14
Protein Metabolism and Turnover
D. Attaix,
1
D. Re
´
mond
1
and I.C. Savary-Auzeloux
2
1
Institut National de la Recherche Agronomique, Unite
´
de Nutrition et
Me
´
tabolisme Prote
´
ique, Theix, 63122 Ceyrat, France;
2
Institut National
de la Recherche Agronomique, Unite
´
de Recherches sur les Herbivores,
Theix, 63122 Ceyrat, France
Introduction
All cellular proteins are in a continuous state of turnover in which they are
synthesized and degraded (Waterlow et al., 1978). Thus, the intracellular
concentration of any protein, and the tissue, organ or whole-body protein
mass, are determined by the relative synthetic and degradation rates. It should
be pointed out that a change in the size of a given protein pool only depends on

5. The intracellular abundance of key proteins (e.g. enzymes, cyclins or
transcription factors) is tightly regulated so that major biological processes are
precisely controlled.
A major challenge is to understand both general and tissue/organ-specific
mechanisms, which are responsible for these adaptations. In vitro studies have
provided detailed information on the regulatory mechanisms of protein turn-
over. In vivo studies are inevitably more descriptive, and experiments in animal
production are mostly designed to optimize protein deposition efficiency in
skeletal muscle (meat) or milk production. Furthermore, the cost of research in
large animal species has clearly impeded our understanding of protein metab-
olism in ruminants, so that most available information remains fragmentary.
Mechanisms of Protein Turnover
The precise mechanisms of protein synthesis, which include transcription,
translation and post-translational modifications, have been extensively studied
and are detailed in many textbooks of biochemistry. The mechanisms that
regulate protein breakdown are much more obscure. First, there are several
proteolytic pathways within cells (e.g. lysosomal, Ca
2þ
-dependent, ubiquitin–
proteasome-dependent (see Fig. 14.1), etc.), and many proteases remain to
be discovered or characterized. In addition, the relative contribution of proteo-
lytic pathways to the rate of overall proteolysis is tissue specific. The lysosomal
pathway plays a prominent role in liver (Attaix et al., 1999), while the ubiqui-
tin–proteasome system has a major importance in skeletal muscle (Attaix and
Taillandier, 1998; Jagoe and Goldberg, 2001). Second, there are many alter-
native routes within a given proteolytic process (Attaix et al., 1999). Third,
in vivo, different proteolytic systems may either independently degrade a given
protein substrate (Attaix et al., 1999), or sequentially participate to its com-
plete hydrolysis into free amino acids (Attaix et al., 2002).
Protein synthesis requires the hydrolysis of both ATP and GTP. However,

n
E3 +
DUB (4)
n Ub + protein
26S Proteasome (5)
nUb + peptides
E1 +
ATP
Protein
(1) (2) (3)
(6)
TPP II
Free AA
+ AP
Fig. 14.1. Schematic representation of the ubiquitin (Ub)–proteasome-dependent proteolytic
pathway. Polyubiquitination of the substrate is achieved in sequential steps (1) to (3). (1) The
Ub-activating enzyme, E1, forms a thiol–ester bond with Ub. (2) The activated Ub is then
transferred to an Ub-conjugating enzyme, E2, which also forms a thiol–ester linkage with Ub.
(3) In the presence of an Ub–protein ligase, E3, that specifically recognizes the substrate, the E2
and / or E3 covalently binds a polyUb chain (Ub)
n
to the target protein. (4) A huge family of
deubiquitinating enzymes (DUB) can remove the polyUb degradation signal, so that the substrate is
not degraded and free ubiquitin is recycled. (5) More generally, the polyUb degradation signal is
recognized by the 26S proteasome, and the substrate is cut into peptides with recycling of free Ub.
(6) The peptides generated by the proteasome are finally hydrolysed into free amino acids (AA) by
the tri-peptidyl peptidase II (TPP II) and several associated aminopeptidases (AP) (see Attaix et al.,
2002 for more detailed information).
Protein Metabolism and Turnover 375
ILR ¼ Synthesis(S) þ Oxidation(O) ¼ Breakdown(B) þ Intake(I)

low, and is low (0.6 to 0.3) in tissues where
protein turnover is rapid (liver, gut)). In
(b), this problem is minimized over a short
period of time, and this ratio is usually over
0.7, including when protein turnover is a
rapid process (see Attaix and Arnal, 1987).
(b)
Plasma
Plasma
Muscle
Liver
Free label specific activity
Time after label administration
10 20 30
40
50 min
(a)
Muscle
Liver
1234 h5
Free AA Protein
(I)
(O)
(S)
(B)
Tracer
Fig. 14.3. Two-pool model used for the estimation of the whole-body irreversible loss rate (ILR)
and tissue protein fractional synthesis rate (FSR) in vivo, see text. Amino acid (AA) fluxes, which
are inputs into the free amino acid pool (e.g. intake (I) and protein breakdown (B)), and outputs
from this pool (e.g. protein synthesis (S) and amino acid oxidation (O)) are shown. The tracer,

ures across the portal-drained viscera (PDV) and liver in sheep (Lobley et al.,
1996). Such procedures require extensive surgery, but they allow repeated
measurements within the same animal.
Tissue and organ protein turnover
Protein synthesis
To measure fractional rates of protein synthesis (FSR, usually expressed in %
per day) in vivo the specific radioactivity (or enrichment) of the labelled amino
acid must be measured in both the precursor and the protein pools (Waterlow
et al., 1978). Except for skeletal muscle and skin, in which biopsies can be
easily performed, slaughter is usually required to collect internal samples. Two
techniques have provided most of the data available in ruminants.
The most commonly used is the constant tracer infusion analysis, as in the
ILR technique (see above and Fig. 14.2a). The difficulty is to estimate the activity
of the precursor pool for protein synthesis. The activity of the actual pool, the
charged aminoacyl-tRNAs, is technically very difficult to determine. Based on
experiments performed in vitro and in vivo, it is generally assumed that ami-
noacyl-tRNAs are charged from both extracellular (plasma) and intracellular
Protein Metabolism and Turnover 377
(tissue homogenate) free amino acid pools (Waterlow et al., 1978). However, as
the label is diluted by the unlabelled amino acid used as a marker, which arises
from protein breakdown, there are large differences between the isotopic activ-
ities in these pools (Fig. 14.2a). This is especially true when protein turnover is
high (liver, GIT). Consequently there are also large differences between FSR
calculated by using the isotopic activity of the free label in the plasma and the
tissue homogenates. In addition, since the label is infused during several hours,
secreted or export proteins, which are for example synthesized in the liver and
the intestines, are not taken into account in the measurements.
To overcome all these problems, the label can be injected with a large or
flooding dose of the same unlabelled amino acid. This results in nearly constant
and close isotopic activity of the tracer, both in the plasma and in tissue hom-

residues in actin and in myosin heavy chains of fast-twitch glycolytic skeletal
muscles. In the rat and cattle, but not all species (see below), the urinary excretion
of 3-methylhistidine provides an index of myofibrillar protein breakdown.
Unfortunately, the visceral smooth muscles of the GIT and other tissues such
as skin contain significant amounts of actin. These tissues contribute dispropor-
tionately for their size to 3-methylhistidine urinary excretion, because of
378 D. Attaix et al.
their high rates of protein turnover. In addition, changes in renal clearance of 3-
methylhistidine may affect the interpretation of the data (see Attaix and Taillan-
dier, 1998). Finally, in some species (e.g. in pigs and to a lesser extent in sheep),
a high proportion of 3-methylhistidine is retained in muscle as a dipeptide,
balenine (Harris and Milne, 1987). A compartmental model of 3-methylhistidine
metabolism has been developed, which involves the assessment of muscle
proteolysis and 3-methylhistidine kinetics without the collection of urine (Rath-
macher and Nissen, 1998). However, due to the numerous limitations of the 3-
methylhistidine approach, caution must be exercised.
Non-quantitative approaches
Non-quantitative approaches may be of special interest in ruminant tissues, due
to the costs of experiments with isotopic amino acids. As a very crude rule, the
control of protein synthesis occurs mainly at the transcriptional level. Therefore
the quantification of the mRNA(s) of a given protein by molecular biology
techniques is often used as an index of protein synthesis. However, many
mRNAs are also subject to translational control, and the relative amount of
any mRNA depends on both rates of transcription and of mRNA breakdown.
Finally, there are frequent discrepancies between mRNA levels and the corre-
sponding protein levels and/or activities. Similarly, changes in mRNA levels for
many proteolytic genes, in particular within the muscle ubiquitin–proteasome-
dependent pathway, closely mimic variations of proteolytic rates measured with
incubated rodent muscles (see Attaix and Taillandier, 1998). These observa-
tions, together with the use of specific inhibitors of lysosomal and Ca

BW
0:75
for sheep and steers, respectively) the slope of the relationship is very
similar (e.g. 13–14 g of protein synthesized per MJ ME). However, below main-
tenance, protein synthesis decreases in sheep but is not altered in steers (Lapierre
et al., 1999). Above maintenance requirements, the calculated whole-body pro-
tein degradation rate (protein synthesis minus deposition) increases in both sheep
and steers (Harris et al., 1992; Lapierre et al., 1999). Below maintenance
protein breakdown decreases in sheep (Harris et al., 1992), but increases in steers
(Lapierre et al., 1999). Besides species differences, the duration of the under-
feeding period, the composition of the diet and the age of animals may account for
these discrepancies. Nevertheless, whole-body protein loss was similar (about
1g/day/kgBW
0:75
) in both underfed (0.6Â maintenance) steers and sheep.
Tissue Protein Metabolism
Portal-drained viscera
On average, the portal net release of essential amino acids accounts for only
two-thirds of their apparent disappearance from the small intestine (MacRae
0
10
20
30
40
50
60
70
−100 102030405060708090100
WB protein synthesis (g/day/kg BW
0.75

30
35
40
0 200
(a)
(b)
400 600 800 1000 1200
Metabolizable energy (kJ/day/kg BW
0.75
)
WB protein synthesis (g/day/kg BW
0.75
)WB protein synthesis (g/day/kg BW
0.75
)
Lobley et al. (1987)
Hammond et al. (1987)
Dawson et al. (1998)
Lapierre et al. (1999)
Lobley et al. (2000) (Angus)
Lobley et al. (2000) (Charolais)
Fig. 14.5. Effect of metabolizable energy intake on whole-body (WB) protein synthesis in cattle
(a) and sheep (b).
Protein Metabolism and Turnover 381
metabolism of dietary amino acids as well as PDV use of systemic amino acids
significantly impact the quantitative and qualitative supply of amino acids to
other tissues or organs. The portal vein drains heterogeneous tissues (GIT,
pancreas, spleen, omentum), but the GIT is by far the major contributor to
PDV protein synthesis. For this reason, only GIT protein metabolism is
reviewed below.

Jejunum
Ileum
Caecum
Colon
Protein fractional synthesis rate (% per day)
1-week-old 8-week-old 8-month-old
Fig. 14.6. Protein fractional synthesis rates in the gastrointestinal tract from milk-fed
(1-week-old) and weaned lambs. (Data from Attaix, 1988; Lobley et al., 1994.)
382 D. Attaix et al.