus muscle after 6 days, but
not in the psoas major for pigs fed extra vitamin E. Cannon
et al. (1996) reported that vitamin E significantly reduced
TBARS of pork after 3 and 5 days of retail display after
0, 14, 28 and 56 days of vacuum storage. However, the
effects of feeding vitamin E on sensory panel attributes
were negligible. Lanari et al. (1995) noted that the improve-
ment in pork muscle color stability produced by dietary a-
tocopherol supplementation was not as profound as has
been reported for beef muscle.
In more recent research since the review published by
Dikeman (2003), Guo et al. (2006a) showed that drip loss
was decreased with increased vitamin E (200 and 400 vs.
40 IU Vitamin E kg 1 of feed) and that both drip loss
and pH were improved with 6 or 9 weeks of feeding vs. 3
Semimembranosus Gluteus medius Dose
42 days 126 days 42 days 126 days
1.0 2.3 1.0 1.7 2.3a
3.3 5.3 2.0 3.0 4.3b
3.3 5.3 1.0 4.3 4.6b
6.3 8.3 4.7 6.0 7.3c
3.5a 5.3b 2.2a 3.8b
4.4e 2.9f
weeks. Also, less lipid oxidation was detected in fresh
ground pork from pigs fed greater concentrations of vita-
min E after 4 days of storage, and in pork chops after 6
months of freezer storage. In another study, Guo et al.
(2006b) evaluated the effects of 200 IU vitamin E kg 1 of
feed and three levels of fat in the diet. These authors stated
that meat quality characteristics did not differ among treat-
ments, but the high level of vitamin E had a beneficial effect
on the oxidative stability of pork as indicated by TBARS
values. Also, adding vitamin E acetate led to greater mono-
unsaturated and total unsaturated fatty acid proportions in
neutral lipids of muscle and adipose tissues.
Supplementing cattle and pig diets with supra-nutri-
tional levels of vitamin E consistently shows an advantage
in color display life and reduced oxidative deterioration of
meat in various packages and chilled states but the effects
in beef are more pronounced than in pork. Williams, Frye,
Frigg, Schaefer, and Scheller (1992) conducted a blind
study of consumers regarding beef from cattle fed
500 IU steer 1 d 1 of vitamin E for 100–120 days and beef
from cattle not fed vitamin E. They found 3.6 percentage
points reduction in losses in retail value from the cattle
fed supplemental vitamin E. Liu, Lanari, and Schaefer
(1995) described the cost/benefit ratio of this technology
for the US beef industry. The cost of supplementing
500 IU of vitamin E d 1 for 126 days is estimated to be
$3 per animal. If retailers could improve their receipts by
3.6% (Williams et al., 1992), this suggests a financial gain
to the US beef industry of $792 million annually. The ben-
efit/cost ratio for the packing, fabrication, distribution, and
retail marketing segments of the beef industry would be
10.4:1. The only issue that needs to be worked out is for
cattle feeders and swine producers to be compensated for
the additional cost of feeding higher levels of vitamin E.
It may also require a method to rapidly verify that cattle
actually received adequate vitamin E supplementation
when marketed with that guarantee. The entire production,
processing, and retail segments of the meat industry would
gain from including vitamin E in the diets of livestock.
6. Somatotropin
In general, somatotropin (ST) administration does not
significantly alter growth or composition in avian species
 M.E. Dikeman / Meat Science 77 (2007) 121–135
(Beermann, 1994). Recombinantly produced porcine
somatotropin (rpST) has been approved for use in swine
production in several countries, but not in the US. It is
delivered as an injection one to three times per week. Aver-
age daily gain is increased by 20% with 150 lg rpST kg 1
body weight per day and feed conversion efficiency is
improved throughout an even greater dose range (Beer-
mann, 1994). Carcass protein accretion rates are increased
up to 74% concurrent with an 82% decrease in lipid accre-
tion rate when rpST is administered from 30 to 90 kg body
weight. The increase in protein deposition generally is con-
sidered a result of increased protein synthesis instead of
decreased protein degradation. On the other hand, dressing
percentage is not increased with rpST administration, pri-
marily because of increased organ weights. Growing rumi-
nants also respond to exogenous rbST administration in a
dose-dependent manner, but responses are generally of les-
ser magnitude than those observed in pigs (Crooker et al.,
1990). Currently, rbST and roST are not approved for beef
cattle or lamb production.
Dunshea et al. (2005) conducted a meta-analysis of the
literature and concluded that rpST decreased intramuscu-
lar fat an average of 12%, increased shear force an average
of 9%, but reduced drip loss by 6%. In addition, the limited
literature on consumer preferences suggests that there is a
9% decrease in tenderness. However, the increases in shear
force and decreases in consumer sensory tenderness can be
as much as 35%, depending on dosage, genotype, and other
management variables. There were no negative effects on
consumer-evaluated flavor, but consumer-evaluated juici-
ness was decreased in a majority of the studies; as much
as 20% in some studies.
Dikeman (2003) concluded that the large majority of
studies show that shear force is increased and sensory panel
tenderness is reduced for chops from pigs treated with
rpST. Thiel (1991) reported a rather dramatic difference
of 2.27 kg in the mean values between the controls and
the highest rpST dose.
McKeith, Lan, and Beermann (1994) concluded that ST
decreases intramuscluar lipid content in both pork and beef
in a dose-dependent manner by 20–50%. Processing of bel-
lies into bacon also can be a problem because of the thicker
skin of rpST treated pigs and because of thinner bellies.
However, the sensory properties of cured products from
rpST treated pigs do not appear to be affected.
Growing sheep also respond to exogenous bovine
somatotropin (rbST) and ovine somatotropin (roST), and
carcass fatness is significantly reduced (Beermann et al.,
1990; McLaughlin et al., 1994). Beermann et al. (1990)
found that ewe lambs exhibited greater reductions in fat
accretion and greater responses in growth rate than wethers
when roST was administered over an 8-week period prior
to slaughter.
Clearly, roST would have to be effective in an implant or
prolonged release form before the sheep industry would
adopt this technology. Sheep often graze or run in larger
lots than pigs, and their quick, flighty nature would make
127
it extremely impractical to inject them daily or bi-weekly
with oST. An implant release form of roST might be
accepted readily by the sheep industry, because no effective
anabolic steroid implant is currently available for use in
sheep.
There are no scientific data to justify preventing the
approval of somatotropin for use in swine, sheep, and beef
cattle production. Somatotropin is extremely effective in
improving growth performance and meat yield percentage.
It will decrease marbling significantly and decrease tender-
ness a majority of the time. It has a neutral to slightly neg-
ative effect on color and firmness and a negative effect on
commercial bacon production because of the thicker skin
and thinner bellies.
7. b-agonists
Only ractopamine hydrochloride for use in pigs (Pay-
lean) and cattle (Optaflexx) and zilpaterol hydrochloride
(Zilmax) for use in cattle will be discussed in this manu-
script. Only these two are discussed because they are
approved for use in the US and several other countries,
and zilpaterol does not have the potential negative potency
or pharmalogical activity of products like clenbuterol or
cimaterol. Zilpaterol is synthetic and is not a phenethanol-
amine like clenbuterol or cimaterol. Dikeman (1991) dis-
cussed significant negative effects of the b-agonists
clenbuterol and cimaterol on meat tenderness in ruminants
and potential toxicity effects in humans from consuming
meat or other edible organs from animals fed these b-
agonists.
Ractopamine hydrochloride for pigs has positive effects
on growth rate, feed efficiency, dressing percent, and car-
cass composition when fed at dosages of 5.0 and 9.9 mg
of ractopamine kg 1 of feed for 30–50 days before slaugh-
ter, but most of the research for approval was done in the
1980s and early 1990s on moderate-lean-growth pigs.
McKeith et al. (1994) summarized the few studies that have
been conducted on the effects of ractopamine hydrochlo-
ride on visual meat quality and sensory traits of both beef
and pork and stated that the effects on marbling, color and
firmness in both species are neutral to positive. However,
some studies have shown an increase (8.33 and 4.80 N) in
WBSF when ractopamine was fed at its highest level
(20 mg kg 1) (Aalhus et al., 1990; Uttaro et al., 1993).
Stites et al. (1991) and Uttaro et al. (1993) stated that feed-
ing ractopamine hydrochloride will increase carcass lean-
ness and should increase processed ham yields.
Schluter, Preston, Davis, Ramsey, and Miller (1991)
studied the effects of feeding ractopamine (marketed as
OptaflexxTM in the US) at 0, 10, 20 or 30 ppm 46 days
before slaughter on feedlot cattle. At 20 and 30 ppm,
growth rate, feed efficiency, and final live and carcass
weights were increased over the controls. However, the
average daily gain of those steers was relatively low
(1.05 kg d 1). Ractopamine did not decrease USDA qual-
ity grade. Avendano-Reyes et al. (2006) evaluated the
128 M.E. Dikeman / Meat Science 77 (2007) 121–135
Fig. 4. Effects of different slaughter and post-slaughter scenarios on tenderness of the longissimus thoracis and semitendinosus muscles from cattle fed
zilpaterol for 30 days vs. controls. (From Strydom et al., 1998).
effects of feeding ractopamine at 300 mg steer 1 d 1 for the
last 33 days on feed and reported decreased (P < 0.05) feed
intake, improved (P < 0.01) gain/feed ratio, increased
(P < 0.05) dressing percent and carcass weight, and
increased (P < 0.05) shear force (46.6 vs. 42.3 N) compared
to controls.
Zilpaterol hydrochloride (Zilmax) distinctly improves
growth performance, dressing percent, and carcass mus-
cling in cattle (Dikeman, 2003). However, Strydom, Osler,
Nel, and Leeuw (1998) reported that zilpaterol supplemen-
tation resulted in a decrease in M. longissimus dorsi muscle
tenderness, but the negative effect seemed to be a function
of ineffective electrical stimulation and inadequate post-
mortem aging. When electrical stimulation and adequate
aging were used, the reduction in tenderness was minor
(Fig. 4). Strydom and Nel (1999) supplemented diets with
0.15 mg zilpaterol for the final 15, 30 or 45 days of the feed-
lot period until 48 h before slaughter. Sensory evaluated
tenderness and juiciness, and shear force of the longissimus
muscle were negatively affected by feeding zilpaterol for 45
days but not for 15 or 30 days. Strydom, Buys, and Stry-
dom (2000) studied the effects of zilpaterol on color and
discoloration of three muscle types during vacuum storage
and subsequent display. The diet supplemented with zilpa-
terol for 30 or 50 days significantly enhanced the color shelf
life of loin and rump steaks and topside mince during retail
display at 4 C, at 0 and 28 days aging. However, the neg-
ative effects of 45 days feeding on tenderness suggest that
30 days of zilpaterol supplementation would be optimum.
Avendano-Reyes et al. (2006) evaluated the effects on
performance, carcass traits, and meat quality of including
60 mg steer 1 d 1 zilpaterol for 33 days before slaughter
of steers fed a high grain diet. Zilpaterol improved
(P < 0.01) gain/feed ratio and ADG by 26% compared to
control steers. In addition, hot carcass weight was
increased (P < 0.05) as well as longissimus area. Meat from
the zilpaterol fed steers had greater (P < 0.01) shear force
values than meat from controls (49.6 vs. 42.6 N) but meat
color was not affected. In contrast, O’Neill (2001) reported
that there were no differences in tenderness between steers
fed zilpaterol and controls. Avendano-Reyes et al. (2006)
compared the effects of ractopamine and zilpaterol and
reported that steers fed ractopamine consumed less feed
than control steers, and tended to have greater fat thickness
than steers fed zilpaterol. Shear values were similar
between steers fed zilpaterol and those fed ractopamine.
Zilpaterol is approved for use in Mexico, South Africa,
and the US (for weight gain, feed efficiency, and carcass
leanness).
Including either ractopamine or zilpaterol in finishing
diets of cattle for a maximum of 30 days before slaughter
would be a very cost effective, convenient, and efficient
way to increase growth rate, improve feed efficiency,
increase dressing percent, and improve carcass meat yield.
However, meat tenderness could be reduced unless proper
electrical stimulation and aging or other mechanical means
are used to counter the potential negative effects on
tenderness.
8. ‘Designer’ lipids
Considerable research has been conducted in recent
years on the effects of dietary conjugated linoleic acid
(CLA) on growth, carcass traits, and meat quality in pigs.
Linoleic acid (C18:2) is at a relatively high concentration in
typical feedstuffs and fat sources used in pig diets. It is not
synthesized by pigs nor is it significantly modified before
being deposited in fat. It contributes to soft, oily fat that
is more susceptible to oxidative rancidity than saturated
fat. Cook et al. (1998) demonstrated a 20% reduction in
backfat for pigs fed CLA and about a 7% increase in lean
muscle mass.
Part of the interest in feeding increased levels of conju-
gated linoleic acid is because of the proposed human health
benefits from consuming CLA. Cook (1999) stated that
CLA is widely recognized as a potent anti-cancer fatty acid
 M.E. Dikeman / Meat Science 77 (2007) 121–135
in many systems and it also reduces fatty streak formation
in the aortas of arteriosclerosis models. The main isomers
of CLA in meat are cis-9, trans-11 and trans-10, cis-12,
and both have human health benefits. Rumen biohydroge-
nation of linoleic acid was originally believed to be the
source of CLA cis-9, trans-11 (Hartfoot & Hazelwood,
1998), but Griinari et al. (2000) and Palmquist (2001) have
shown that the primary source of CLA is through D9-desat-
urase acting on trans vaccenic acid during rumen biohydro-
genation. Scollan et al. (2006) stated that it is recognized
that trans vaccenic acid (trans-11, 18:1) is the precursor
for tissue synthesis of beneficial CLA in both humans
and animals. McGuire, Duckett, Andrae, Giesy, and Hunt
(1998) showed that increasing both forage level and high-
oil corn in feedlot diets of cattle increased CLA content
by 24% in intramuscular fat. Enser et al. (1999) added
60 g linseed oil kg 1 diet dry matter to the diet of finishing
steers and increased CLA content in longissimus muscle
three-fold (35.6 vs. 11.3 mg 100 g 1 muscle) compared to
Megalac supplement (calcium salts of palm oil fatty acids,
mainly palmitic, oleic and linoleic). Lorenzen et al. (2007)
found CLA content to be higher in cooked muscles from
cattle fed on pasture than those fed a feedlot diet and the
CLA content varied among the three muscles studied.
Dugan, Aalhus, Schaefer, and Kramer (1997) found that
CLA in the diet (2%) of pigs repartitions nutrients from car-
cass fat to lean. In a later study, Dugan, Aalhus, Jeremiah,
Kramer, and Schaefer (1999) fed either CLA (2%) or sun-
flower oil (2%) from 61.5 to 106 kg live weight. Feed intake
was reduced, feed efficiency was improved, and growth rate
was not changed. Objective chroma values for subcutane-
ous fat were reduced but longissimus thoracis shear force,
drip loss, and color were not affected by diet for pigs fed
CLA. The longissimus thoracis muscle had higher marbling
scores and increased ether extractable lipid (Table 3). Diet
did not affect any meat palatability trait. In a review of
the literature, Dunshea et al. (2005) stated that the greater
the initial backfat depth, the greater the reduction in back-
Table 3
Effects of diets containing conjugated linoleic acid (CLA) or sunflower oil
on Longissimus objective color, subjective color, structure score, and
marbling score (106-kg pigs)
Parameter Diet
CLA Sunflower
L* 53.2 52.8
Hue 41.0 41.1
Chroma 9.05a 8.21b
Color score 2.96 2.94
Structure score 2.97 2.95
Marbling scorec 434a 390b
Wet Matter Basis (g kg 1)
Intramuscular fat 19.2a 15.5b
Shear force (kg cm2) 5.88 5.95
Drip loss (g kg 1) 50.3 45.1
From Dugan et al. (1999).
a,b
Means with different letters within row are different (P < 0.05).
c
Interaction between diet and gender is significant (P < 0.01).
129
fat in response to dietary CLA supplementation. For the
already extremely lean, hybrid genotypes now used in many
countries, the use of dietary CLA may be limited, unless the
improvements in quality and value of pork with higher
CLA content is rewarded. In a meta-analysis of published
studies, Dunshea et al. (2005) showed an average reduction
in backfat of 6% and an increase in intramuscular fat and
marbling score by 7% and 11%, respectively. However, con-
sumer perception of juiciness was reduced by 12%. Thiel,
Sparks, Wiegand, Parrish, and Ewan (1998) found an
improvement in growth rate as well as in feed efficiency
from feeding 0.12 and 1.0% CLA. An additional advantage
found in that study was increased belly firmness as CLA
was increased linearly in the diet.
Dunshea et al. (2005) summarized the literature to con-
clude that dietary supplementation also results in modified
fatty acid composition of pork tissue, mostly in subcutane-
ous fat. O’Quinn et al. (1999b) studied the effects of mod-
ified tall oil, a rich source of CLA, and vitamin E on
performance and carcass traits of finishing pigs. Pigs fed
modified tall oil had increased ADG and reduced backfat,
regardless of vitamin E level. Woodworth et al. (1999)
found that modified tall oil decreased average daily feed
intake and improved feed efficiency. In addition, pigs fed
modified tall oil (O’Quinn et al., 1999b; Woodworth
et al., 1999) or dietary CLA (Thiel et al., 1998) had firmer
bellies, which would be an advantage to processors.
Joo, Lee, Ha, and Park (2002) fed pigs diets containing
0, 1, 2.5 or 5% CLA for 28 days before slaughter. They
found that dietary CLA reduced the concentration of lino-
leic acid and increased CLA concentration in intramuscu-
lar fat of pork loins. Intramuscular fat was increased by
the 5% CLA in the diet and less purge was observed with
samples from CLA-fed pigs. In addition, dietary CLA
improved the color stability of pork loins during cold stor-
age, likely because of lower TBARS. These authors con-
cluded that dietary CLA offers human health benefits and
it also improves pork color and water-holding capacity.
These authors did not conduct palatability evaluations.
According to Gillis et al. (2004), feeding rumen pro-
tected CLA offers a potential means of increasing the
CLA content of meat from ruminants. In an excellent
review, Scollan et al. (2006) concluded that, despite the
high levels of ruminal biohydrogentation of dietary PUFA,
nutrition is the major route for increasing the content of
beneficial fatty acids in beef. Scollan et al. (2006) stated
that finishing cattle on grass or on concentrates containing
linseed (rich in a-linolenic acid, 18:3n3) increases the con-
tent of 18:3n3 and its longer chain derivative eicosapentae-
noic acid (EPA, 20:5n3) in beef muscle and adipose tissue,
resulting in a lower n6:n3 ratio (Table 4). These authors
further stated that feeding PUFA-rich lipids that are pro-
tected from ruminal biohydrogentation result in further
enhancement of the PUFA in meat with concomitant ben-
eficial improvements in the ratio of PUFA:SFA (P:S ratio)
and n6:n3 ratio (Table 5). However, Scollan et al. (2006)
further stated that as the content of n3 PUFA increases,
130 M.E. Dikeman / Meat Science 77 (2007) 121–135

Table 4
1
Influence of forage on the fatty acid composition (mg 100 g tissue) of beef longissimus muscle
(i) Grass vs. concentrate (Dunshea et al., 2005; adapted from Warren et al. (2003))
Fatty acids Grass Concentrate SED Significance
Total 2581 1724 139.3 ***

18:2n 6 62.0 146.9 6.68 ***

18:3n 3 32.0 7.2 1.60 ***

20:5n 3 17.7 4.5 1.05 ***

22:5n 3 10.8 10.8 1.28 ***

22:6n 3 5.0 1.3 0.30 ***

n 6:n 3 1.2 8.9 0.24 ***

P:S 0.09 0.24 0.010 ***

1
(ii) Proportion of grass (g kg DM) in the diet (Dunshea et al., 2005; adapted from French et al. (2000))
1
Fatty acids Grass (g kg DM) SED Significance
0 510 770 1000
Total 3410 4490 4020 4360 650.5 NS
18:2n 6 120.5 105.8 94.4 85.9 6.05 **(linear)

18:3n 3 29.3 35.4 41.1 46.0 1.78 **(linear)

20:5n 3 4.9 11.0 9.8 9.4 1.32 *(quadratic)

22:6n 3A – – – –
n 6:n 3 4.15 2.86 2.47 2.33 0.197 **(linear)

P:S 0.09 0.10 0.11 0.13 0.010 **(linear)

NS, not significant.

* P < 0.05.
** P < 0.01.
*** P < 0.001.

sensory attributes such as grassy, greasy, and fishy are
increased and color shelf life may be reduced. High levels
of vitamin E are then necessary to help offset these negative
sensory attributes, either in the form of specific grasses or
concentrates supplemented with vitamin E. Scollan et al.
(2006) concluded that opportunities exist to enhance the
content of health promoting fatty acids in beef, which
results in opportunities to add value and product differen-
tiation. However, it is imperative that these approaches to
deliver ‘‘functional’’ attributes do not compromise on the
health value (lipoperoxidation) or the taste of beef.
It should be remembered that the amount of CLA in
meat is small relative to the amount recommended daily
for health benefits, which is approximately 3500 mg d 1
(exraploated from studies with animals) (Ha, Grimm, &
Pariza, 1989). Except for low values found for bulls, the
range of CLA in meat from steers from several published
manuscripts summarized by Scollan et al. (2006) is from
35 to 134 mg 100 g 1 of fresh muscle. Nevertheless, meat
and dairy products are the primary sources of CLA in
human diets.
Although, not all studies show all of the same benefits, the
reported benefits of feeding CLA to pigs include improved
feed efficiency, some reduction in backfat thickness,
increased marbling and ether extractable lipid, increased
fat firmness, improved muscle color, and reduced TBARS.
No detrimental effects on performance or visual meat qual-
ity have been reported. The effects of CLA on meat palat-

Table 5
Influence of fat sources on the fatty acid composition (mg/100 g tissue) of beef longissimus muscle
(ii) Plant oils supplement (PLS) protected from ruminal biohydrogenation (Scollan et al., 2006; adapted from Scollan et al. (2004))
Control PLS (g d 1) SED Significance
400 800 1000
Total 4685 4976 4880 4895 737.0 NS
18:2n 6 120 255 279 305 23.4 ***

18:3n 3 29a 102b 118b,c 139c 13.1 ***

20:5n 3 13a 15a 14a 16a 1.1 *

22:5n 3 23a,b 24b 20a 20a 1.7 *

22:6n 3 1.9 1.8 1.5 1.6 0.272 NS

n 6:n 3 2.27c 2.02b 2.00b 1.88a 0.055 ***

P:S 0.07a 0.18b 0.20b 0.22b 0.018 ***

NS, not significant.

* P < 0.05.
** P < 0.01.
*** P < 0.001.
 M.E. Dikeman / Meat Science 77 (2007) 121–135
ability appear to be minimal. However, at the time this
review was written, the use of CLA per se in diets of pigs
or other meat animals as a metabolic modifier was minimal.
Some hydrogenated vegetable oils have a rather high con-
tent of CLA, but are not marketed as containing CLA. If
the benefit/cost ratio of feeding CLA is proven favorable,
swine producers should consider adding it to pig diets.
9. Chromium, carnatine, magnesium, niacin, manganese and
betaine
Chromium is an essential trace element for normal
metabolism. Boleman et al. (1995) found that feeding ele-
vated levels of chromium picolinate to pigs increased per-
centage of muscle, decreased backfat, and had no effect
on tenderness or sensory traits. Reduced purge loss was
observed in the loin muscle by Mathews, Southern, Bidner,
and Persica (2001).
Carnatine is a vitamin-like compound that aids in the
transport of long-chain fatty acids to the mitochrondial
matrix. Supplementing swine diets with L-carnatine
decreased backfat thickness without affecting growth per-
formance (Owen et al., 1994; Smith et al., 1994) and
increased lean deposition (Owen et al., 1994). O’Quinn
et al. (1999a) evaluated the effects of modified tall oil, chro-
mium nicotinate, and L-carnatine in growing-finishing pig
diets. L-carnatine did not have any affect on growth perfor-
mance or carcass measurements. Chromium nicotinate
improved feed efficiency but had no effects on carcass or
meat traits. Waylan et al. (1999) evaluated meat traits from
the pigs used in the study by O’Quinn et al. (1999a) and
found no differences for longissimus color display, TBARS,
or shear force. Bacon from pigs fed chromium had more
after taste than bacon from pigs not fed chromium. The
results of these studies in which chromium nicotinate and
L-carnatine were included in diets suggest little advantage
from including these metabolic modifiers in the diets of pigs.
Apple, Maxwell, deRodas, Watson, and Johnson (2000)
conducted two experiments on the effects of dietary supple-
mentation of magnesium mica during the growing-finishing
period on pig performance and pork carcass characteris-
tics. Magnesium mica had no effect on performance but
decreased fat thickness and increased muscle percentage
in one study, but color scores improved linearly with
increasing levels of magnesium mica. Magnesium aspartate
supplementation was shown by D’Souza, Warner, Leury,
Table 6
Effect of sex and immunocastration on eating quality of pork loin steaks (D’Souza et al., 1999)
Sex Boar
a
Odor 56
Flavora 58
Tendernessa 52
Juicinessa 60
Overall acceptabilitya 58
a
Acceptability score (line scale) for all attributes, 0 = dislike extremely and 100 = like extremely.
b
Least significant difference.
131
and Dunshea (1998) to increase muscle pH in the longissi-
mus muscle at 40 min and 24 h after slaughter. In addition,
pigs fed the magnesium aspartate supplemented diet had
lower percentages of drip loss, lower surface lightness L*,
and had no PSE carcasses compared to pigs fed the control
diet. However, little is known about the effects of supple-
mental magnesium on palatability of pork. According to
Gardner, Jacob, and Pethick (2001), 1% magnesium oxide
supplementation in sheep 4 days before commercial slaugh-
ter resulted in higher muscle glycogen concentrations in
muscle and reduced ultimate pH in 2 of 3 trials. They con-
cluded that magnesium oxide supplementation is a viable
option for reducing the stress associated with commercial
sheep slaughter.
Real et al. (2002) conducted two experiments to deter-
mine the effects of added dietary niacin on growth perfor-
mance and meat quality in finishing pigs. Dietary
treatments consisted of a corn-soybean meal-based control
diet or the control diet with 13, 28, 55, 110 or 550 mg kg 1
of added niacin. In one experiment, increasing niacin con-
tent improved feed efficiency, subjective color, and pH of
meat, and reduced carcass shrink,
Apple et al. (2007) evaluated the effects of dietary man-
ganese inclusion level on pork quality traits during retail
display. They found that chops from pigs fed 80 ppm man-
ganese received higher (P < 0.05) American and Japanese
color scores than pigs fed 0 or 40 ppm but TBARS values
did not differ among dietary manganese levels.
Betaine is used quite widely in some countries to
improve growth performance of pigs, but little data exists
regarding its effect on pork meat quality.
10. Immunocastration
Dunshea et al. (2005) gave an excellent review of the
mechanisms and effects of immunocastration with Impro-
vacTM on growth performance, carcass composition, and
meat quality of pigs. Improvac contains a modified form
of GnRF in a low reactogenic adjuvant system, which allows
pigs to receive a secondary immunization relatively close to
slaughter. He reviewed original work in which he was
involved (Dunshea et al., 2001) that demonstrated a reduc-
tion in the production and accumulation of both androstre-
none and skatole in pork carcasses. Taint substances already
present are metabolized and this allows boars to be slaugh-
tered at heavier weights without taint, but still allows boars
Barrow Immunocastrate LSDb P-value
62 62 6.13 0.093
62 66 7.01 0.101
59 62 7.44 0.016
59 64 7.05 0.304
62 67 6.41 0.025
132 M.E. Dikeman / Meat Science 77 (2007) 121–135
to benefit from testicular testosterone on growth and carcass
composition before they are immunized.
Dunshea et al. (2005) cited the work of Oliver et al.
(2003) who reported that Improvac-treated boars had up
to 35% increased feed intake from 2 to 4 weeks after the
second immunization, but feed efficiency was not different.
This contrasts a reduction in feed intake of pigs adminis-
tered rpST. These results support real benefits from
immunocastration with Improvac by increasing feed
intake, capitalizing on the growth and leanness advantages
of boars before immunocastration, and may have potential
to improve marbling. In the original work by Dunshea
et al. (2001), most boars treated with either Improvac or
a placebo had appreciable levels of testosterone (>2 nM)
at the time of the second immunization. However, within
2 weeks the placebo treated boars had fat androstenone
levels almost eight times higher than the Improvac treated
boars, which in turn, were not significantly different from
the barrows. In addition, these researchers concluded that
99% and 100% of the Improvac-treated boars had skatole
and androstenone levels below a recommended threshold
of 0.22 and 1.0 lg g 1 established by the European Union.
D’Souza, Hennessy, Danby, McCauley, and Mullan (2000)
reported that immunocastrated boars had higher marbling
scores than non-immunocastrated boars. Dunshea et al.
(2005) cited work from D’Souza, Hagan, Hooper, Nicholls,
and Mullan (1999) showing that flavor, tenderness, and
overall acceptability of pork loin steaks for immunoca-
strates were superior to those from intact boars and not
different from barrows (Table 6). These results suggest that
immunocastration is very effective at preventing boar taint
and it may improve marbling, with meat palatability equal
to meat from barrows.
11. Summary
In a thorough review by Dunshea et al. (2005), the
authors stated that conservative use of each of the meta-
bolic modifiers used to increase carcass leanness will result
in a 5–10% increase in shear force and a similar reduction
in the perception of tenderness. The authors did not, how-
ever, state whether there were any additive negative effects
if two or more metabolic modifiers were used simulta-
neously. Anabolic steroid implants are very cost effective
and improve the efficiency of cattle production. They are
too effective for most of the beef industry not to use them.
In general, the more ‘aggressive’ implants and implant
strategies decrease marbling compared to non-implanted
controls. In addition, aggressive implants or implant strat-
egies may tend to make cattle more susceptible to stress
and increase the incidence of dark cutters when other con-
ditions are unusually stressful. Tenderness and marbling
also usually are reduced in meat from cattle implanted with
aggressive combination implants late in the finishing phase,
or implanted several times. Not following the manufactur-
ers’ recommendations for implanting types and sequences
certainly can cause negative effects.
Feeding vitamin D3 to cattle or pigs will improve tender-
ness early postmortem, but the advantage in tenderness is
minor after adequate aging. The depressions in feed intake
and performance reported in some trials and the concerns
about human toxicity from consuming too much vitamin
D3 likely will prohibit its use until more research is con-
ducted. Including supra-nutritional levels of vitamin E in
the finishing diets of both cattle and pigs appears to be very
beneficial in extending shelf life and reducing oxidative ran-
cidity of meat. The livestock industry should incorporate
vitamin E in all finishing diets, and meat processors and
retailers should reward the industry for this practice. A
low level of vitamin A in cattle diets shows potential to
increase marbling in cattle but may decrease shelf-life of
meat.
Feeding ractopamine hydrochloride (Paylean) to pigs
for 28 days before harvest will increase growth rate, dress-
ing percent, and carcass leanness. It may also improve pro-
cessing yields, and have a neutral effect on meat
palatability. Ractopamine hydrochloride (Optaflexx) will
also improve growth performance of cattle. Zilpaterol (Zil-
max) distinctly improves growth performance, dressing
percent, and carcass muscling. When fed for only 15–30
days and when effective electrical stimulation and adequate
aging are used, its negative effects on meat palatability
should be minor. Meat color may be improved.
Somatatropin is extremely effective in improving growth
performance and meat yield percentage of pigs. However,
it negatively affects marbling, tenderness, and bacon pro-
duction. Including conjugated linoleic acid in diets of pigs
generally has positive effects on carcass composition,
water-holding capacity, and lipid oxidation. In addition,
healthfulness of pork is improved. It should be adapted
by the industry if approved and proven cost effective.
Immunocastration shows very good potential for pre-
venting boar taint and improving marbling and still capi-
talize on the growth, feed efficiency and carcass leanness
of boars up to the point of immmunocastration. Inclusion
of magnesium in the diets of pigs shows potential to
improve pH and color of pork.
References
Aalhus, J. L., Jones, S. D. M., Schaefer, A. L., Tong, A. K. W.,
Rovertson, W. M., Merrill, J. K., et al. (1990). The effect of
ractopamine on performance, carcass composition, and meat quality.
Canadian Journal of Animal Science, 70, 943.
Adachi, K., Kawano, H., Tsuno, K., Nomura, Y., Yamamoto, N.,
Arikawa, A., et al. (1999). Relationship between serum biochemical
values and marbling scores in Japanese Black cattle. Journal of
Veterinary Medical Science, 61(8), 961–964.
Apple, J. K., Maxwell, C. V., deRodas, B., Watson, H. B., & Johnson, Z.
B. (2000). Effect of magnesium mica on performance and carcass
quality of growing-finishing swine. Journal of Animal Science, 78,
2135–2143.
Apple, J. K., Roberts, W. J., Maxwell, C. V., Jr., Rakes, L. K., Friesen, K.
G., & Fakler, T. M. (2007). Influence of dietary inclusion level of
manganese on pork quality during retail display. Meat Science, 75,
640–647.
 M.E. Dikeman / Meat Science 77 (2007) 121–135
Arnett, A., Dikeman, M. E., Spaeth, C., Johnson, B. J., & Hildabrand, B.
(2006). Effects of vitamin A supplementation in young lambs in
performance, serum lipid, and carcass adipose tissue attributes. Journal
of Animal Science, 84(Suppl. Abst.), 243.
Arnold, R. N., Scheller, K. K., Arp, S. C., Williams, S. N., Buege, D. R.,
& Schaefer, D. M. (1992). Effect of long- or short-term feeding of a-
tocopherol acetate to Holstein and crossbred beef steers on perfor-
mance, carcass characteristics, and beef color stability. Journal of
Animal Science, 70, 3055.
Asghar, A., Gray, J. I., Booren, A. M., Gomaa, E. A., Abouzied, M. M.,
Miller, E. R., et al. (1991). Effects of supranutritional dietary vitamin
E levels on subcellular deposition of aa-tocopherol in the muscle and
on pork quality. Journal of the Science of Food and Agriculture, 57, 31.
Avendano-Reyes, L., Torres-Rodriguez, V., Meraz-Murill, F. J., Perez-
Linares, C., Figueroa-Saavedra, F., & Robinson, P. H. (2006). Effects
of two – adrenergic agonists on finishing performance, carcass
characteristics, and meat quality of feedlot steers. Journal of Animal
Science, 80, 3259.
Barham, B. L., Brooks, J. C., Blanton, J. R., Jr., Herring, A. D., Carr, M.
A., Kerth, C. R., et al. (2003). Effects of growth implants on consumer
perceptions of meat tenderness in beef steers. Journal of Animal
Science, 81(12), 3052–3056.
Bedo, G., Santisteban, P., & Aranda, A. (1989). Retinoic acid regulates
growth hormone gene expression. Nature (London), 339, 231–877.
Beermann, D. H. (1994). Carcass composition of animals given partition-
ing agents. In H. D. Hafs & R. G. Zimbleman (Eds.), Low Fat Meats
(pp. 203–222). San Diego: Academy Press, Inc.
Beermann, D. H., Hogue, D. E., Fishell, V. K., Aronica, S., Dickson, H.
W., & Schricker, B. R. (1990). Exogenous human growth hormone
releasing factor and ovine somatotropin improve growth performance
and composition of gain in lambs. Journal of Animal Science, 68, 4122.
Boleman, S. L., Boleman, S. J., Bidner, T. D., Southern, L. L., Ward, T.
L., Pontif, J. E., et al. (1995). Effect of chromium picolinate on
growth, body composition, and tissue accretion in pigs. Journal of
Animal Science, 73, 2033–2042.
Cannon, J. E., Morgan, J. B., McKeith, F. K., Smith, G. C., Sonka, S.,
Heavner, J., et al. (1995). Pork chain quality audit packer survey:
Quantification of pork quality characteristics. Journal of Muscle Foods,
6, 369–402.
Cannon, J. E., Morgan, J. B., Schmidt, G. R., Tatum, J. D., Sofos, J. N.,
Smith, G. C., et al. (1996). Growth and fresh meat quality character-
istic of pigs supplemented with vitamin E. Journal of Animal Science,
74, 98–105.
Cheah, K. S., Cheah, A. M., & Krausgril, D. H. (1995). Effect of dietary
supplementation of vitamin E on pig meat quality. Meat Science, 39,
255–264.
Cook, M. E. (1999). Conjugated linoleic acid. Proceedings from Reciprocal
Meat Conference, 52, 43–46.
Cook, M. E., Jerome, D. L., Crenshaw, T. D., Buege, D. R., Pariza, M.
W., Albright, K. J., et al. (1998). Feeding conjugated linoleic acid
improves feed efficiency and reduces carcass fat in pigs. Federation of
American Society for Experimental Biology Abstract.
Crooker, B. A., McGuire, M. A., Cohick, W. S., Harkins, M., Bauman, D.
E., & Sejrsen, K. (1990). Effect of dose of bovine somatotropin on
nutrient utilization in growing dairy heifers. Journal of Nutrition, 120,
1253–1256.
Dalke, B. S., Roeder, R. A., Kasser, T. R., Veenhuizen, J. J., Hunt, C. W.,
Hinman, D. D., et al. (1992). Dose-response effects of recombinant
bovine somatotropin implants on feedlot performance in steers.
Journal of Animal Science, 70(7), 2130–2137.
Dikeman, M. E. (1991). Growth, carcass composition and meat quality. In
Proceedings of the 37th international congress on meat science and
technology (Vol. 1, pp. 1–38), Kulmbach, Germany.
Dikeman, M. E. (1997). Reducing the fat content by production practices.
Production and processing of healthy meat, poultry and fish products.
In A. M. Pearson & T. R. Dutson (Eds.). Advances in Meat Research
Series (Vol. 11, pp. 150–190). London: Blackie Academic and
Professional, Chapman and Hall.
133
Dikeman, M. E. (2003). Metabolic modifiers and genetics: Effects on
carcass traits and meat quality. Brazilian Journal of Food Techology
(Special issue), 6, 1–38.
D’Souza, D. N., Hagan, C. R., Hooper, J. A., Nicholls, R. R., & Mullan,
B. P. (1999). Influence of genotype and sex on pork eating quality: a
consumer taste panel assessment. In P. D. Cranwell (Ed.), Manipulat-
ing pig production VII (pp. 177). Werribee: Australasian Pig Science
Association.
D’Souza, D. N., Hennessy, D. P., Danby, M., McCauley, I., & Mullan, B.
P. (2000). The effect of Improvac on pork quality. Journal of Animal
Science, 78(Suppl. 1), 158.
D’Souza, D. N., Pethick, D. W., Dunshea, F. R., & Mullan, B. P. (2003).
Nutritional manipulation increases intramuscular fat levels in the
longissimus muscle of female finisher pigs. Australian Journal of
Agricultural Research, 54(8), 745–749.
D’Souza, D. N., Warner, R. D., Leury, B. J., & Dunshea, F. R. (1998).
The effect of dietary magnesium aspartate supplementation on pork
quality. Journal of Animal Science, 76(1), 104–109.
Dugan, M. E. R., Aalhus, J. L., Jeremiah, L. E., Kramer, J. K. G., &
Schaefer, A. L. (1999). The effects of feeding conjugated linoleic acid
on subsequent pork quality. Canadian Journal of Animal Science, 79,
45–51.
Dugan, M. E. R., Aalhus, J. L., Schaefer, A. L., & Kramer, J. K. G.
(1997). The effect of conjugated linoleic acid on fat to lean
repartitioning and feed conversion in pigs. Canadian Journal of Animal
Science, 77, 723–725.
Dunshea, F. R., Colantoni, C., Howard, K., McCaulley, L., Jackson, P.,
Long, K. S., et al. (2001). Vaccination of boars with a GnRH vaccine
(Improvac) eliminates boar taint and increases growth performance.
Journal of Animal Science, 79, 2524–2535.
Dunshea, R. R., D’Souza, D. N., Pethic, D. W., Harper, G. S., & Warner,
R. D. (2005). Effects of dietary factors and other metabolic modifiers
on quality and nutritional value of meat. Meat Science, 71, 8–38.
Ellis, M., & McKeith, F. K. (1999). Non-ruminant nutrition and meat
quality. In Proceedings of the recipocal meat conference (Vol. 52, pp.
15–23).
Enser, M., Scollan, N. D., Choi, N. J., Kurt, E., Hallet, K., & Wood, J. D.
(1999). Effect of dietary lipid on the content of conjugated linoleic acid
in beef muscle. British Journal of Animal Science, 69, 143–146.
Faustman, C., Cassens, R. G., Schaefer, D. M., Buege, D. R., Williams, S.
N., & Scheller, K. K. (1989). Improvement of pigment and lipid
stability in Holstein steer beef by dietary supplementation with vitamin
E. Journal of Food Science, 54, 858.
Fisher, A. V., Wood, J. D., & Whelehan, O. P. (1986). The effects of a
combined androgenic-oestrogenic anabolic agent in steers and bulls. 1.
Growth and carcass composition. Animal Production, 42, 203.
Foutz, C. P., Gill, D. R., Dolezal, H. G., Gardner, T. L., & Botts, R. T.
(1990). Synovex-S and trenbolone acetate implants for feedlot steers.
Journal of Animal Science, 68(Suppl. 1, Abstr.), 472.
French, P. C., Stanton, C., Lawless, F., O’Riordan, G., Monahan, F. J.,
Caffrey, P. J., et al. (2000). Fatty acid composition, including
conjugated linoleic acid, of intramuscular fat from steers offered
grazed grass, grass silage or concentrate-based diets. Journal of Animal
Science, 78, 2849–2855.
Gardner, G. E., Jacob, R. H., & Pethick, D. W. (2001). The effect of
magnesium oxide supplementation on muscle glycogen metabolism
before and after exercise and at slaughter in sheep. Australian Journal
of Agricultural Research, 52, 723–729.
Gillis, M. H., Duckett, S. K., Sackmann, J. R., Realini, C. E., Keisler, D.
H., & Pringle, T. D. (2004). Effects of supplemental rumen-protected
conjugated linoleic acid or linoleic acid on feedlot performance,
carcass quality, and leptin concentrations in beef cattle. Journal of
Animal Science, 82, 851–859.
Gorocica-Buenfil, M. A., Fluharty, F. L., & Loerch, S. C. (2005). Effect of
dietary vitamin A intake on marbling. Journal of Animal Sci-
ence(Suppl. 1, Abstr.), 370.
Griinari, J. P., Corl, B. A., Lacy, S. H., Chouinard, P. Y., Nurmela, K. V.,
& Bauman, D. E. (2000). Conjugated linoleic acid is synthesized
134 M.E. Dikeman / Meat Science 77 (2007) 121–135
endogenously in lactating dairy cows by Delta(9)-desaturase. Journal
of Nutrition, 130, 2285–2291.
Guidera, J., Kerry, J. P., Buckley, D. J., Lynch, P. B., & Morissey, P. A.
(1997). The effect of dietary vitamin E supplementation on the quality
of fresh and frozen lamb. Meat Science, 45, 33–43.
Guo, Q., Richert, B. T., Burgess, J. R., Webel, D. M., Orr, D. E., Blair,
M., et al. (2006a). Effect of dietary vitamin E supplementation and
feeding period on pork quality. Journal of Animal Science, 84(11),
3071–3078.
Guo, Q., Richert, B. T., Burgess, J. R., Webel, D. M., Orr, D. E.,
Blair, M., et al. (2006b). Effects of dietary vitamin E and fat
supplementation on pork quality. Journal of Animal Science, 84(11),
3089–3099.
Ha, Y., Grimm, N., & Pariza, M. (1989). Newly recognized anticarcino-
genic fatty acids: identification and quantification in natural and
processed cheeses. Journal of Agriculture and Food Chemistry, 37,
75–81.
Hartfoot, C. G., & Hazelwood, G. P. (1998). Lipid metabolism in the
rumen. In P. N. Hobson (Ed.), The Rumen Microbial Ecosystem
(pp. 383–426). London: Elsevier Science.
Hayden, J. M., Bergen, W. G., & Merkel, R. A. (1992). Skeletal muscle
protein metabolism and serum growth hormone, insulin, and cortisol
concentrations in growing steers implanted with estradiol-17, trenbo-
lone acetate, or estradiol-17 plus trenbolone acetate. Journal of Animal
Science, 70, 2109.
Hoving-Bolink, A. H., Eikelenboom, G., van Diepen, J. Th. M.,
Jongbloed, A. W., & Houben, J. H. (1998). Effect of dietary vitamin
E supplementation on pork quality. Meat Science, 49, 205–212.
Jensen, C., Guidera, J., Skovgaard, I. M., Staun, H., Skibsted, L. H.,
Jensen, S. K., et al. (1997). Effects of dietary a-tocopherol acetate
supplementation on a-tocopherol deposition in porcine m. psoas major
and m. longissimus dorsi and on drip loss, colour stability, and
oxidative stability of pork meat. Meat Science, 45, 491–500.
Joo, S. T., Lee, J. I., Ha, Y. L., & Park, G. B. (2002). Effects of dietary
conjugated linoleic acid on fatty acid composition, lipid oxidation,
color, and water-holding capacity of pork loin. Journal Animal
Science, 80, 108–112.
Kruk, Z. A., Siebert, B. D., Pitchford, W. S., Davis, J., Harper, G. S., &
Bottema, C. D. K. (2004). The effect of vitamin A on fatness in lot-fed
cattle: (I) Fat Quantity. In Proceedings of the 50th international
congress of meat science and technology, Helsinki, Finland.
Lanari, M. C., Schaefer, D. M., & Scheller, K. K. (1995). Dietary vitamin
E supplementation and discoloration of pork bone and muscle
following modified atmosphere packaging. Meat Science, 41, 237–250.
Lee, C. Y., Lee, H. P., Jeong, J. H., Baik, K. H., Jin, S. K., Lee, J. H.,
et al. (2002). Effects of restricted feeding, low-energy diet, and
implantation of trenbolone acetate plus estradiol on growth, carcass
traits, and circulating concentrations of insulin-like growth factor
(IGF)-I and IGF-binding protein-3 in finishing barrows. Journal of
Animal Science, 80, 84–93.
Liu, Q., Lanari, M. C., & Schaefer, D. M. (1995). A review of dietary
vitamin E supplementation for improvement of beef quality. Journal of
Animal Science, 73, 3131–3140.
Liu, Q., Scheller, K. K., Arp, S. C., Schaefer, D. M., & Frigg, M. (1996).
Color coordinates for color assessment of dietary vitamin E effects on
beef color stability. Journal Animal Science, 74, 106–116.
Lorenzen, C. L., Golden, J. W., Martz, F. A., Grun, I. U., Ellersieck,
M. R., Gerrish, J. R., et al. (2007). Conjugated linoleis acid content
of beef differs by feeding regime and muscle. Meat Science, 76,
159–167.
Lynch, M. P., Kerry, J. P., Buckley, D. J., Faustman, C., & Morrissey, P.
A. (1998). Effect of dietary vitamin E supplementation on the colour
and lipid stability of fresh, frozen and vacuum-packaged beef. Meat
Science, 52, 95–99.
Mathews, J. O., Southern, L. L., Bidner, T. D., & Persica, M. S. (2001).
Effects of betaine, penspace and slaughter handling method on growth
performance, carcass traits and pork quality of finishing barrows.
Journal of Animal Science, 79, 967–974.
McGuire, M. A., Duckett, S. K., Andrae, J. G., Giesy, J. G., & Hunt, C.
W. (1998). Effect of high-oil corn on conjugated linoleic acid (CLA) in
beef. Journal of Animal Science, 76(Suppl. 1, Abstr.), 301.
McKeith, F. K., Lan, Y. H., & Beermann, D. H. (1994). Sensory
characteristics of meat from animals given partitioning agents. In H.
D. Hafs & R. G. Zimbleman (Eds.), Low Fat Meats (pp. 233–252). San
Diego, CA: Academic Press, Inc.
McLaughlin, C. L., Hedrick, H. B., Veenhuizen, J. J., Hintz, R. L.,
Munyakazi, L., Kasser, T. R., et al. (1994). Performance, clinical
chemistry, and carcass responses of finishing lambs to formulated
sometribove (methionyl bovine somatotropin). Journal of Animal
Science, 72, 2544.
Monahan, F. J., Asghar, J. A., Gray, J. I., Buckley, D. J., & Morrissey, P.
A. (1992). Influence of dietary vitamin E (a-tocopherol) on the colour
stability of pork chops. Proceedings of the international congress for
meat science and technology, 38, 543.
Monahan, F. J., Asghar, J. A., Gray, J. I., Buckley, D. J., & Morrissey, P.
A. (1994). Effect of oxidized dietary lipid and viatmin E on the colour
stability of pork chops. Meat Science, 37, 205–215.
Morgan, J. B. (1997). Implant program effects on USDA beef carcass
quality grade traits and meat tenderness. Proceedings of Oklahoma
State University Implant Symposium, Stillwater.
Muramoto, T., Nakanishi, N., Shibata, M., & Aikawa, K. (2003). Effect of
dietary b-carotene supplementation on beef color stability during
display of two muscles from Japanese Black steers. Meat Science, 63,
39–42.
Nade, T., Hirabara, S., Okumura, T., & Fujita, K. (2003). Effects of
vitamin A on carcass composition concerning younger steer fattening
of Wagyu cattle. Asian-Australasian Journal of Animal Science, 16,
353–358.
Nakai, A., Kita, K., Hasegawa, K., & Hiramitsu, M. (1992). Vitamin A
deficiency in Japanese black cattle. Journal of Clinical Veterinary
Medicine, 10, 2143–2148.
Oka, A., Maruo, Y., Miki, T., Yamasaki, T., & Saito, T. (1998). Influence
of vitamin A on the quality of beef from the Tajima strain of Japanese
Black cattle. Meat Science, 48, 159–167.
Oliver, W. T., McCauley, I., Harrell, R. J., Suster, D., Kerton, D. J., &
Dunshea, F. R. (2003). A gonadotropin-releasing factor vaccine
(Improvac) and porcine somatropin have synergistic and additive
effects on growth performance in group-housed boars and gilts.
Journal of Animal Science, 81, 1959–1966.
O’Neill, H. A. (2001). The effect of Zilpaterol hydrochloride on dietary N-
requirements and the quality and nutritional value of meat compo-
nents. M.S. Thesis. Pretoria, South Africa: University of Pretoria.
O’Quinn, P. R., Waylan, A. T., Goodband, R. D., Nelssen, J. L., Unruh,
J. A., Woodworth, J. C., et al. (1999a). Effects of modified tall oil,
chromium nicotinate, and L-carnitine on growth and carcass traits of
finishing pigs. Journal of Animal Science, 77(Suppl. 1, Abstr.), 43.
O’Quinn, P. R., Waylan, A. T., Nelssen, J. L., Goodband, R. D., Unruh,
J. A., Woodworth, J. C., et al. (1999b). The effects of modified tall oil
and vitamin E on growth performance and carcass characteristics of
finishing pigs. Journal of Animal Science, 77(Suppl. 1, Abstr.), 28.
Owens, F. N., & Gardner, B. A. (1999). Ruminant nutrition and meat
quality. Proceedings of the reciprocal meat conference, 52, 25–36.
Owen, K. Q., Smith, J. W., II, Nelssen, J. L., Goodband, R. D., Tokach,
M. D., Friesen, K. G., et al. (1994). The effect of L-carnitine on
growth performance and carcass characteristics of growing-finishing
swine. Journal of Animal Science, 72(Suppl. 1, Abstr.), 274.
Pairault, J., Quignard-Boulange, A., Dugail, I., Lasnier, F., et al. (1988).
Differential effects of retinoic acid upon early and late events in adipose
conversion of 3T3 preadipocytes. Experimental Cell Research, 177,
27–36.
Palmquist, D. L. (2001). Ruminal and endogenous synthesis of CLA in
cows. Australian Journal of Dairy Technology, 56, 134–137.
Preston, R. L. (1999). Hormone containing growth promoting implants in
farmed livestock. Advanced Drug Delivery Reviews, 38, 123.
Puls, R. (1994). Vitamin Levels in Animal Health. Clearbrook, BC,
Canada: Sherpa International, pp. 84.
 M.E. Dikeman / Meat Science 77 (2007) 121–135
Real, D. E., Nelssen, J. L., Unruh, J. A., Tokach, M. D., Goodband, R.
D., Dritz, S. S., et al. (2002). Effects of increasing dietary niacin on
growth performance and meat quality in finishing pigs reared on two
different environments. Journal of Animal Science, 80, 3203–3210.
Roeber, D. L., Cannell, R. C., Belk, K. E., Miller, R. K., Dean, K.,
Tatum, J. D., & Smith, G. C. (1999). Impact of feedlot growth
promotant implant strategies on carcass grade characteristics and
subsequent cooked beef palatability traits when applied to small/
medium framed, 3-way British crossbred steers. Final Report to the
Beef Quality Assurance Advisory Board. Denver, CO: National
Cattlemen’s Beef Association.
Samber, J. A., Tatum, J. D., Wray, M. I., Nichols, W. T., Morgan, J. B., &
Smith, G. C. (1996). Implant program effects on performance and
carcass quality of steer calves finished for 212 days. Journal of Animal
Science, 74, 1470–1476.
Schluter, A. R., Preston, R. L., Davis, G. W., Ramsey, C. B., & Miller, M.
F. (1991). Ractopamine effects on feedlot performance, carcass traits
and chemical composition of feedlot steers and heifers. Proceedings of
the reciprocal meat conference, 44(Abstract), 206.
Scollan, N. D., Enser, M., Richardson, I., Gulati, S., Hallet, K. G., &
Wood, J. D. (2004). The effects of including ruminally protected lipid
in the diet of Charolais steers on animal performance, carcass quality,
and the fatty acid composition of longissimus dorsi muscle. In
Proceedings of the British society of animal science meetings, p. 87.
Scollan, N., Hocquette, J. J. F., Nuernberg, K., Dannenberger, D.,
Richardson, I., & Moloney, A. (2006). Innovations in beef production
systems that enhance the nutritional and health value of beef lipids and
their relationship with meat quality. Meat Science, 74, 17–33.
Seibert, B. D., Pitchford, W. S., Kuchel, H., Kruk, Z. A., & Bottema, C.
D. K. (2000). The effect of beta-carotene on desaturation of ruminant
fat. Asian-Australasian Journal of Animal Science, 13, 185–188.
Smith, K. R., Duckett, S. K., Asain, M. J., Sonon, R. N., Jr., & Pringle, T.
D. (2007). The effect of anabolic implants on intramuscular lipid
deposition in finished beef cattle. Journal of Animal Science, 85,
430–440.
Smith, J. W., II, Owen, K. Q., Nelssen, J. L., Goodband, R. D., Tokach,
M. D., Friesen, K. G., et al. (1994). The effects of dietary carnitine,
betaine, and chromium nicotinate supplementation on growth and
carcass characteristics in growing-finishing pigs. Journal of Animal
Science, 72(Suppl. 1, Abstr.), 1054.
Stites, C. R., McKeith, F. K., Singh, S. D., Bechtel, P. J., Mowrey, D. H.,
& Jones, D. J. (1991). The effect of ractopamine hydrochloride on the
carcass cutting yields of finishing swine. Journal of Animal Science, 69,
3094–3101.
Strydom, P. E., & Nel, E. (1999). The effect of supplementation period of a
beta-agonist (zilpaterol), electrical stimulation, and ageing period on
meat quality characteristics. In Proceedings of the 45th international
congress of meat science and technology (Vol. 2, pp. 58–59). Yoko-
hama, Japan.
Strydom, P. E., Osler, E. H., Nel, E., & Leeuw, K. -J. (1998). The effect of
supplementation period of a beta-agonist (zilpaterol) on growth
135
performance, carcass yield, and meat quality characteristics. In
Proceedings of the 44th international congress of meat science and
technology (pp. 894–895). Barcelona, Spain.
Strydom, P. E., Buys, E. M., & Strydom, H. F. (2000). The effect of a beta-
agonist (zilpaterol) on meat colour shelf-life. In Proceedings of the 46th
international congress of meat science and technology (pp. 148–149).
Buenos Aires, Argentina.
Thiel, L. F. (1991). The effects of exogenous somatotropin administration
on pork yield, composition, quality, and sensory characteristics. PhD.
Dissertation. Ithaca, New York: Cornell University.
Thiel, R. L., Sparks, J. C., Wiegand, B. R., Parrish, F. C., Jr., & Ewan, R.
C. (1998). Conjugated linoleic acid improves performance and body
composition in swine. Journal of Animal Science, 76(Suppl., Abstr.),
61.
Thompson, J., Polkinghorne, R., Porter, M., Hunter, & Burrow. (in
press). The effect of repeated implants of estradiol-17b (Compudose)
on the palatability of steaks from Brahman and Brahman cross steers.
Australian Journal of Experimental Agriculture.
Torri, S., Matsui, T., & Yano, H. (1996). Development of intramuscular
fat in Wagyu beef cattle depends on adipogenic or antiadipogenic
substances present in serum. Journal of Animal Science, 63, 73–78.
Uttaro, B. E., Ball, R. O., Dick, P., Rae, W., Vessie, G., & Jeremiah, L. E.
(1993). Effect of ractopamine and sex on growth, carcass character-
istics, processing yield, and meat quality characteristics of crossbred
swine. Journal of Animal Science, 71, 2439–2449.
Warren, H. E., Enser, M., Richardson, I., Wood, J. D., & Scollan, N. D.
(2003). Effect of breed and diet on total lipid and selected shelf-life
parameters in beef muscle. In Proceedings of British society of animal
science, p. 23.
Watson, R. (submitted for publication). Meta-analysis of studies regard-
ing effects of hormone promotant use on beef palatability measured by
objective testing and by consumer studies. Australian Journal of
Experimental Agriculture.
Waylan, A. T., O’Quinn, P. R., Unruh, J. A., Goodband, R. D., Nelssen,
J. L., Woodworth, J. C., et al. (1999). The effects of swine dietary
supplementation of modified tall oil, chromium nicotinate, and L-
carnitine on longissimus muscle quality characteristics and display
color stability. Journal Animal Science, 77(Suppl. 1, Abstr.), 25.
Williams, S. N, Frye, T. M., Frigg, M., Schaefer, D. M., Scheller, K. K., &
Liu Q. (1992). Vitamin E as an in situ post-mortem pigment and lipid
stabilizer in beef. In Proceedings of the Pacific Northwest animal
nutrition conference (p. 149), Spokane, WA.
Woodworth, J. C., Goodband, R. D., Unruh, J. A., Nelssen, J. L.,
Tokach, M. D., Waylan, A. T., et al. (1999). Interactive effects
between fat source and modified tall oil on growth performance and
carcass characteristics of finishing pigs. Journal of Animal Sci-
ence(Suppl. 1, Abstr.), 118.
Wulf, D. M., Morgan, J. B., Sanders, J. B., Sanders, S. K., Tatum, J. D.,
Smith, G. C., et al. (1995). Effects of dietary supplementation of
vitamin E on storage and caselife properties of lamb retail cuts. Journal
of Animal Science, 73(2), 399–405.
 MEAT
SCIENCE
Meat Science 77 (2007) 136–147
www.elsevier.com/locate/meatsci

Cured meat products without direct addition of nitrate or nitrite:

what are the issues?
a,* b,1
Joseph G. Sebranek , James N. Bacus
a
Department of Animal Science, Department of Food Science and Human Nutrition, 215 Meat Laboratory, Iowa State University,
Ames, IA 50011, United States
b
Technical Ingredient Solutions, LLC, 3518 NW 97th Blvd., Gainesville, FL 32606, United States

Received 2 March 2007; received in revised form 29 March 2007; accepted 30 March 2007

Abstract

The growing popularity of food products marketed in the United States as ‘‘natural’’ and ‘‘organic’’ has resulted in a proliferation of
marketing efforts to meet consumer demands for these foods. Because natural and organic foods are not permitted to use chemical pre-
servatives, the traditional curing agents used for cured meats, nitrate and/or nitrite, cannot be added to natural and organic processed
meat products. However, alternative processes that utilize ingredients with high nitrate content, such as vegetable-based ingredients, and
a nitrate-reducing starter culture can produce processed meats with very typical cured meat properties. Because it is not possible to ana-
lytically measure the amount of nitrite produced by this process, several potential issues deserve consideration. Regulations, for example,
should permit labeling that accurately reflects the process and products, manufacturing procedures must be standardized to achieve
product consistency, marketing efforts should clearly communicate the nature of these products to consumers, product quality must
be maintained, and microbiological safety must be assured.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Organic; Natural; Cured meats; Nitrite; Nitrate

1. Introduction that consumers associate with conventionally produced

In many parts of the world, natural and organic foods
have been experiencing an explosive market growth. Natu-
ral and organic processed meats have been a very signifi-
cant part of that growth, and in the United States, have
made up the fastest growing category of natural and
organic foods (Mitchell, 2006; Organic Trade Association,
2006). Several studies have documented that consumer
preferences for organic and natural foods are based on con-
cerns about antibiotics, pesticides, hormones, genetic mod-
ifications in plants and animals, and chemical additives
*
Corresponding author. Tel.: +1 515 294 1091; fax: +1 515 294 5066.
E-mail addresses: sebranek@iastate.edu (J.G. Sebranek), jbacus@
tingsol.com (J.N. Bacus).
1
Tel.: +1 352 505 5550.
foods (Bourn & Prescott, 2002; Devcich, Pedersen, & Pet-
rie, 2007; Dreezens, Martijn, Tenbult, Kok, & deVries,
2005; Saher, Lindeman, & Hursti, 2006; Siderer, Maquet,
& Anklam, 2005; Winter & Davis, 2006). Consumers are
willing to pay significant premiums for organic and natural
foods. Premiums of 10–40% for organic foods over conven-
tional products are common (Winter & Davis, 2006) but
for meat and poultry, premiums may reach 200% (Bacus,
2006) or even more. In one such example, the average retail
price for four brands of organic broilers in the Midwest
during April and May, 2006 was $3.19/lb. compared to
$1.29/lb. for conventionally produced broilers, a 247% dif-
ference (Husak, 2007). The large premiums that consumers
are willing to pay for natural and organic foods have
resulted in a rapid proliferation of new products and
increased marketing by retailers (Petrak, 2005).

0309-1740/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.meatsci.2007.03.025
 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
In the US, natural and organic foods must be produced
and processed according to United States Department of
Agriculture (USDA) regulations that define these products.
In most cases, natural and organic foods are very similar to
conventional products and do not differ in the typical char-
acteristics expected by consumers. However, in the case of
processed meat products such as hams, bacon, frankfurters,
bologna and others that are typically cured by addition of
sodium/potassium nitrite or nitrate, the requirements for
natural or organic labeling do not permit addition of nitrite
or nitrate. Nitrite, added directly or derived from nitrate, is
a unique, distinctive cured meat ingredient for which there
is no substitute, consequently significant process and prod-
uct changes are necessary to produce natural or organic
processed meats that provide consumers with the properties
expected of traditional cured meat products. These changes,
combined with labeling requirements for these products,
have resulted in a category of processed meats in the US
that is confusing, and perhaps even misleading, to consum-
ers. Further, because of the essential role that nitrite plays in
cured meat quality and safety, changes in the products need
to be carefully examined in light of the processing changes
that are being introduced for manufacturing natural and
organic processed meats. Consequently, several consider-
ations including regulatory, manufacturing, marketing,
quality and safety issues need to be addressed.
2. Background
2.1. Definitions of natural and organic processed meats
The requirements for processed meats such as hams,
bacon, frankfurters and bologna to qualify as natural or
organic in the US have resulted in unique approaches to
the development of these products. This is because, while
‘‘natural’’ and ‘‘organic’’ are two separate and distinct cat-
egories of meat and poultry products in terms of USDA
regulations and labels, neither of these product categories
are permitted to be manufactured with added sodium (or
potassium) nitrite or nitrate. However, the USDA permits
the manufacture of uncured versions of typical cured meats
according to the Code of Federal Regulations (2006), 9
CFR 319.2, which reads:
‘‘Any product, such as frankfurters and corned beef, for
which there is a standard in this part and to which
nitrate or nitrite is permitted or required to be added,
may be prepared without nitrate or nitrite and labeled
with such standard name when immediately preceded
with the term ‘‘Uncured’’ in the same size and style of
lettering as the rest of such standard name: Provided,
That the product is found by the Administrator to be
similar in size, flavor, consistency and general appear-
ance to such products as commonly prepared with
nitrate and nitrite: And providing further, That labeling
for such products complies with the provisions of
317.17 (C) of this subchapter’’.
137
Thus, there is another category of processed meats, sep-
arate from ‘‘natural’’ and ‘‘organic’’, and that category is
‘‘uncured’’. The definitions of natural and organic require
that ‘‘uncured’’ be included on the label for products
labeled with a standardized cured product name (i.e.,
uncured bacon), but it is important to note that not all
products labeled ‘‘uncured’’ are natural or organic.
Processed meats that are labeled ‘‘natural’’ must comply
with the definition of the term provided by the USDA
Food Standards and Labeling Policy Book (USDA,
2005). This definition requires that a natural product . . .
. . .‘‘does not contain any artificial flavor or flavoring,
coloring ingredient, or chemical preservative (as defined
in 21 CFR 101.22), or any other artificial or synthetic
ingredient; and the product and its ingredients are not
more than minimally processed’’.
The term ‘‘minimally processed’’ includes
‘‘. . .traditional processes used to make food edible or to
preserve it or to make it safe for human consumption,
e.g., smoking, roasting, freezing, drying, and ferment-
ing, or those physical processes which do not fundamen-
tally alter the raw product. . ., e.g. grinding meat. . .’’.
(USDA, 2005).
The definition of natural has been controversial. For
example, in the 2005 edition of the USDA Food Standards
and Labeling Policy Book, a note was added indicating
that sugar, sodium lactate (from a corn source) and natural
flavorings from oleoresins or extractives are acceptable for
‘‘all natural’’ claims (USDA, 2005). However, because lac-
tate is widely recognized as an antimicrobial ingredient,
such use may conflict with the ‘‘no chemical preservatives’’
requirement for labeling of a product as natural. This was
the basis for a petition submitted to the USDA in October,
2006 after which the Agency removed lactate from the
guidance statement provided for natural claims. The
USDA will, however, consider use of lactate for natural
foods on a case-by-case basis for applications where the
ingredient may function as a flavoring rather than a preser-
vative. Further, the Agency is currently planning to initiate
new rulemaking processes in the near future for the use of
‘‘natural’’ to clarify these uses as well as the use of natural
claims relative to livestock production practices (O’Con-
nor, 2006).
Products labeled as organic are much better defined and
controlled in the US than the products labeled with natural
claims because organic products are governed by the
USDA Organic Foods Production Act (OFPA), first estab-
lished in 1990 as part of the 1990 Farm Bill (Winter &
Davis, 2006). The OFPA created a National Organic Stan-
dards Board, which established a National List of Allowed
and Prohibited Substances, and developed National
Organic Program Standards. The standards, implemented
in 2002, specify methods, practices and substances that
may be used for production, processing and handling of
organic foods. Meat, for example, must be raised under
138 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
organic management and come from a USDA-certified
farm. Ingredients used for processed products are clearly
defined as permitted or prohibited in the OFPA National
List. Organic products may be labeled as; (1) ‘‘100%
organic’’ which must contain only organically produced
ingredients; (2) ‘‘organic’’ which must contain at least
95% organically produced ingredients; or (3) ‘‘made with
organic ingredients’’ which must have at least 70% organic
ingredients. Products with less than 70% organic ingredi-
ents are not allowed to be labeled as organic and are per-
mitted only to list those ingredients that are organic on
the label. Those products that qualify for the ‘‘organic’’
and ‘‘100% organic’’ labeling are permitted to use the
USDA organic seal as part of the label.
2.2. Definitions of cured and uncured processed meats
The term ‘‘cured’’ relative to processed meats is univer-
sally understood to mean the addition of nitrite or nitrate
with salt and other ingredients to meat for improved preser-
vation (Pegg & Shahidi, 2000). While several ingredients
including sugar, spices, phosphates and others are typically
included in cured meats, it is the addition of nitrite in one
form or another that results in the distinctive characteristics
of cured meat (Cassens, 1990). The typical color, flavor,
shelf life and safety of ham, bacon, frankfurters, bologna
and other cured products are so widely recognized by con-
sumers that these product names are considered ‘‘standard-
ized’’ and ‘‘traditional’’ by the USDA for product labeling
and to not require any further clarification to communicate
the expected product properties to consumers. On the other
hand, products that are similar but made without nitrite or
nitrate, must be clearly labeled as ‘‘uncured’’ as described
earlier. This is because ‘‘uncured’’ versions of standardized
products like ham, bacon, frankfurters and bologna are sig-
nificantly different from the traditional products that they
emulate. At the same time, there are a number of processed
meats that are traditionally manufactured without nitrite or
nitrate, and that are not labeled as uncured because the
standardized product name effectively communicates that
the product is not cured. Fresh sausage, such as pork sau-
sage, for example, is not labeled as ‘‘uncured’’ because these
products are standardized, traditional and the common
name is clearly understood.
The advent of natural and organic processed meat prod-
ucts, both of which prohibit direct addition of nitrite or
nitrate, but that also resemble traditional cured meat has
made it necessary to require ‘‘uncured’’ as part of the tra-
ditional product name. However, because current meat
processing technology has developed means by which
nitrate and nitrite can be indirectly added to these products
to achieve very typical cured meat properties, the labeling
designations for these products as uncured is confusing
and technically inaccurate. Further, because the indirect
addition of nitrate and nitrite to natural and organic pro-
cessed meats has not been thoroughly investigated in terms
of nitrite chemistry and subsequent product properties, a
number of important questions remain to be answered,
particularly in regard to quality and safety.
3. Review of conventional cured meat ingredients and
processes
Conventionally – cured meat products are characterized
by addition of nitrate and/or nitrite. While other ingredi-
ents, particularly sodium chloride, are essential parts of
typical cured meat formulations, it is the nitrate/nitrite that
provides the distinctive properties that are common to all
cured meat products. The role of nitrate/nitrite is so com-
monly understood in the meat industry that the term
‘‘cure’’ is used as both a noun and a verb, meaning either
nitrate/nitrite as chemical entities, or the addition of these
ingredients to meat, respectively.
3.1. Nitrate
Numerous reviews of the history of meat curing have
suggested that meat curing originally developed from use
of salt contaminated with sodium or potassium nitrate
(Binkerd & Kolari, 1975; Cassens, 1990; National Acad-
emy of Sciences, 1982; Pegg & Shahidi, 2000; Pierson &
Smoot, 1982; Sebranek, 1979).
While it is not clear when saltpeter (potassium nitrate)
was first recognized as a curing agent, it is clear that
nitrate, either as saltpeter or as a contaminant of sodium
chloride, was used to cure meat for centuries before
research chemists began to unravel the chemistry of meat
curing. In the late 1800s, it was discovered that nitrate
was converted to nitrite by nitrate-reducing bacteria, and
that nitrite was the true curing agent. The first half of the
20th century brought a gradual shift from nitrate to nitrite
as the primary curing agent for cured meats as the advan-
tages of faster curing time for increased production capac-
ity became more important, and as nitrite chemistry
became better understood. By the early 1970s, relatively lit-
tle nitrate was being used for cured meats (Binkerd &
Kolari, 1975). The late 1960s and early 1970s also brought
a watershed event for the cured meat industry when it
became obvious that nitrite could result in formation of
carcinogenic n-nitrosamines in cured meat. Subsequent
research demonstrated that a significant factor in nitrosa-
mine formation was residual nitrite concentration, and
consequently, nitrate was eliminated from most curing pro-
cesses to achieve better control over residual nitrite concen-
trations (Pegg & Shahidi, 2000). Today, nitrate is rarely
used and then only in a few specialty products such as
dry cured hams and dry sausage where long, slow curing
processes necessitate a long-term reservoir for nitrite that
can be slowly released over the course of the process.
3.2. Nitrite
The chemistry of nitrite in cured meat is an extremely
complex mixture of interactive chemical reactions involv-
 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
ing several different reactants. Nitrite is a highly reactive
compound that can function as an oxidizing, reducing or
a nitrosylating agent, and can be converted to a variety
of related compounds in meat including nitrous acid, nitric
oxide and nitrate (Honikel, 2004). To further complicate
understanding of nitrite chemistry, it has become clear that
the formation of nitric oxide (NO) from nitrite is a neces-
sary prerequisite for most meat curing reactions (Møller
& Skibsted, 2002). Fortunately, fundamental research on
nitric oxide has become one of the most active research
areas in biology because nitric oxide has been found to play
crucial roles in several physiological functions in living
organisms. Fundamental research on nitric oxide in biolog-
ical systems since the early 1990s has facilitated better
understanding of nitrite and nitric oxide in cured meat
(Møller & Skibsted, 2002; Stamler & Meissner, 2001).
The most effective way to consider nitrite chemistry in
cured meat is to consider the practical effects of the addi-
tion of nitrite to meat. The first and most obvious effect
is that of cured color development. Close examination of
the chemical reactions involved with color development
immediately make it obvious that the chemistry of nitrite
in meat is a phenomenally complex event. For example,
nitrite does not act directly as a nitrosylating (transfer of
nitric oxide) agent in meat but first forms intermediates
such as N2O3 (Honikel, 2004)in the mildly acidic condi-
tions typical of postmortem muscle, and NOCl (Fox, Seb-
ranek, & Phillips, 1994; Møller & Skibsted, 2002; Sebranek
& Fox, 1985, 1991) in the presence of salt.
Formation of NO from the intermediates is facilitated
by reductants such as ascorbate, and the NO will react with
the iron of both myoglobin (Fe+2) and metmyoglobin
(Fe+3) to form cured meat pigments and cured color. These
reactions demonstrate two of the most important factors
governing nitrite reactions in conventionally cured meat
products; namely pH and reductants. However, several
other nitrite reactions are involved in cured meats and con-
tribute to nitric oxide production. For example, when
nitrite is added to comminuted meat, the meat quickly
turns brown due to metmyoglobin formation because
nitrite acts as a strong heme pigment oxidant and is, in
turn, reduced to NO. The NO reacts with metmyoglobin
and subsequent reduction reactions convert the oxidized
heme to reduced nitric oxide myoglobin for typical cured
color following cooking. Further, nitrite can also react with
sulfhydryl groups on proteins to release nitric oxide in an
oxidation–reduction reaction that results in a disulfide
(Pegg & Shahidi, 2000).
In addition to the above reactions of nitrite in meat, all
of which affect the rate and/or extent of cured color devel-
opment, nitrite plays a key role in cured meat as a bacterio-
static and bacteriocidal agent. Nitrite is strongly inhibitory
to anaerobic bacteria, most importantly Clostridium botu-
linum and contributes to control of other micro organisms
such as Listeria monocytogenes. The effects of nitrite and
the likely inhibitory mechanism differs in different bacterial
species (Tompkin, 2005). The effectiveness of nitrite as an
139
antibotulinal agent is dependent on several environmental
factors including pH, sodium chloride concentration,
reductants and iron content among others (Tompkin,
2005). While the means by which nitrite achieves microbial
inhibition is not clear and many mechanisms have been
proposed, all of the factors that impact nitrite inhibitory
effects are also important to the known reactions that gen-
erate nitric oxide for cured color. Thus, the reaction
sequences involving nitric oxide are probably an important
part of the antimicrobial role of nitrite in cured meat. For
example, some researchers have suggested that nitrous acid
(HNO2) and/or nitric oxide (NO) may be responsible for
the inhibitory effects of nitrite (Tompkin, 2005). Because
it appears that nitrite reactivity is key to microbial inhibi-
tion (one indicator of this is the strong dependence on
pH), there has been some question whether ingoing or
residual nitrite is most critical to antimicrobial effects.
Tompkin (2005) concluded that residual nitrite at the time
of product temperature abuse is critical to antibotulinal
effects and that depletion of residual nitrite during product
storage will reach some point at which inhibitory effects are
also depleted.
The nitrite reaction sequences involved with cured color
development probably also play a key role in the strong
anti-oxidant function of nitrite in cured meat, because pro-
posed mechanisms for the antioxidant effect of nitrite
include reaction with heme proteins and metals, and for-
mation of nitroso- and nitrosyl-compounds that have anti-
oxidant properties (Pegg & Shahidi, 2000). It is likely that
these proposed mechanisms are dependent upon the same
initial reactions of nitrite that form nitric oxide for cured
color.
Nitrite is also responsible for the production of charac-
teristic cured meat flavor, though this is probably the least
well understood aspect of nitrite chemistry (Pegg & Shah-
idi, 2000). It is easy to distinguish cooked, cured ham from
fresh roast pork on the basis of flavor but the chemical
identity of distinguishing flavor components in cured meat
has eluded numerous researchers. Some of the flavor differ-
ence may be due to the suppression of lipid oxidation by
nitrite but other antioxidants do not produce cured meat
flavor. If nitrite does, in fact, form some volatile flavor fac-
tors, this would represent yet another reaction product of
nitrite in cured meat.
In addition to the nitrite reactions which result in cured
meat color, microbial inhibition, antioxidant effects and
flavor, it has been demonstrated that addition of nitrite
to meat results in formation of nitrate and nitrogen gas
as well as reaction with carbohydrates and lipids (Honikel,
2004; Pegg & Shahidi, 2000).
The point of this brief discussion of nitrite chemistry
and the functions of nitrite in cured meat is to emphasize
that nitrite is a highly reactive ingredient when combined
with meat, and results in a complex mixture of reaction
products. Because nitrite, particularly as nitric oxide, so
readily reacts with a wide variety of substrates, reaction
kinetics could be an important determinant of how nitrite
140 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
is proportioned among the various competitive substrates
and reaction products. A slow formation of nitrite (such
as from nitrate) in meat might be significantly different in
terms of nitrite reaction products than the direct, one-time
addition of a full load of nitrite. If, for example, the fastest-
reacting substrates consumed a greater share of the nitrite
during slow nitrite formation than in the case where nitrite
is added directly, then the end products of the more reac-
tive substrates might achieve greater final concentration.
3.3. Past and current safety issues associated with nitrite
Issues that have been raised concerning the safety of
using nitrate and nitrite for cured meat have included
chemical toxicity, formation of carcinogens in food or after
ingestion, and reproductive and developmental toxicity.
None of these issues represent relevant concerns for nitrate
or nitrite in light at the current regulated levels of use in
processed meats. While nitrite is recognized as a potentially
toxic compound, and there have been cases where nitrite
was mistakenly substituted for other compounds in food
or drink at concentrations great enough to induce toxicity
symptoms, the normally controlled use of nitrite in pro-
cessed meats represents no toxicity risk.
However, the issue of carcinogenic nitrosamines formed
from nitrite in cured meat was a very serious concern in the
1970s. Fortunately, changes in manufacturing practices
and reduced levels of nitrite used in curing solved the prob-
lem of nitrosamine formation in cured meat. Yet, a back-
ground concern about nitrite has lingered, and in the
1990s, a series of epidemiological studies reported that con-
sumption of cured meat was related to childhood leukemia
and brain cancer (Peters et al., 1994; Preston-Martin &
Lijinsky, 1994; Preston-Martin et al., 1996; Sarasua &
Savitz, 1994). Further, in 1998, nitrite was proposed as a
developmental and reproductive toxicant under Califor-
nia’s Proposition 65 (Safe Drinking Water and Toxic
Enforcement Act). Fortunately, both issues (nitrite as a
carcinogen and as a developmental/reproductive toxicant)
have been largely resolved by subsequent studies and care-
ful scientific review (Milkowski, 2006).
The issue of ingested nitrate and nitrite first arose in the
1970s when it was recognized that carcinogenic nitrosa-
mines could be formed in the stomach following ingestion.
Subsequent work has shown that less than 5% of the nitrite
and nitrate typically ingested comes from cured meat, the
rest coming from vegetables and saliva (Archer, 2002; Cas-
sens, 1997a; Milkowski, 2006). Nevertheless, in 2006, the
International Agency for Research on Cancer (IARC) con-
cluded that ‘‘Ingested nitrate or nitrite under conditions
that result in endogenous nitrosation is probably carcino-
genic to humans’’ (Coughlin, 2006). While the IARC
report is still in progress, the conclusions are likely to ramp
up questions and concerns about nitrite as a food additive.
In light of the anticipated challenges to nitrite in cured
meat, it is imperative that as much information as possible
is developed for all processed meat applications where
nitrite and/or nitrate have a role.
3.4. Current US regulations on nitrite and nitrate
Current regulations on use of nitrite and nitrate in the
United States vary depending on the method of curing used
and the product that is cured. For comminuted products,
the maximum ingoing concentration of sodium or potas-
sium nitrite is 156 parts per million (ppm) or 0.25 oz. per
100 lbs. (7 g/45.4 kg), based on the green weight of the
meat block (USDA, 1995). Maximum ingoing nitrate for
these products is 1718 ppm. Sodium and potassium nitrite
and nitrate are limited to the same amount despite the
greater molecular weight of the potassium salts, which
means that less nitrite or nitrate will be included when
the potassium salt is used. For immersion cured, and mas-
saged or pumped products, maximum ingoing sodium or
potassium nitrite and nitrate concentrations are 200 ppm
and 700 ppm, respectively, again based on the green weight
of the meat block. Dry cured products are limited to
625 ppm and 2187 ppm of nitrite and nitrate respectively.
If nitrite and nitrate are both used for a single product,
the ingoing limits remain the same for each but the combi-
nation must not result in more than 200 ppm of analyti-
cally measured nitrite, calculated as sodium nitrite in the
finished product.
Bacon is an exception to the general limits for curing
agents because of the potential for nitrosomine formation.
For pumped and/or massaged bacon without the skin,
120 ppm of sodium nitrite or 148 ppm of potassium nitrite
is required along with 550 ppm of sodium ascorbate or
sodium erythorbate which is also required. It is important
to note that this is a specifically required amount whereas
other nitrite limits are maximum amounts. To accommo-
date variation in pumping procedures and brine drainage
from pumped products, the regulations for pumped and/
or massaged bacon permit ±20% of the target concentra-
tions at the time of injecting or massaging. For example,
sodium nitrite concentrations within the range of 96–
144 ppm are acceptable. Nitrate is not permitted for any
bacon curing method. There are two exceptions to these
regulations for pumped and/or massaged bacon: first,
100 ppm of sodium nitrite (or 123 ppm of potassium
nitrite) with an ‘‘appropriate partial quality control pro-
gram’’ is permitted and, second, 40–80 ppm of sodium
nitrite or 49–99 ppm of potassium nitrite is permitted if
sugar and a lactic acid starter culture are included. Immer-
sion cured bacon is limited to 120 ppm of sodium nitrite or
148 ppm of potassium nitrite while dry cured bacon is lim-
ited to 200 ppm or 246 ppm, respectively. For bellies cured
with the skin on, nitrite and reductant concentrations must
be reduced by 10%, based on the assumption that skin
comprises approximately 10% of the belly weight.
It is important to note that the regulations also require a
minimum of 120 ppm of ingoing nitrite for all cured ‘‘Keep
Refrigerated’’ products ‘‘unless safety is assured by some
 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
other preservation process, such as thermal processing, pH
or moisture control’’. The establishment of minimum ingo-
ing nitrite concentration is considered critical to subse-
quent product safety. This is a significant consideration
for natural and organic cured meat products.
On the other hand, for cured products that are pro-
cessed to ensure shelf stability (stored at ambient tempera-
ture and do not require refrigeration), there is no minimum
ingoing nitrite level. The USDA Processing Inspector’s
Calculations Handbook (USDA, 1995) suggests that, for
shelf-stable products, ‘‘. . .40 ppm nitrite is useful in that
it has some preservative effect. This amount has also been
shown to be sufficient for color-fixing purposes. . .’’.
4. Ingredients used for natural and organic cured meats
Because of the negative perceptions of nitrite-cured meat
held by some consumers, the ‘‘uncured’’ natural and
organic versions of typical cured meats have enjoyed
wide-spread market acceptance. A survey of 56 commercial
‘‘uncured’’ meat products including bacon, ham, frankfurt-
ers, bologna, braunschweiger, salami, Polish sausage,
Andouille sausage and snack sticks showed that most of
these products demonstrated typical cured meat color
and appearance (Sindelar, 2006b). Review of the product
ingredient statements showed that sea salt, evaporated cane
juice, raw sugar or turbinado sugar, lactic acid starter cul-
ture, natural spices or natural flavorings, and celery juice or
celery juice concentrate were used in many of the products.
Analyses of samples of four selected commercial brands
each of natural or organic bacon, hams and frankfurters
showed that all samples except one sample of bacon con-
tained residual nitrite at concentrations ranging from
0.9 ppm to 9.2 ppm. Residual nitrate was found in all prod-
ucts at concentrations of 6.8 ppm to 44.4 ppm (Sindelar,
2006a). Residual nitrite was lower in most of the natural
or organic products at the time of sampling than in compa-
rable commercial products made with conventional addi-
tion of nitrite. Other cured meat properties including
instrumental color, cured pigment concentration, lipid oxi-
dation and sensory properties were, in general, similar for
the natural or organic products relative to the convention-
ally cured products but greater variation in the natural and
organic products was obvious. Most notable was low color
values, low cured pigment content and low sensory scores
for those products that contained little or no residual
nitrite. It is important to note that because these were com-
mercial products selected at retail, the time of manufacture
and storage history of each was unknown. Nevertheless,
these results suggest that: (1) there is wide variation among
the natural and organic processed meats that simulate con-
ventionally cured products; and (2) a large majority of nat-
ural and organic processed meats demonstrate typical
cured meat properties, including cured color, flavor and
significant concentrations of residual nitrite and nitrate.
Thus, it is clear that nitrite and nitrate are being introduced
141
to these products indirectly as components of other
ingredients.
4.1. Unique ingredients in natural and organic processed
meats
The most common ingredient observed in review of the
product labels of natural and organic processed meats was
sea salt. Sea salt is derived directly from evaporation of sea
water, unrefined without addition of free-flow additives
and retains the natural trace minerals characteristic of the
source (Heinerman & Anderson, 2001; Kuhnlein, 1980).
Several varieties of sea salt are available and differ depend-
ing on the geographical origin of the water used and the
mineral content (Saltworks, 2006). While sea salt has been
suggested as a likely source of nitrate, limited analytical
information suggested that nitrate content of sea salt is rel-
atively low. Herrador, Sayago, Rosales, and Asuero (2005)
reported that Mediterranean sea salt contained 1.1 ppm of
nitrate and 1.2 ppm of nitrite. Cantoni, Berretta, and Bian-
chi (1978) analyzed 10 samples each of 3 grades of sea salt
and found nitrate and nitrite concentrations of 0.3–
1.7 ppm and 0–0.45 ppm, respectively.
The second most common ingredient observed in natu-
ral and organic processed meat ingredient lists was raw
sugar, most often shown as turbinado sugar. Turbinado
sugar is a raw sugar obtained from evaporation of sugar
cane juice followed by centrifugation to remove surface
molasses. Remaining molasses gives turbinado sugar a
light brown color and flavor similar to brown sugar. While
it seems possible that raw sugar could include nitrate, there
appears to be no evidence of significant nitrate or nitrite
concentrations in raw sugar.
Natural flavorings or spices, and celery juice or celery
juice concentrate were frequently listed as ingredients,
and because these are plant/vegetable products, the poten-
tial contribution of nitrate from these sources is very signif-
icant. Vegetables are well-known as a source of nitrate with
concentrations as high as 1500–2800 ppm (National Acad-
emy of Sciences, 1981) in celery, lettuce and beets. Vegeta-
ble juices and vegetable powders are commercially
available and may be used as ingredients in natural and
organic foods. Analysis of some commercially available
vegetable juices showed that carrot, celery, beet and spin-
ach juice contained 171 ppm, 2114 ppm, 2273 ppm and
3227 ppm of nitrate respectively (Sebranek, 2006). After
10 days of storage at room temperature, nitrate levels in
these juices declined by 14–22%. Nitrite was not detected
initially but concentrations of 128–189 ppm of nitrite were
found after 10 days at room temperature, probably result-
ing from bacterial reduction of nitrate. Analysis of com-
mercial celery juice powder indicated a nitrate content of
27,462 ppm or about 2.75%, reflecting the increased con-
centration following drying (Sindelar, 2006a). Clearly, veg-
etable products offer the greatest potential to introduce
natural sources of nitrate into processed meats. Juices
and powders have advantages in supplying nitrate in
142 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
concentrated form. Celery juice and celery powder appear
to be highly compatible with processed meat products
because of very little vegetable pigment (as opposed to
beets, for example) and a mild flavor profile similar to
raw celery that does not detract greatly from finished prod-
uct flavor.
A critical ingredient for processed meats with natural
nitrate sources is a nitrate-reducing bacterial culture if typ-
ical cured meat properties are the final objective. The neces-
sity of bacterial reduction of nitrate to nitrite for meat
curing has long been recognized, and nitrate-reducing cul-
tures have been commercially available for several years.
Most applications of these cultures have been for dry sau-
sage where a long-term reservoir of nitrite during drying is
desirable and where subtle flavor contributions from the
culture are considered important (Olesen, Meyer, & Sta-
hnke, 2004). The lactic acid starter cultures used for fer-
mented sausage, primarily Lactobacillus plantarum and
Pediococcus acidilactici, do not reduce nitrate. However,
cultures of coagulase-negative cocci such as Kocuria (for-
merly Micrococcus) varians, Staphylococcus xylosus, Staph-
ylococcus carnosus and others will reduce nitrate to nitrite.
These organisms can achieve nitrate reduction at 15–20 °C
but are much more effective at temperatures over 30 °C
(Casaburi, Blaiotta, Mauriello, Pepe, & Villani, 2005).
The typical recommended holding temperature for com-
mercial nitrate-reducing cultures is 38–42 °C to minimize
the time necessary for adequate nitrite formation. Recent
research has documented that time is a critical parameter
in the development of typical cured meat properties from
natural sources of nitrate. Sindelar (2006a) reported that
holding time at 38 °C was more critical than the amount
of naturally-added nitrate for development of cured meat
properties in frankfurters and hams. Time appeared to be
more critical for the small diameter frankfurters that
reached internal temperature of 38 °C quickly than for
the large diameter hams where internal temperature
increased to 38 °C more slowly.
Sindelar, Cordray, Sebranek, Love, and Ahn (in press a)
evaluated several quality characteristics of frankfurter-style
cooked sausages manufactured with starter culture and
either 0.2% or 0.4% celery juice powder, each held at
38 °C for either 30 min or 120 min. The products were eval-
uated during 90 days of refrigerated, vacuum-packaged
storage and compared with conventionally processed prod-
ucts manufactured at the same time with added sodium
nitrite. Color measurements (Hunter a+ values reflectance
ratios, cured pigment concentrations) indicated that treat-
ments with short incubation time resulted in less cured
color/redness than the nitrite-cured control though this dif-
ference was not always significant. Cured color/redness was
comparable to the nitrite-cured control with the longer
incubation time for the first 14 days following manufactur-
ing but the difference became non-significant during
extended storage. Residual nitrite following incubation
was dramatically different with 5.6 ppm and 7.7 ppm found
for the 0.2% and 0.4% celery powder levels, respectively,
after 30 min but 24.5 ppm and 46.0 ppm observed after
120 min. No differences were noted for lipid oxidation
between any of the treatments and the control. The
nitrite-cured control received the highest sensory scores
though differences were not significant for all sensory
properties.
A similar experiment with hams (Sindelar, Cordray,
Sebranek, Love, & Ahn, in press b) was conducted using
either 0.2% or 0.35% celery powder and incubation time
of 0 min or 120 min. The treatment with no incubation
time was included because the extended thermal process
(3 hrs., 35 min.) used for hams relative to small diameter
frankfurter-style sausage was expected to result in adequate
nitrate reduction by the culture. Results showed that there
were no treatment differences in objective color measure-
ments or cured pigment concentrations and all products
were similar in color properties to the nitrite-cured control.
Residual nitrite, following the 120 minute incubation for
the 0.2% and 0.35% celery juice powder additions, was
19.5 ppm and 36.1 ppm, respectively. The residual nitrite
was significantly less for the hams with celery juice powder
(21.0–36.0 ppm at day 0; 7.2–21.3 ppm after 90 days) rela-
tive to the nitrite-cured control (63.4 ppm at day 0: 34.1
ppm after 90 days). However, residual nitrite was greater
in hams with a greater amount of added celery juice
(27.7–36.0 ppm from 0.35% celery powder vs. 19.3–
21.0 ppm from 0.20% celery powder at day 0 compared
with 11.7–21.3 ppm vs. 7.2–8.8 ppm, respectively, for each
after 90 days). Sensory panel evaluation indicated that
the greater celery powder treatment (0.35%) resulted in
greater vegetable aroma and flavor with less ham aroma
and flavor. The treatments with a low level of celery pow-
der (0.2%) were similar to the nitrite-cured control for all
sensory properties evaluated.
The authors concluded that the celery juice powder/star-
ter culture treatment was an effective alternative to the
direct addition of sodium nitrite to small-diameter, frank-
furter-style cured sausage but that incubation time at
38 °C is an important factor for product quality. The celery
juice powder/starter culture treatment was also effective for
hams but in this case the amount of celery juice powder
proved to be more critical. For large diameter products
such as hams, it appears that the slow temperature increase
that is part of a typical thermal process may provide
enough time for the culture to achieve nitrate-to-nitrite
reduction. Further, the delicate flavor profile of hams
makes these products more susceptible to flavor contrib-
uted by vegetable products.
The authors also pointed out that the concentration of
celery juice powder used (0.2%, 0.35% and 0.4% – total
formulation weight basis) could provide, with 100%
nitrate-to-nitrite conversion, maximum ingoing nitrite con-
centrations of 69 ppm, 120 ppm and 139 ppm (meat block
basis), respectively, based on the initial nitrate concentra-
tion of 27,462 ppm in the celery powder. Because these
nitrite concentrations are, at best, significantly less than
the 156–200 ppm normally included in cured comminuted
 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
products or injected products, it seems likely that product
quality differences could occur in some circumstances. It
is also worth noting that the actual amounts of nitrite
formed from nitrate when natural nitrate sources are used
could be a concern relative to microbiological safety. The
shelf life of processed meats manufactured with natural
nitrate sources in generally less than nitrite-cured products
because less nitrite is present and other typical preserva-
tives such as phosphates, lactate, curing accelerators and
antioxidants are not included (Bacus, 2006).
Ingredients that might be considered as curing adjuncts
for natural or organic processed meats include vinegar,
lemon juice solids, and cherry powder. Acidulants such
as vinegar have potential to accelerate nitrite reactions
because of the impact of pH. However, reducing pH in
these products is also a concern for reduced moisture reten-
tion because phosphates and many of the traditional water
binders cannot be used for natural or organic products.
Lemon juice solids or powder are typically significant
sources of citric acid which could have similar pH effects
as vinegar. Cherry powder, on the other hand, is high in
ascorbic acid which functions as a strong nitrite reductant
but does not have a large impact on product pH.
An evaluation of a cured pork product manufactured
with a natural nitrate source (celery powder) and with or
without 0.28% cherry powder showed that including the
cherry powder reduced residual nitrite by about 50% (Bas-
eler, 2007). Residual nitrite declined from 61 ppm to
32 ppm during 12 weeks of storage for a nitrite-cured con-
trol, 18–10 ppm for the celery powder treatment and 10–
3 ppm for the celery powder/cherry powder treatment.
Addition of cherry powder did not alter the product pH.
Other product properties (color, lipid oxidation) were not
consistently different though the nitrite-cured treatment
showed greater redness (Hunter a* values) after about 4
weeks of storage.
Natural antioxidants such as rosemary may be used to
provide flavor protection and to retard lipid oxidation in
processed meats. However, these compounds do not con-
tribute directly to nitrate/nitrite reactions in meat systems.
Further, it appears that the nitrite generated from natural
nitrate sources as reported by Sindelar et al. (in press a,
in press b) was sufficient to provide strong antioxidant
effects as typically observed in nitrite-cured meats. Past
research has shown that as little as 50 ppm added nitrite
has a highly significant effect on lipid oxidation (Morrissey
& Techivangana, 1985). Thus, relatively small amounts of
nitrite formed from nitrate probably provide an important
antioxidant role in natural and organic processed meats.
4.2. Processes for naturally-cured meats
Most processors that utilize ‘‘natural curing’’ are follow-
ing processing procedures that are, in general, similar to
those processes that include chemical nitrites and nitrates.
Naturally-cured products typically utilize natural sources
of nitrate, but some natural ingredients may also contain
143
nitrite. If sufficient nitrite is consistently available from a
natural source, no changes in the normal process are
required.
Naturally-cured meat products that utilize natural ingre-
dients for a nitrate source need an ingredient that contains
a relatively high natural nitrate content. The nitrate ion is
much more available, more consistent in concentration
and more stable than nitrite, and can be found in a wide
variety of natural ingredients. When using a natural nitrate
source, conversion of the nitrate to nitrite is required. Typ-
ically, this conversion is accomplished by specific microor-
ganisms (with a nitrate reductase enzyme), as described
earlier, that are also acceptable food ingredients. When
using these microorganisms, the conversion process
requires time, with the specific amount of time depending
upon the temperature, the environment, and the concentra-
tion of the reactants, namely the microorganisms and the
naturally-occurring nitrate. The conversion time can be
decreased by increasing the reactant concentrations with
the amount of starter culture the most critical.
In all natural curing processes, good distribution of both
the nitrate source and the starter culture is essential to
achieve uniform curing. The natural nitrate source, if dry,
is usually either blended with the dry seasoning component
for comminuted products, or added directly to curing
brines. The starter culture commonly is diluted first with
good quality water (i.e. distilled, or low in chlorine or other
bacteriocidal chemicals) prior to the addition to commi-
nuted products (the USDA permits a maximum 0.5% com-
bined water and starter culture without labeling the added
water) or may also be added directly to curing brines. Also,
it is recommended that the starter culture should not be
pre-blended with anything that might affect viability (i.e.
spices, salt), and thus the nitrate-reducing activity of the
culture. The naturally-occurring nitrate is soluble, but the
starter culture is not soluble, being water-dispersible, there-
fore some agitation is recommended for brines to achieve
optimal distribution of the culture in the meat product.
With curing brines, the pH of the brine is critical to achiev-
ing optimal natural curing as well as final product texture,
because the phosphates or other buffering agents typically
used with nitrite-added products cannot be included for
products labeled natural or organic. Generally, low pH
brines (i.e. <5.5) are not desirable, thus the pH effect of
any added natural ingredients should be considered. With
comminuted meat products, the pH effect of directly-added
ingredients is not as critical due to the buffering capacity of
the meat. Liquid sources of naturally-occurring nitrates
(vegetable juices) also are utilized but these ingredients
pose some manufacturing issues. Typically, most of these
liquids are not shelf stable, and are supplied in frozen form.
Secondly, the added water that is a component of the juices
must be considered.
For small diameter cooked sausage, formulation and
processing are essentially the same as for nitrite-added
products, except for the nitrate source and the culture,
and a smokehouse process that includes an ‘‘incubation’’
144 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
period of about 1 h at 42 °C to achieve nitrate reduction
prior to a smoke-cook thermal process. For injected prod-
ucts, the physical injection of the brine with the culture is
critical because the culture will not migrate significantly
from injection sites. Consequently, good distribution by
injection is necessary to avoid uncured spots. Increased
injection percentage of brine is generally preferred to facil-
itate better distribution. For thermal processing, no ‘‘incu-
bation’’ period is likely to be necessary for large diameter
products due to the longer process time involved with rel-
atively slow increase in internal temperature. As product
diameter is decreased, the heat process may need to be
adjusted to achieve optimal nitrate conversion. With
injected bellies, for example, a short ‘‘incubation’’ period
at 42–46 °C may be necessary prior to the usual heating
cycle. Because bacon is not fully cooked, relatively high
bacterial counts from the added culture will remain in the
finished product. Fermented sausage products do not typ-
ically require any adjustments in processing because the
fermentation step already incorporated into the process
provides adequate nitrate conversion. Because most starter
cultures, particularly in the US consist of lactic acid-
producing bacteria only, it is imperative to confirm that
the added starter culture is composed of a mixed culture
including nitrate-reducing organisms as well as those
included to ferment the added sugars. Many mixed starter
cultures are available to accomplish both functions, or the
nitrate-reducing culture can be added ‘‘on top’’ of a cur-
rently used acid-producing culture.
5. Current issues with natural and organic cured meats
5.1. Regulatory
The current US regulatory requirements concerning
‘‘organic’’ meat products are well defined, thus processors
desiring to make such products must adhere to a fixed set
of regulations outlining permitted ingredients. With ‘‘natu-
ral’’ meat products, however, the rules for permitted ingre-
dients recently have become more confusing. The petition
submitted to the USDA in October, 2006, suggested that
the 2005 revisions to the agency’s ‘‘natural’’ policy created
inconsistencies by allowing foods carrying the ‘‘natural’’
label to contain synthetic ingredients and preservatives.
Much of the concern expressed by the petitioner was the
allowance of sodium and potassium lactates in ‘‘natural’’
products, since these ingredients would be considered
‘‘chemical preservatives’’. The petition to the USDA pro-
posed that extensive rulemaking should be initiated for
meat and poultry products labeled as ‘‘natural’’ in much
the same way that had been done with products labeled
as ‘‘organic’’. Currently, the USDA has begun the rule-
making process for new regulations on natural products
with a public meeting in Washington in December, 2006,
and a comment period that concluded in March, 2007.
The issue of lactates as ‘‘chemical preservatives’’ also
raised the issue of other dual-function ingredients, whereas
the ingredient may be considered as a natural ingredient for
flavor and/or function, but can also have a dual function as
a ‘‘natural’’ preservative. The issue of ‘‘natural preserva-
tive’’ vs. ‘‘chemical preservative’’ has not been defined, as
yet. Many natural compounds that exist in the environ-
ment can serve to inhibit microorganisms, retard oxidation,
and thus ‘‘preserve’’ the product and would be valuable
ingredients in food products that are labeled as ‘‘natural’’.
Until this issue of ‘‘natural’’ vs. ‘‘chemical’’ preservatives is
resolved, the current US regulatory environment is retard-
ing innovative product development and may compromise
food safety as well.
5.2. Manufacturing
When manufacturing ‘‘natural’’ and ‘‘organic’’ meat
products using natural ingredients, the inherent variability
of natural ingredients must be considered. In the natural
curing process, whereby naturally-occurring nitrates are
converted by starter cultures to nitrites, the concentration
of the nitrate in the source plus the nitrate-reducing activity
of the starter culture will effect the degree of curing achieved.
Typically, in products cured with direct addition of sodium
nitrite in the US, the ingoing nitrite is limited to 156 ppm in
most meat products and 120 ppm in injected bacon. In nat-
urally-cured meat products, the ingoing nitrate level is most
often between 40 ppm and 60 ppm, thus there is, at best, sig-
nificantly less nitrite in these products than in the typical
nitrite-cured products. Measurable residual nitrite concen-
trations in both types of cured products are often similar
without major differences in color or color stability. How-
ever, product quality will be very dependent upon actual
nitrite formation in the product during the nitrate reduction
phase of the process. While reduced shelf-life has been
observed in naturally-cured meats, a significant part of the
change is probably due to the lack of other traditional ingre-
dients that are not permitted in ‘‘natural’’ products (i.e.
phosphate, ascorbate, erythorbate, citric acid, synthetic
antioxidants) as well as reduced concentration of nitrite.
5.3. Marketing
The issue of consumer understanding of what is meant
by ‘‘natural’’ meat products is difficult to define. Many
consumers may not comprehend that natural ingredients
often contain naturally-occurring chemicals virtually iden-
tical in chemical nature to those chemicals synthetically
produced. One of the current concerns with ‘‘naturally-
cured’’ meat and poultry products is that these products
often contain residual nitrate and nitrite, even though
labeled as ‘‘no nitrates or nitrites added’’. According to
the US Code of Federal Regulations (2006) in 9 CFR
319.2, the processor has no choice but to label such prod-
ucts (i.e. ‘‘. . . to which nitrite or nitrate is permitted or
required to be added. . .’’) as ‘‘uncured’’ and no ‘‘nitrates
or nitrites added’’ even though the processor may be utiliz-
ing a natural curing process.
 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
To provide the consumer with the most accurate infor-
mation, more appropriate labeling would be to use a term
such as ‘‘naturally-cured’’ and eliminate the ‘‘uncured’’ and
‘‘no nitrates or nitrites added’’ requirement that currently
exists in the US.
5.4. Quality
The quality characteristics expected of traditional cured
meats that are unique to these products include the red-
dish-pink color of cooked denatured nitrosylhemochrome,
a flavor that is distinct from uncured products, and long-
term flavor protection resulting from the strong antioxi-
dant effect of nitrite on meat systems. The fixation of
desirable color is the first and most obvious effect of nitrite
when added to meat and is considered an essential function
because color is a critical component affecting consumer
retail purchases (Cornforth & Jayasingh, 2004). As little
as 2–14 ppm of nitrite (depending on species) can induce
pink coloration in cooked meats, though at these levels
the color is often sporadic and likely to fade with time.
Extensive research in the 1970s showed that 25–50 ppm
of ingoing nitrite was adequate to develop relatively stable
cured color (National Academy of Sciences, 1982). While
there are indications that cured color may be less intense
with 40–50 ppm of nitrite instead of 150–200 ppm depend-
ing on product type, 40–50 ppm is generally considered
adequate for cured color development in most products.
Thus, it would appear that cured color development can
be achieved relatively easily in processed meat using natu-
ral sources of nitrate and a nitrate-reducing culture. A
related question, however, concerns the long-term stability
of the cured color formed in these products. One of the dif-
ficulties with assessing potential cured color intensity or
stability with nitrate-based cures is that the absolute
amount of nitrite formed from nitrate cannot be deter-
mined due to the reactive nature of nitrite in meat. Sindelar
et al. (in press a) reported that for frankfurter-type sau-
sages, products made with celery powder and culture had
9.3–31.9 ppm of residual nitrate remaining when 69 ppm
of nitrate was added as part of the celery powder, and
12.2–81.4 ppm when 139 ppm was added. So, if 100% of
the nitrate that was depleted was in fact reduced to nitrite,
the ingoing nitrite concentrations ranged from 37 ppm to
127 ppm. This is sufficient nitrite to generate desirable
cured meat color characteristics in most processed meat
products. Similar results were observed for color with hams
(Sindelar et al., in press b) where the residual nitrate con-
centrations suggested formation of nitrite in the range of
45–119 ppm. Thus, the quality of cured color in terms of
intensity and stability is not likely to be a major issue in
processed meats using natural sources of nitrate if appro-
priate processing procedures are followed to achieve nitrate
reduction, and if adequate packaging (oxygen removal by
vacuum or gas flushing and high oxygen-barrier films) is
used (Møller et al., 2003).
145
Cured flavor is an important quality attribute of cured
meats that is derived from addition of nitrite, though the
chemical nature of the flavor has never been established.
It is clear, however, that relatively low concentrations of
nitrite result in significant cured flavor. Several researchers
have reported acceptable cured meat flavor in products for-
mulated with 40 ppm of ingoing nitrite (Pegg & Shahidi,
2000). In a series of reports, MacDonald, Gray, Stanley,
and Usborne (1980), MacDonald, Gray, Kakuda, and
Lee (1980) and MacDonald, Gray, and Gibbins (1980)
concluded that addition of 50 ppm of nitrite to hams was
sufficient to produce significant cured meat flavor and anti-
oxidant protection. Thus, in addition to color, it appears
that 40–50 ppm or more of ingoing nitrite will result in a
significant flavor contribution to cured meat.
The third quality contribution of nitrite to cured meat is
the often-overlooked, but highly effective role of nitrite as
an antioxidant, and it is clear that nitrite is again effective
at relatively low concentrations. Morrissey and Techivang-
ana (1985), for example, using cooked, ground beef, pork,
chicken and fish muscle, reported that 50 ppm of nitrite
reduced TBA values by 50–64% for beef, pork and chicken,
and about 35% for fish. Nitrite concentrations of 100 ppm
resulted in TBA reductions of 57–72%, and 200 ppm
reduced TBA values by 87–91%. There was a very clear
relationship between saturated:unsaturated fat ratios and
the TBA values, with more unsaturated fats resulting in
greater TBA values regardless of the nitrite concentration.
While nitrite is effective as an antioxidant at 50 ppm, it is
more effective at greater concentrations up to 200 ppm. A
point to note is that the antioxidant function of nitrite in
cured meat, while highly effective, is not as unique as the
color and flavor contributions. There are a number of other
antioxidants including natural compounds that can protect
meat lipids from oxidation and flavor deterioration.
If at least 50 ppm of nitrite is formed from nitrate during
processing of meat products with natural nitrate sources, it
appears that the typical quality characteristics expected of
cured meat (color, flavor, flavor stability) will be achieved.
A question that is more difficult to answer is the long-term
stability of those quality characteristics. It is well recog-
nized that when nitrite is fully depleted from cured meat,
color fading and flavor changes typically occur. Some
residual nitrite is essential to maintaining typical cured
meat properties during extended product storage, and 5–
15 ppm residual nitrite has been reported for commercial
cured meats in the US (Cassens, 1997b). It is important
to keep in mind that packaging and environmental condi-
tions, particularly temperature and exposure to light, are
critical to long-term cured meat quality, and become more
critical when residual nitrite is reduced.
5.5. Safety
The safety of processed meats that simulate traditional
cured meats by using natural sources of nitrate is a signif-
icant issue for two reasons; first, nitrite is a very effective
146 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
antimicrobial agent, particularly for preventing toxin pro-
duction by C. botulinum and second; residual nitrite con-
centration is a well-known risk factor in the potential
formation of carcinogenic nitrosamines. In both cases,
ingoing and residual nitrite concentrations must be care-
fully controlled to provide product safety.
The antimicrobial role of nitrite in cured meat has been
well documented. While nitrite plays a key role in inhibi-
tion of C. botulinum, extensive research has demonstrated
that pH, reductants (ascorbate and erythorbate), salt,
phosphates and thermal processing are all highly interac-
tive with nitrite for achieving inhibitory effects (Tompkin,
2005). It has been suggested that both the amount of added
nitrite and the amount of nitrite present as residual nitrite
at the time of temperature abuse are important to antibo-
tulinal protection. Because ingoing nitrite is depleted over
time in cured meat, the importance of ingoing nitrite is
probably due to the increased residual nitrite that results.
Christiansen (1980), in a review of botulinal inhibition by
nitrite, concluded ‘‘that any change in nitrite usage which
reduces the level of residual nitrite or increases the rate
of nitrite depletion could increase the above mentioned
(botulinal) theoretical risk’’.
The issue for processed meats that utilize natural sources
of nitrate is that the true amount of nitrite formed is
unknown and impossible to measure because nitrite reacts
quickly with meat components. While the amount of
detectable residual nitrite in these products is significant,
it is often less than that found in nitrite-cured products
(Sindelar et al., in press a, in press b) depending on process-
ing conditions. On the other hand, the nitrite reactions
means that there are variable pools of nitrite-modified
compounds in cured meat that remain available as reactive
sources of nitric oxide (Kanner & Juven, 1980; Møller &
Skibsted, 2002). Consequently, the microbial safety of pro-
cessed meats manufactured with natural sources of nitrate
is very difficult to assess without microbiological challenge
studies. This is a very significant current research need for
effectively assessing the safety of these products.
The second potential safety issue that should be consid-
ered with these products is the implications of higher-than-
usual nitrite concentrations. Sindelar et al. (in press a)
found that a frankfurter-style sausage processed with
0.4% celery powder and an extended incubation time
resulted in significantly more residual nitrite than a
nitrite-cured control throughout 90 days of storage. Ele-
vated residual nitrite in bacon is a potential risk for nitro-
samine formation and actual ingoing nitrite concentrations
need to be carefully controlled to avoid this potential prob-
lem. The nature of the time–temperature relationship for
reduction of nitrate to nitrite by a starter culture makes
the concentration of nitrite a variable entity. Further, veg-
etable products are recognized as extremely variable in
nitrate content as a result of different environmental condi-
tions that occur during plant growth (National Academy of
Sciences, 1981). Consequently, more information is needed
on the best means by which to control nitrite formation in
processed meats manufactured with natural sources of
nitrate to assure that excess concentrations of nitrite do
not become a safety issue.
References
Archer, D. L. (2002). Evidence that ingested nitrate and nitrite are
beneficial to health. Journal of Food Protection, 65, 872–875.
Bacus, J. (2006). Natural ingredients for cured and smoked meats. In
Proceedings of the 59th reciprocal meat conference (pp. 77–78), June
18–21, 2006. Champaign-Urbana, IL, USA.
Baseler, L. J. (2007) Unpublished data. Iowa State University, Ames, IA,
USA.
Binkerd, E. F., & Kolari, O. E. (1975). The history and use of nitrate and
nitrite in the curing of meat. Food and Cosmetics Toxicology, 13,
655–661.
Bourn, D., & Prescott, J. (2002). A comparison of the nutritional value,
sensory qualities and food safety of organically and conventionally
produced foods. Critical Reviews in Food Science and Nutrition, 42(1),
1–34.
Cantoni, C., Berretta, G., & Bianchi, M. A. (1978). Ammonia, nitrites and
nitrates in sodium chloride. Archivio Veterinario Italiano, 29(5/6),
166–167 (Italian with English summary).
Casaburi, A., Blaiotta, G., Mauriello, G., Pepe, O., & Villani, F. (2005).
Technological activities of Staphylococcus carnosus and Staphylococcus
simulans strains isolated from fermented sausages. Meat Science, 71,
643–650.
Cassens, R. G. (1990). Nitrite-cured meat. Trumbull, CT: Food and
Nutrition Press, Inc.
Cassens, R. G. (1997a). Composition and safety of cured meats in the
USA. Food Chemistry, 59, 561–566.
Cassens, R. G. (1997b). Residual nitrite in cured meat. Food Technology,
51(2), 53–55.
Christiansen, L. N. (1980). Factors influencing botulinal inhibition by
nitrite. Food Technology, 34(5), 237–239.
Code of Federal Regulations. (2006). Animals and animal products. 9
CFR 317.17, 9 CFR 319.2 and 9 CFR 424.21-23. Washington, DC.
<http://www.access.gpo.gov/nara/cfr/waisidx_06/9cfrv2_06.html#301>.
Accessed 28.12.06.
Cornforth, D. P., & Jayasingh, P. (2004). Colour and pigment. In W. K.
Jensen, C. Devine, & M. Dikeman (Eds.), Encyclopedia of meat
sciences (pp. 249–256). Oxford, UK: Elsevier Ltd.
Coughlin, J. R. (2006). Update on International Agency for Research on
Cancer monograph on ingested nitrite and nitrate. In Proceedings of
the meat industry research conference. <http://www.meatscience.orgP-
ubs/mircarchiv/2006/default.htm>. Accessed 30.12..06. October 4–5,
2006, Hollywood, FL, USA.
Devcich, D. A., Pedersen, I. K., & Petrie, K. J. (2007). You are what you
eat: Modern health worries and the acceptance of natural and synthetic
additives in functional foods. Appetite, 48, 333–337.
Dreezens, E., Martijn, C., Tenbult, P., Kok, G., & deVries, N. K. (2005).
Food and values: an examination of values underlying attitudes
toward genetically-modified and organically grown food products.
Appetite, 44, 115–122.
Fox, J. B., Jr., Sebranek, J. G., & Phillips, J. G. (1994). Kinetic analysis of
the formation of nitrosylmyoglobin. Journal of Muscle Foods, 5, 15–25.
Heinerman, J., & Anderson, G. (2001). Salt: Nature’s fifth element. Chico,
CA: Ocean Press International.
Herrador, M. A., Sayago, A., Rosales, D., & Asuero, A. G. (2005).
Analysis of a sea salt from the Mediterranean Sea. Alimentaria, 360,
85–90.
Honikel, K. O. (2004). Curing agents. In W. K. Jensen, C. Devine, & M.
Dikeman (Eds.), Encyclopedia of meat sciences (pp. 195–201). Oxford,
UK: Elsevier Ltd.
Husak, R. L. (2007). A survey of commercially available broilers
originating from organic, free-range, and conventional production
 J.G. Sebranek, J.N. Bacus / Meat Science 77 (2007) 136–147
systems for meat yield, composition and relative value. M.S. thesis,
Iowa State University, Ames, IA, USA.
Kanner, J., & Juven, B. J. (1980). S-nitrosocysteine as an antioxidant,
color-developing and anticlostridial agent in comminuted turkey meat.
Journal of Food Science, 45, 1105–1108, 1112.
Kuhnlein, H. V. (1980). The trace element content of indigenous salts
compared with commercially refined substitutes. Ecology of Food and
Nutrition, 10(2), 113–121.
MacDonald, B., Gray, J. I., Kakuda, Y., & Lee, M. L. (1980). Role of
nitrite in cured meat flavor: Chemical analysis. Journal of Food
Science, 45, 889–892.
MacDonald, B., Gray, J. I., Stanley, D. W., & Usborne, W. R. (1980).
Role of nitrite in cured meat flavor: sensory analysis. Journal of Food
Science, 45, 885–888, 904.
MacDonald, B., Gray, J. I., & Gibbins, L. N. (1980). Role of nitrite in
cured meat flavor: Antioxidant role of nitrite. Journal of Food Science,
45, 893–897.
Milkowski, A. L. (2006). The controversy about sodium nitrite. In
Proceedings of the meat industry research conference. <http://
www.meatscience.org/Pubs/mircarchiv/2006/default.htm>. Accessed
30.12.06. October 4–5, 2006, Hollywood, FL, USA.
Mitchell, R. (2006). Outrunning supplies. Meat and Deli Retailer.
November, 2006. <http://www.meatanddeliretailer.com>. Accessed
28.12.06.
Møller, J. K. S., & Skibsted, L. H. (2002). Nitric oxide and myoglobins.
Chemical Reviews, 102, 1167–1178.
Møller, J. K. S., Jakobsen, M., Weber, C. J., Martinussen, T., Skibsted, L.
H., & Bertelsen, G. (2003). Optimization of colour stability of cured
ham during packaging and retail display by a multifactorial design.
Meat Science, 63, 169–175.
Morrissey, P. A., & Techivangana, J. Z. (1985). The antioxidant activities
of nitrite and nitrosylmyoglobin in cooked meats. Meat Science, 14,
175–190.
National Academy of Sciences (1981). The health effects of nitrate, nitrite
and n-nitroso compounds. Washington, DC: National Academy Press.
National Academy of Sciences (1982). Alternatives to the current use of
nitrite in foods. Washington, DC: National Academy Press.
O’Connor, M. E. (2006). Natural raising claims – a discussion of current
natural regulations and outlook towards the future. In Proceedings of
the meat industry research conference. <http://www.meatscience.org/
Pubs/mircarchiv/2006/default.htm>. Accessed 30.12.06. October 4–5,
2006, Hollywood, FL, USA.
Olesen, P. T., Meyer, A. S., & Stahnke, L. H. (2004). Generation of flavor
compounds in fermented sausages – the influence of curing ingredients,
Staphylococcus starter culture and ripening time. Meat Science, 66,
675–687.
Organic Trade Association. (2006). OTA Manufacturer Survey – Sum-
mary. <http://www.ota.com>. Accessed 28.12.06.
Pegg, R. B., & Shahidi, F. (2000). Nitrite curing of meat. The n-nitrosamine
problem and nitrite alternatives. Trumbull, CT: Food and Nutrition
Press, Inc.
Peters, J. M., Preston-Martin, S., London, S. J., Bowman, J. D., Buckley,
J. D., & Thomas, D. C. (1994). Processed meats and risk of childhood
leukemia (California, USA). Cancer Causes and Control, 5, 195–202.
Petrak, L. (2005). Retailers go natural. Meat and Deli Retailer. December,
2005. <http://www.meatanddeliretailer.com>. Accessed 30.12.06.
147
Pierson, M. D., & Smoot, L. A. (1982). Nitrite, nitrite alternatives and the
control of Clostridium botulinum in cured meats. Critical Reviews in
Food Science and Nutrition, 17, 141–187.
Preston-Martin, S., & Lijinsky, W. (1994). Cured meats and childhood
cancer (Letters to the Editor). Cancer Causes and Control, 5, 484–485.
Preston-Martin, S., Pogoda, J. M., Mueller, B. A., Holly, E. A., Lijinsky,
W., & Davis, R. L. (1996). Maternal consumption of cured meats and
vitamins in relation to pediatric brain tumors. Cancer Epidemiology, 5,
599–605.
Saher, M., Lindeman, M., & Hursti, U.-K. K. (2006). Attitudes toward
genetically modified and organic foods. Appetite, 46, 324–331.
Saltworks. (2006). Salt information. <http://www.saltworks.us/salt_info/
si_gourmet_reference.asp>. Accessed 5.1.07.
Sarasua, S., & Savitz, D. A. (1994). Cured and broiled meat consumption
in relation to childhood cancer: Denver, Colorado (United States).
Cancer Causes and Control, 5, 141–148.
Sebranek, J. G. (1979). Advances in the technology of nitrite use and
consideration of alternatives. Food Technology, 33(7), 58–62, 93.
Sebranek, J. G. (2006). Unpublished data.
Sebranek, J. G., & Fox, J. B. Jr., (1985). A review of nitrite and chloride
chemistry: Interactions and implications for cured meats. Journal of
the Science of Food and Agriculture, 36, 1169–1182.
Sebranek, J. G., & Fox, J. B. Jr., (1991). Rate of nitric oxide formation
from nitrite as affected by chloride ion concentration. Journal of
Muscle Foods, 2, 149–163.
Siderer, Y., Maquet, A., & Anklam, E. (2005). Need for research to
support consumer confidence in the growing organic food market.
Trends in Food Science and Technology, 16, 332–343.
Sindelar, J. J. (2006a). Investigating uncured no-nitrate-or-nitrite-added
processed meat products. Ph.D. thesis. Iowa State University, Ames,
IA, USA.
Sindelar, J. J. (2006b). Unpublished data. Iowa State University, Ames,
IA, USA.
Sindelar, J. J., Cordray, J. C., Sebranek, J. G., Love, J. A., & Ahn, D. U.
(in press a). Effects of varying levels of vegetable juice powder and
incubation time on color, residual nitrate and nitrite, pigment, pH and
sensory attributes of ready-to-eat uncured, emulsified frankfurter-style
cooked sausage. Journal of Food Science.
Sindelar, J. J., Cordray, J. C., Sebranek, J. G., Love, J. A., & Ahn, D. U.
(in press b). Effects of varying levels of vegetable juice powder and
incubation time on color, residual nitrate and nitrite, pigment, pH and
sensory attributes of ready-to-eat uncured ham. Journal of Food
Science.
Stamler, J. S., & Meissner, G. (2001). Physiology of nitric oxide in skeletal
muscle. Physiology Reviews, 81, 209–237.
Tompkin, R. B. (2005). Nitrite. In P. M. Davidson, J. N. Sofos, & A. L.
Branen (Eds.), Antimicrobials in food (3rd ed.). Boca Raton, FL: CRC
Press, Taylor & Frances Group.
USDA. (1995). Processing Inspector’s Calculations Handbook (FSIS
Directive 7620.3). <http://www.fsis.usda.gov/OPPDE/rdad/FSISDi-
rectives/7620-3.pdf>. Accessed 3.1.07.
USDA. (2005). Food standards and labeling policy book. <www.fsis.usda.
gov/OPPDE/larc/Policies/Labeling_Policy_Book_082005.pdf>.
Accessed 27.12.06.
Winter, C. K., & Davis, S. F. (2006). Organic foods. Journal of Food
Science, 71(9), R117–R124.

Meat Science 77 (2007) I–II
AMSA Mission
AMSA is a broad-reaching organization
of individuals that develops and
disseminates its collective food and
animal science knowledge to provide
meat science education and
professional development.
Purposes/Activities of AMSA
• To provide a distinguished forum for all meat
interests - industry, academia, government,
consumer - to address the needs of all in meat
science technology
• To promote the application of science and
technology to production, processing,
packaging, distribution, preparation, evalua-
tion, and utilization of meat and meat products
• To stimulate the exchange, discussion, and
dissemination of information concerning meat
research
• To encourage sound and useful research and
educational techniques
• To coordinate educational, research, develop-
ment, and service activities in meat science
and related areas
• To encourage recognition of those engaged in
meat science
AMSA Members Receive
• Annual Proceedings of the Reciprocal Meat
Conference (online) and the Meat
Industry Research Conference (online)
• AMSA Directory of Members - a listing of
addresses and phone numbers (online) for
networking with other research, teaching,
extension, regulatory, supplier, and
production members
• Bi-weekly AMSA Newsletter of activities, meat
science and industry updates, and employment
opportunities
• Discounted subscription rates to Meat
Science, the official journal of AMSA
• Unlimited access to all online ‘member-only’
information
www.meatscience.org
0309-1740/$ - see front matter  2003 Published by Elsevier Ltd.
doi:10.1016/S0309-1740(07)00190-8
MEAT
SCIENCE
www.elsevier.com/locate/meatsci
Benefits of AMSA Membership
• Direct, personal contact at AMSA meetings with
meat scientists from academia, industry, and
government
• Opportunities to learn about current meat
research, teaching, extension, industry, and
government activities in the United States
and worldwide
• Opportunity to attend various exclusive short
courses on meat science subjects and an
invitation to attend the International Congress
of Meat Science and Technology (ICoMST)
• Dissemination of scientific information at joint
programs with American Association of Meat
Processors, American Meat Institute, National
Cattlemen’s Beef Association, National Pork
Producers Council, National Meat Association
• Opportunity to become involved in the programs
that shape your career by participating on
committees to determine the structure of
meetings and industry programs
Creating Opportunities For Your Career
The American Meat Science Association
provides you with many valuable
opportunities to enhance your career in the
meat science industry. AMSA provides you
with a forum for continuing education;
opportunities to network with meat science
colleagues from academia, industry, and
government; the latest information and
technology in meat science and related
areas; and the ability to participate in top
quality meetings, courses, and programs.
Types of AMSA Membership
Professional
Any individual who is active in any aspect of meat
science and who displays an interest in support-
ing the objectives of the Association shall be
eligible to become a professional member
Student
Any individual actively pursuing candidacy for a
degree (Associate or higher) with a science or
technology major associated with meat science
II AMSA Membership and Benefits / Meat Science 77 (2007) I–II

American Meat Science Association

Membership Application www.meatscience.org

Please circle one

Name Dr. Mrs. _________________________________________________________________
Mr. Ms. Last First Middle Initial

E-Mail _________________________________________ Title __________________________________________

Organization ___________________________________ Department ___________________________________
q Business Address __________________________________________________________________________
City _______________________________________ State/Province ________________________________
Zip/Postal Code ___________________________ Country _______________________________________
q Home Address _____________________________________________________________________________
City _______________________________________ State/Province ________________________________
Zip/Postal Code ___________________________ Country _______________________________________
Please check one box for preferred mailing address
Business Phone Number _____________________________ Fax Number _____________________________
Home Phone Number ___________________________________________________________________________
Sex: (please circle one) M F Date of Birth _________________________________

Academic Record Institution Degree Year Conferred

Undergraduate Degree(s) _________________________________________________________________________________
Advanced Degree(s) ______________________________________________________________________________________
____________________________________________________________________

MEMBERSHIP DATABASE (this information will allow AMSA to TYPE OF MEMBERSHIP DUES
provide you with more valuable networking opportunities) q Professional $145
AREAS OF SPECIAL INTEREST (please check all that apply) q Student $65
q Animal Physiology q Human Nutrition PLEASE ADD IN ADDITION TO MEMBERSHIP DUES
q Biochemistry-Biophysics q Livestock and q Meat Science Journal (optional) $125
q By-Products Meat Production q International Shipping Costs $45
q Food Chemistry q Microbiology/Food Safety (other than U.S., Canada, and Mexico)
q Fresh Meat Processing q Packaging TOTAL AMOUNT ENCLOSED $ _______________
q Growth Biology q Processed Meats
q Histology/Histochemistry q Teaching and Education PAYMENT
q Other ________________________________________________ q Check (must be drawn on a U.S. bank in U.S. funds)
q Visa q Mastercard q American Express
JOB CODES (please check all that apply) Card #______________________________________________
Government Industry Exp. Date __________________________________________
q Regulatory q Association Signature __________________________________________
q Research q Marketing
The membership dues must accompany this application. The membership
q Packaging year is January 1 through December 31. Dues are not prorated. A
Academic q Plant Manager membership becomes void if not renewed by March 31 of the calendar
q Administration q Purchasing year to which it applies.
q Extension q Quality Control
q Research q Research & Development Date _______________________________________________
q Student q Sales Applicant Signature

q Teaching q Senior Management Signature of Sponsor ________________________________

q Technical Consultant (Professional Member of AMSA) Sponsor Signature

Send application and payment to the

American Meat Science Association
1111 N. Dunlap Ave., Savoy, IL 61874
phone: (217) 356-5368 fax: (217) 465-0688