Am J Physiol Endocrinol Metab 283: E1105–E1112, 2002;

10.1152/ajpendo.00337.2002.

invited review
Human protein metabolism: its measurement and regulation

ZHENQI LIU AND EUGENE J. BARRETT

Division of Endocrinology and Metabolism, Department of Internal
Medicine, and the General Clinical Research Center, University
of Virginia Health Sciences Center, Charlottesville, Virginia 22908

Liu, Zhenqi, and Eugene J. Barrett. Human protein metabolism:

its measurement and regulation. Am J Physiol Endocrinol Metab 283:
E1105–E1112, 2002;10.1152/ajpendo.00337.2002.—The body’s protein
mass not only provides architectural support for cells but also serves
vital roles in maintaining their function and survival. The whole body
protein pool, as well as that of individual tissues, is determined by the
balance between the processes of protein synthesis and degradation.
These in turn are regulated by interactions among hormonal, nutri-
tional, neural, inflammatory, and other influences. Prolonged changes in
either the synthetic or degradative processes (or both) that cause protein
wasting increase morbidity and mortality. The application of tracer
kinetic methods, combined with measurements of the activity of compo-
nents of the cellular signaling pathways involved in protein synthesis
and degradation, affords new insights into the regulation of both protein
synthesis and breakdown in vivo. These insights, including those from
studies of insulin, insulin-like growth factor I, growth hormone, and
amino acid-mediated regulation of muscle and whole body protein turn-
over, provide opportunities to develop and test therapeutic approaches
with promise to minimize or prevent these adverse health consequences.
protein synthesis; proteolysis; insulin; insulin-like growth factor I;
growth hormone; amino acids

IN HUMANS, THE BODY PROTEIN MASS provides architectural
support, enzymes to catalyze metabolic reactions, sig-
naling intermediates within and between cells, and
fuel to allow survival. Because there is no protein
storage pool, proteins that serve vital roles are catab-
olized when the body is faced with a need for additional
fuel expenditure. Subsequently, the body replaces pre-
viously sacrificed proteins during periods of caloric/
nitrogen surfeit, thereby keeping a delicate and dy-
namic protein balance that maintains homeostasis in
the face of environmental challenge.
Protein synthesis and degradation are closely regu-
lated in vivo, and each is affected by physiological and
pathophysiological conditions, such as fasting, feeding,
exercise, disease, and aging. Many diseases can de-
crease protein synthesis and/or enhance protein degra-
dation, causing negative protein balance. While tran-
siently this may afford benefit by supplying amino
acids for gluconeogenesis, wound healing, and synthe-
sis of antibodies and acute-phase proteins, prolonged
negative protein balance is associated with signifi-
cantly increased morbidity and mortality. Our under-
standing of the cellular and integrated regulation of
protein synthesis and degradation has advanced con-
siderably over the past two decades. Here, we briefly
review selected aspects of this progress, particularly
isotopic methods used to measure whole body and
tissue-specific protein turnover as well as newer trac-
er-independent methods for studying the regulation of
body protein metabolism. Finally, we will briefly re-
view how these methods are being applied to studies of
insulin, insulin-like growth factor I (IGF-I), growth
hormone (GH), and amino acid-mediated regulation of
muscle protein turnover.
METHODS FOR MEASUREMENT OF PROTEIN
METABOLISM IN VIVO

Address for reprint requests and other correspondence: E. J. Barrett, For any cell or tissue, protein balance reflects the net
Dept. of Internal Medicine, PO Box 801410, Univ. of Virginia Health of protein synthesis and degradation. The rates of
Sciences Center, Charlottesville, VA 22908 (E-mail EJB8X@virginia.edu). protein synthesis and degradation differ drastically
http://www.ajpendo.org 0193-1849/02 $5.00 Copyright © 2002 the American Physiological Society E1105
E1106 INVITED REVIEW
among organs/tissues, and between the cytosolic, nu-
clear, and mitochondrial compartments of a cell. Whole
body protein balance is the integral of these processes.
Many attempts to quantify either whole body or tissue-
specific protein metabolism in humans in vivo have
been made during the past two to three decades. These
efforts have been almost exclusively based on use of
tracer measurements. However, these complex dy-
namic processes require model assumptions that limit
the accuracy of all available methods, and each method
has inherent shortcomings and limitations.
Whole Body and Organ-Specific Nitrogen Balance
Early studies of whole body nitrogen balance simply
employed chemical analyses of nitrogen intake and
excretion (8). With the development of automated
amino acid analyses, the net balance of all amino acids
across a tissue bed (muscle, liver, brain, and so forth)
became measurable (44). These estimates at steady
state provide a net measurement of the protein accre-
tion/loss, but no insight into the processes of protein
synthesis and degradation at either the whole body or
tissue level. However, because these measurements do
not require the modeling assumptions that accompany
the use of tracer methods, they remain a useful, inde-
pendent method for assessing whole body or organ
responses to nutrient, hormonal, and other manipula-
tions.
Whole Body and Organ-Specific Protein Turnover
To extend investigation beyond the limitations of the
whole body nitrogen or tissue amino acid balance tech-
niques, several approaches were developed. Each in-
volves the use of an essential amino acid tracer and is
based on the assumption that the kinetics of the essen-
tial amino acid tracer used represent the kinetics of the
protein pool. Because essential amino acids are not
synthesized in vivo, the steady-state dilution of the
infused tracer-specific activity/enrichment in blood re-
flects the inflow into blood (or rate of appearance, Ra) of
unlabeled amino acids derived from protein degrada-
tion and dietary intake. Conversely, the rate of tracer
disappearance (Rd) from blood represents the sum of
amino acid oxidation to CO2 and nonoxidative disposal,
which, with the appropriately chosen tracer, will prin-
cipally reflect body protein synthesis.
Whole body protein turnover using isotopic tracers.
Whole body protein turnover is now most commonly esti-
mated using L-[1-*C]leucine (36) or L-[*H5]phenylalanine
labeled with either radioactive or stable isotopic trac-
ers and given as a primed, continuous intravenous
infusion. With a continuous intravenous tracer infu-
sion, the arterial plasma tracer enrichment/specific
activity reaches a plateau (steady state) in ⬃2 h. At
steady state, whole body protein turnover equals the
rate of tracee appearance (Ra) in the plasma free amino
acid pool and the rate of tracee disappearance (Rd) from
the free amino acid pool, and each can be estimated
from the tracer infusion rate divided by the arterial
plasma-specific activity/enrichment. When labeled
AJP-Endocrinol Metab • VOL
leucine is used as the tracer, venous ␣-ketoisocaproic
acid (␣-KIC), the transamination product of leucine,
may better reflect the enrichment of leucine in the
cellular free amino acid pool (52). In addition, with
leucine labeled in carbon 1, provided the loss of label to
CO2 is measured, both protein synthesis and degrada-
tion can be quantified. This approach provides the most
accurate method currently available to simultaneously
estimate whole body protein synthesis and degrada-
tion. With this method, because of the rapid equilibra-
tion of the infused tracer, it is possible to “perturb” the
steady state and estimate dynamic changes in re-
sponse to specific interventions (see Hormonal and
Nutritional Regulation of Protein Turnover in Vivo).
However, there are limitations (6). This method cer-
tainly underestimates the total turnover of tissue pro-
teins. This is partially compensated in the case of
skeletal muscle by measuring the specific activity (or
enrichment) of ␣-KIC (24). It is less clear that this
compensation is satisfactory for other tissues with
lower transaminase activities. The estimates of syn-
thesis are also affected by assumptions regarding the
body bicarbonate pool. These issues aside, the steady-
state [1-*C]leucine infusion method is currently the
accepted standard for measurement of whole body pro-
tein turnover.
Measurement of protein turnover in specific tissues/
organs. Measurements of both protein synthesis and
degradation within specific tissues/organs are obtained
by using a combination of organ balance and tracer ap-
proaches. The methods require arterial sampling to-
gether with sampling of venous blood draining the tissue
being studied. With this method, one gives either a
leucine (12) or phenylalanine (5, 21) tracer by primed,
continuous intravenous infusion until a steady state is
achieved in plasma. During steady-state infusion of
tracer, measurement of the uptake of tracer by the tissue
(Rd) and of the dilution of its specific activity as it
traverses the tissue (Ra) can be used to estimate protein
synthesis and breakdown, respectively. The calculation is
based on the measurement of the concentration and spe-
cific activity (SA)/enrichment of systemically infused
tracer in both arterial blood entering and venous blood
draining the tissue under study, together with blood flow
to the tissue. The net balance of leucine or phenylalanine
simply reflects the difference between its uptake for pro-
tein synthesis and its release from degraded protein. The
formulas used for these calculations are detailed else-
where (4). In these estimates it is assumed that either the
arterial or the venous blood SA/enrichment at steady
state approximates that of the pool being used for protein
synthesis. The amino acid selected as tracer is dictated
somewhat by the tissue under study. For skeletal muscle
(5), heart (48), and adipose tissue (14), phenylalanine
appears most appropriate. Phenylalanine is neither syn-
thesized nor concentrated by the muscle and is well
represented in muscle protein; its only metabolic fate in
muscle is to be incorporated into protein, and the only
endogenous source of phenylalanine released into blood
traversing muscle is protein. For liver, leucine would be a
preferred tracer, because the fraction of its metabolic fate
283 • DECEMBER 2002 • www.ajpendo.org
 INVITED REVIEW
directed other than to protein synthesis (42) is less than
that of phenylalanine.
The major advantage of this method is that it allows
simultaneous estimation of both synthesis and degra-
dation. Because it relies on steady-state conditions
being reached in the plasma pool, repeated sampling
can be performed to assure this condition, and because
this requires only ⬃2 h, it is amenable for the study of
the effects of physiological interventions.
In studies directed more specifically at examining
the process of protein synthesis in tissues in vivo,
several methods are currently regularly used. The first
is an adaptation of the steady-state leucine or phenyl-
alanine tracer method just described. It involves a
continuous tracer infusion that persists until a steady
state of enrichment or specific activity is reached in the
pool considered to be the precursor for protein synthe-
sis (43). Then a tissue sample is taken, and a second
sample is taken hours later, with the tracer remaining
at steady-state enrichment between samplings. The
proteins in the biopsy samples are analyzed for tracer
enrichment. The synthesis rate is calculated by divid-
ing the change in tracer enrichment in the protein-
bound amino acid by the steady-state tracer enrich-
ment in the precursor pool over a given period of time.
Optimally, labeling of the amino acyl-tRNA pool would
provide the precursor pool for protein synthesis. How-
ever, its small size relative to the bulk amino acid pool
and technical difficulties with its measurement have
precluded this approach in most circumstances. In-
stead, enrichment or specific activity of the amino acid
label in the total tissue extract is typically used hope-
fully to approximate that of the tRNA pool. To circum-
vent this problem, Garlick and colleagues pioneered
first in animals (20) and later in humans (38) the
so-called “flooding dose” technique, which employs a
bolus intravenous injection of large doses of labeled
(tracer) and unlabeled (tracee) amino acid to rapidly
force equilibration of labeling between the intra- and
extracellular amino acid pools. The tracer enrichment
in these pools is assumed to be nearly identical to the
true precursor, i.e., the aminoacyl-tRNA pool, in all
tissues. Protein-bound tracer amino acid is then ex-
tracted from tissues (or proteins), and the fractional
synthesis rate (FSR) is calculated. A potential draw-
back of this technique is that rapid injection of a large
amount of amino acid tracee may itself stimulate pro-
tein synthesis. Most investigators use phenylalanine
as the tracer, since large doses of leucine are capable of
activating translation initiation and increasing the
FSR of mixed muscle protein. Despite use of phenylal-
anine, the FSR measured with this technique is usu-
ally 1.5- to 2-fold higher than that calculated from the
continuous infusion method, a finding not yet ade-
quately resolved.
These latter two methods afford the advantage of
likely providing a more accurate estimate of protein
synthetic rates. However, limitations include the inva-
sive nature of the studies, requiring multiple biopsies
that (at least for clinical investigations) limit studies to
muscle, adipose tissue, and skin; the lack of informa-
AJP-Endocrinol Metab • VOL
E1107
tion regarding protein degradation; and the concern
that it is not trivial to ascertain when the precursor
pool labeling has achieved a steady state.
For muscle, an additional method has been used to
estimate proteolysis that involves measurement of uri-
nary excretion of 3-methylhistidine (3-MH). Histidyl
residues in actin, and to a lesser extent in myosin, are
posttranslationally methylated, and methyl-histidines
are released when these proteins are degraded. More
than 90% of 3-MH residues are in skeletal muscle, and
the urinary excretion rate of 3-MH provides noninva-
sive in vivo estimates of muscle protein degradation in
humans that are consistent with rates estimated by
other methods. However, the analysis requires quanti-
tative collection of urine, a carefully controlled meat-
free diet for several days, and normal renal function
and is limited by the quantities of 3-MH in urine
contributed by nonskeletal muscle sources, which vary
with the 3-MH pool sizes (depending on lean body
mass), their relative contents of 3-MH, and the protein
turnover rates of different tissues as well (7).
Measurement of the Synthesis of Specific Proteins or
Classes of Proteins
For some relatively abundant plasma (albumin, fi-
brinogen, apolipoprotein B-100, and the like) (15) or
tissue (myosin and actin in muscle) (11) proteins, as
well as for several classes of cellular proteins (mito-
chondrial and contractile proteins in muscle) (49), it is
possible to incorporate sufficient amino acid label (gen-
erally as a stable isotope) to allow measurement of
their synthesis rates. These estimates are based upon
many of the same assumptions as the whole body and
whole tissue measures. Despite the slow turnover of
these proteins, their abundance allows reasonable
quantitation of their turnover and responses to inter-
ventions like insulin and feeding.
Nontracer Techniques Applied to Studies of Protein
Metabolism in Vivo
Because of difficulties that appear intrinsic to tracer
estimates of protein metabolism, efforts to study the
regulation of protein synthesis and degradation in vivo
have recently been extended to methods that examine
the activity of the pathways involved in synthesis
(mRNA transcription and activation of translational
regulatory elements) and degradation. In both experi-
mental animals and humans, measurement of tran-
scription of specific mRNAs has been accomplished by
classical Northern blotting, PCR, and RNAse protec-
tion assays, and more recently by use of microarray
methods. For translation, the phosphorylation states of
proteins involved in regulating translation initiation
have been coupled with estimates of protein synthesis
to provide a more comprehensive picture of the regu-
lation of this portion of the synthetic pathway.
Efforts to study the regulation of protein degradation
have also progressed and include estimates of the ac-
tivities of one or more of the pathways involved in
proteolysis or transcription of genes that code for pro-
283 • DECEMBER 2002 • www.ajpendo.org
E1108
teins involved in one or another proteolytic pathway (3,
40). Progress with these nontracer measures of prote-
olysis has not kept pace with measures of transcrip-
tion/translation, in part because of the complexity of
the multiple pathways involved in protein degradation,
and this will be an area for future investigation, espe-
cially as molecular probes for components for the cal-
pain, proteasomal, and lysosomal pathways are devel-
oped. This will be particularly important because one
action of inflammatory cytokines appears to involve a
marked increase in proteolytic activity, and this might
contribute to the wasting that occurs in chronic inflam-
matory disorders (13).
All of these methods involve tissue sampling and
provide information on regulatory changes that are
relatively stable in nature, i.e., changes that are pre-
served during tissue sampling and assay. To the extent
that allosteric mechanisms are involved in modulating
the rates of either synthesis or degradation, their im-
pact remains elusive.
SELECTED APPLICATIONS OF IN VIVO METHODS TO
STUDIES OF HUMAN PROTEIN METABOLISM
Whole Body Protein Turnover and Energy Balance
With their significant limitations, what information
can be obtained from whole body protein turnover
methods? Here we point out one simple phenomenon
that we have come to call the “protein paradox,” which
derives quite simply from steady-state measurements
of fuel turnover in postabsorptive humans. It begins
with a comparison of the turnover rates of the body’s
three principal fuels: protein, carbohydrate, and fats.
Figure 1 gives the Ra in plasma of amino acids, free
fatty acids, and glucose in the circulation after an
overnight fast in healthy humans. In each case, the
estimates are derived from steady-state tracer infusion
methods. Among these three fuels, protein is unique in
that there is no storage form that is not already serving
another significant purpose. On this basis, it might be
expected that it would be relatively preserved. This
Fig. 1. Rates of fuel turnover and relative contribution of each fuel to
daily caloric consumption. Turnover rates of whole body protein are
much faster, yet its oxidation is substantially less than that for
either fat or carbohydrate. FFA, free fatty acids.
AJP-Endocrinol Metab • VOL
INVITED REVIEW
would seem more advantageous given the large
amount of energy required to synthesize a protein
relative to that required to store energy as either
carbohydrate or fats (a minimum of 4 ATP equivalents
per peptide bond made and broken vs. 2 for a glycosidic
linkage in glycogen and 2 per ester linkage in triglyc-
eride). Paradoxically, the turnover rate of the body’s
protein pool is substantially higher than that of either
of the other two principal fuels (see Fig. 1). However,
the fraction of amino acids liberated from body protein
via one or another proteolytic pathway that is subse-
quently oxidized fully is substantially less than that
fraction for either fat or carbohydrate. Thus a selective
advantage appears to accrue to the organism when the
unitary building blocks of proteins (amino acids) are
used sparingly as an oxidative fuel, whereas the pro-
tein pool per se turns over rapidly at considerable cost
of energy to the body. The energy expenditure involved
in this turnover of the body protein pool is not insig-
nificant when compared with total body resting energy
expenditure. It can be estimated that, at a minimum,
the turnover of the body protein pool will account for
5% of our resting energy expenditure (33). The nature
of the selective advantage afforded by this plasticity of
our body protein pool is unclear. However, estimates
of this sort require the use of tracer measurements
as described above and cannot be derived by other
methods.
Hormonal and Nutritional Regulation of Protein
Turnover in Vivo
Protein synthesis and degradation are each regu-
lated by multiple hormonal and nutritional factors,
and protein balance of individual tissues and the whole
body changes constantly. Recent studies at the molec-
ular and cellular levels, coupled with tracer methods,
have dramatically advanced our understanding of the
regulation of protein metabolism in vivo. Here we will
selectively review studies relating to the actions of
three peptide hormones, insulin, IGF-I, and GH, and
amino acids that have been more extensively studied
and that affect body protein metabolism acutely and
hence are amenable to study by use of some of the
tracer and biopsy methods we have described. Other
hormones (e.g., thyroid hormones, glucocorticoids, go-
nadal steroids), as well as cytokines and interleukins,
also impact body protein metabolism, but information
on the actions of these effectors is much less complete;
review of these is beyond the scope of the current work.
Our focus in reviewing these three hormones and
amino acids is in part to provide a “status report” of
current information but also to highlight the un-
knowns that will hopefully be areas for further inves-
tigations.
Insulin. In humans, the rates of whole body (54),
splanchnic (42), skeletal (21), and heart (37) muscle
protein degradation are decreased in response to phys-
iological increases in plasma insulin. Conversely, the
rates of protein degradation increase transiently in
heart and skeletal muscle in diabetic animals (2), and
283 • DECEMBER 2002 • www.ajpendo.org
 INVITED REVIEW
the whole body protein turnover is increased in insulin-
withdrawn type 1 diabetic patients (10). Decreasing
insulin availability secondary to either experimental
diabetes or starvation reduces by 40–50% the rate of
protein synthesis in rat skeletal muscle. This is rapidly
reversed with insulin treatment. Numerous in vitro
experiments in cultured cells or isolated perfused tis-
sues have consistently demonstrated that insulin at
high concentrations stimulates protein synthesis and
abrogates proteolysis (26, 28). Surprisingly, whereas
the in vivo antiproteolytic effect of insulin in adult
animals or humans has been repeatedly shown, the
anabolic effect of insulin on protein synthesis has not
been consistently demonstrated. A satisfactory expla-
nation for this apparent discrepancy between in vivo
and in vitro studies is wanting. Factors including the
insulin concentrations used in vitro (typically between
2 and 10 mU/ml), preincubation of tissues in the com-
plete absence of insulin in vitro before insulin addition,
or the overall highly catabolic state of tissues incu-
bated in vitro may all contribute. On the other hand,
an effect of insulin in vivo may be underestimated if
insulin’s antiproteolytic effect significantly decreases
amino acid availability, as amino acids themselves
regulate the rate of mRNA translation (55).
Recent studies that build on early work from the
laboratory of Jefferson and colleagues (Kimball et al.,
Ref. 29) have shed light on the effect of insulin to
regulate mRNA translation, a critical step in insulin’s
action on protein synthesis in isolated muscle, heart,
and liver. Availability of phosphospecific antibodies
and measurement of electrophoretic mobility shift as-
says have allowed linkage of insulin action to stimulate
protein synthesis with its action on key signaling mol-
ecules downstream of phosphatidylinositol 3-kinase
(PI-3 kinase) and protein kinase B (PKB or Akt) in the
insulin-signaling cascade. At the translational level,
insulin has been shown to modulate the phosphoryla-
tion states of several key intermediates in the protein
AJP-Endocrinol Metab • VOL
E1109
Fig. 2. Simplified schematic diagram illustrating the
regulation of translation initiation and protein synthe-
sis by insulin, IGF-I, and amino acids. 2, stimulatory
signal; ——|, inhibitory signal. See text for definition of
terms.
synthetic pathway, among them eukaryotic initiation
factor (eIF) 4E-binding protein 1 (4E-BP1) and ribo-
somal protein S6 kinase (p70S6K). The phosphorylation
of 4E-BP1 and p70S6K, both downstream of PI 3-ki-
nase-Akt-mammalian target of rapamycin (mTOR) in
the insulin-signaling pathway, is closely correlated
with the rate of protein synthesis (Fig. 2). Inhibition of
PI 3-kinase with wortmannin or mTOR with rapamy-
cin blocks insulin-stimulated phosphorylation of
4E-BP1 and p70S6K and partially inhibits protein syn-
thesis. Readers are referred to several excellent recent
reviews that have detailed the regulation of protein
synthesis by insulin and nutrients through modulating
translation initiation (39, 47).
We have examined the effects of insulin, at both
pharmacological and physiological concentrations, on
the phosphorylation of 4E-BP1 and p70S6K in both
humans and laboratory rats. Interestingly, we found
that insulin stimulated the phosphorylation of p70S6K,
but not 4E-BP1, at physiological concentrations (22),
whereas the phosphorylation of both proteins is en-
hanced at pharmacological concentrations (33). Be-
cause ribosomal S6 protein is a principal physiological
substrate for p70S6K, and phosphorylation of S6 en-
hances the translation of a restricted subset of proteins
with oligopyrimidine sequences at the transcriptional
start site, it may be that p70S6K is required for main-
taining the apparatus required for ongoing protein
synthesis. This is consistent with previous suggestions
that physiological concentrations of insulin play a sig-
nificant role in maintaining the protein synthetic ap-
paratus (2). We have also found that infusion of IGF-I
increases the phosphorylation of both 4E-BP1 and
p70S6K (W. Long and E. J. Barrett, unpublished obser-
vation), again raising the issue of possible cross-stim-
ulation of IGF-I and/or IGF-I/insulin hybrid receptors
by high-dose insulin.
The cellular mechanisms underlying insulin’s anti-
proteolytic actions are poorly understood. Mounting
283 • DECEMBER 2002 • www.ajpendo.org
E1110 INVITED REVIEW
evidence suggests that the ATP-dependent ubiquitin-
proteasome proteolytic pathway plays a central role in
insulin-regulated protein degradation, particularly in
muscle (46). Blockade of lysosomal and calcium-acti-
vated proteases resulted in only minimal decrease in
proteolysis, whereas depletion of ATP or inhibition of
proteasomal activity with either MG-132 or lactacystin
blocked the increase in proteolysis due to insulinope-
nia. In insulin-deficient rats, increased mRNAs for
ubiquitin and proteasome subunits have also been
demonstrated (31, 34). However, insulin may also exert
a direct inhibitory effect on proteasomal peptidase ac-
tivity independent of an effect on transcription.
For insulin, there appears to be a dissociation be-
tween the doses of insulin that affect protein synthesis
vs. those that affect protein degradation. Whether we
consider whole body measures of synthesis and degra-
dation as estimated with labeled leucine or measures of
these processes in skeletal muscle with phenylalanine
tracer and the limb balance method, it appears that
proteolysis is more sensitive than synthesis to small
changes in plasma insulin within its physiological
range. Unknown is where in the insulin-signaling sys-
tem the pathway for regulation of synthesis and deg-
radation diverges. Dissecting this will require a more
detailed understanding of the early signaling events
that mediate insulin’s antiproteolytic action.
IGF-I. Like insulin, IGF-I has been repeatedly
shown to have anabolic properties on protein metabo-
lism (30). Systemically administered IGF-I inhibits
whole body protein breakdown and causes hypoamino-
acidemia and hypoglycemia. When given acutely in
vivo at doses that have no stimulatory effect on glucose
uptake, IGF-I stimulates muscle protein synthesis (17)
and is very effective in activating 4E-BP1 and p70S6K
phosphorylation in both skeletal and cardiac muscle
(W. Long and E. J. Barrett, unpublished observation).
In contrast, there is a lack of stimulation of whole body
protein synthesis with systemic infusion of IGF-I (16),
which likely results from the fall in circulating insulin
and/or amino acid concentrations that accompanies
systemic IGF-I infusion in vivo. When amino acids (50)
or insulin (25) is replaced simultaneously, IGF-I
clearly stimulates whole body protein synthesis. Mus-
cle protein breakdown seems to be inhibited by the
higher doses of IGF-I that also promote glucose uptake.
Inasmuch as insulin and IGF-I exert anabolic ac-
tions on both protein synthesis and proteolysis, the
question naturally arises as to whether there is cou-
pling between the processes of synthesis and degrada-
tion. In this regard, it is particularly intriguing to
examine the dose-response characteristics of each.
Thus modest increases in plasma insulin diminish
whole body and skeletal muscle proteolysis (35) and
increase glucose disposal in muscle, but higher doses
appear required to stimulate protein synthesis. In con-
trast, increases in plasma IGF-I that are without effect
on glucose uptake stimulate protein synthesis in mus-
cle and the whole body (17) and, at least in muscle,
higher doses are required to retard proteolysis. This
suggests that the cellular signals that mediate insulin
AJP-Endocrinol Metab • VOL
and IGF-I action in muscle, although similar, diverge
significantly, yet both can influence both the synthesis
and degradation pathways.
GH. Chronic GH treatment increases lean body mass
and decreases fat mass, especially when used to treat
antecedent GH deficiency (27, 51) and/or prevent glu-
cocorticoid-induced protein wasting (23). In healthy
humans, acute infusions of GH lasting from 6 h to 3–7
days modestly stimulate muscle (18) and whole body
protein synthesis without significantly affecting plasma
amino acid concentrations in healthy humans. The
mechanisms of GH-stimulated protein synthesis re-
main unclear. Systemic infusion not only increases
systemic IGF-I concentrations but also stimulates local
generation of IGF-I. GH may also exert its protein
anabolic actions directly, independently of IGF-I gen-
eration. GH has been shown to phosphorylate insulin
receptor substrate 1 (IRS-1), the principal substrate of
the insulin receptor, activate PI 3-kinase, and increase
the phosphorylation of p70S6K, actions similar to those
of insulin and IGF-I. When GH is infused locally into
the brachial artery, forearm muscle protein synthesis
increases, whereas systemic IGF-I concentrations and
muscle protein degradation are unaffected. This sug-
gests that GH-stimulated protein synthesis is not de-
pendent on increases in systemic IGF-I concentrations,
but the local generation of IGF-I in the skeletal muscle
remains a possible mechanism (30). Although GH in-
fusion provokes an increase in serum insulin concen-
tration, this does not retard protein degradation in
human muscle. Indeed, when GH and insulin are co-
infused, GH can block the effect of insulin to retard
proteolysis (19). Thus, for protein metabolism, both an
“insulin-like” and an insulin-antagonistic action of GH
can be seen when GH is given acutely to humans.
At the current time, it is difficult to construct an
integrated, mechanistic picture of GH’s action on hu-
man protein turnover. However, as additional tools
have become available for measuring cellular media-
tors of GH action, including the activity of the families
of the janus and signal transducer and activator of
transcription kinases, it will become increasingly pos-
sible to track responses in accessible tissues in humans
and animals that will allow a more detailed under-
standing of how GH affects protein synthesis and deg-
radation.
Amino acids. Ingestion of proteins or infusion of
mixed amino acids (AA) has significant protein ana-
bolic effects in healthy humans. At a systemic level, AA
administration stimulates insulin, GH, and glucagon
secretion. Independently of these hormonal effects, in
vitro AA directly stimulate protein synthesis (9, 53),
inhibit proteolysis (41), and enhance the sensitivity of
protein synthesis to insulin. Among all AA, branched-
chain amino acids (BCAA), especially leucine, have a
particularly important role in mediating these protein
anabolic effects (1). Recent data in cultured cell and
animal studies have demonstrated that both mixed AA
and BCAA can act as direct initiators of signal trans-
duction through mTOR-dependent phosphorylation of
4E-BP1 and p70S6K (Fig. 2), thereby enhancing protein
283 • DECEMBER 2002 • www.ajpendo.org
 INVITED REVIEW
synthesis at a translational level. BCAA or leucine
alone can activate the mTOR-4E-BP1/p70S6K pathway.
AA deprivation results in reversible inactivation of
p70S6K, increased binding of 4E-BP1 to eIF4E and
dissociation of the eIF4F complex, decreased phosphor-
ylation of eIF4E, and increased phosphorylation of
eukaryotic elongation factor 2 (eEF2). In aggregate,
these actions retard protein synthesis. AA also regu-
late global protein synthesis through modulation of
eIF2B activity. Despite these seemingly concordant
effects between AA mixture and BCAA on the activa-
tion of the mTOR-4E-BP1/p70S6K pathway, in vivo
infusions of AA mixtures and BCAA evoke quite differ-
ent patterns of anabolic response. Combining the fore-
arm arteriovenous tracer balance technique and mus-
cle biopsy, our laboratory has found that, in humans,
systemic infusion of a physiological amount of AA
increases protein synthesis with concurrent enhance-
ment of 4E-BP1 and p70S6K phosphorylation, without
affecting proteolysis (31a), whereas systemic infusion of
BCAA decreases proteolysis without affecting protein
synthesis, despite significant increase in 4E-BP1 and
p70S6K phosphorylation (32). The mechanisms under-
lying this difference are not yet clear. The lack of
anabolic action on protein synthesis during BCAA in-
fusion is possibly due to the decline in plasma (perhaps
intracellular) AA concentrations secondary to suppres-
sion of protein degradation, because the availability of
AA per se regulates protein synthesis. The mecha-
nism(s) by which BCAA retards proteolysis is poorly
understood. Moreover, why a complete mixture of AA
should be less effective in suppressing proteolysis than
BCAA alone is not known. AA mixtures have been
shown to modify, or even antagonize, insulin’s action at
sites early in the insulin-signaling cascade, including
the insulin receptor per se (45). Whether this might
account for the lack of antiproteolytic action of infused
AA requires further investigation. At this point, it
cannot be discounted that the tracer methods being
used to study the turnover of protein may be compro-
mised in some manner not currently apparent.
CONCLUSIONS AND FUTURE DIRECTIONS
We are early along in our understanding of how the
body regulates its protein mass. Major unanswered
questions abound. Some are basic to our understand-
ing of this homeostatic system. For example, why are
the turnover rates of proteins so high relative to those
of either carbohydrates or fats? Likewise, there is a
major question as to whether there is coupling between
the processes of protein synthesis and protein degra-
dation that needs to be more directly addressed. This
arises from observations that some hormones (e.g.,
insulin, IGF-I) have anabolic actions on both processes,
whereas others (glucocorticoids) have dual anti-ana-
bolic actions. Is there overlap in the signaling path-
ways mediating this? What are the early steps in the
pathway of insulin’s and IGF-I’s anti-proteolytic ef-
fects? Questions focusing more particularly on patho-
logical processes involved in body protein wasting also
AJP-Endocrinol Metab • VOL
E1111
warrant further investigation. Addressing some of
these questions will require new tools. These may
include newer generations of mass spectrometers that
will allow more facile measurement of the synthesis of
specific proteins. In addition, better probes for address-
ing the functional status of the ubiquitin-proteasomal
proteolytic pathways will need to be developed to ex-
amine the role of this potentially very important path-
way in the in vivo regulation of protein metabolism.
This work was supported by National Institutes of Health Grants
RR-15540 (Z. Liu), DK-38578, and DK-54058 (E. J. Barrett) and
RR-00847 (University of Virginia General Clinical Research Center).
REFERENCES
1. Anthony JC, Yoshizawa F, Anthony TG, Vary TC, Jeffer-
son LS, and Kimball SR. Leucine stimulates translation initi-
ation in skeletal muscle of postabsorptive rats via a rapamycin-
sensitive pathway. J Nutr 130: 2413–2419, 2000.
2. Ashford AJ and Pain VM. Effect of diabetes on the rates of
synthesis and degradation of ribosomes in rat muscle and liver in
vivo. J Biol Chem 261: 4059–4065, 1986.
3. Attaix D, Aurousseau E, Combaret L, Kee A, Larbaud D,
Ralliere C, Souweine B, Taillandier D, and Tilignac T.
Ubiquitin-proteasome-dependent proteolysis in skeletal muscle.
Reprod Nutr Dev 38: 153–165, 1998.
4. Barrett EJ and Gelfand RA. The in vivo study of cardiac and
skeletal muscle protein turnover. Diabetes Metab Rev 5: 133–
148, 1989.
5. Barrett EJ, Revkin JH, Young LH, Zaret BL, Jacob R, and
Gelfand RA. An isotopic method for in vivo measurement of
muscle protein synthesis and degradation. Biochem J 245: 223–
228, 1987.
6. Bier DM. Intrinsically difficult problems: the kinetics of body
proteins and amino acids in man. Diabetes Metab Rev 5: 111–
132, 1989.
7. Bilmazes C, Uauy R, Haverberg LN, Munro HN, and Young
VR. Muscle protein breakdown rates in humans based on N-
methylhistidine (3-methylhistidine) content of mixed proteins in
skeletal muscle and urinary output of N-methylhistidine. Metab-
olism 27: 525–530, 1978.
8. Cahill GF Jr. Starvation in man. N Engl J Med 282: 668–675,
1970.
9. Campbell LE, Wang X, and Proud CG. Nutrients differen-
tially regulate multiple translation factors and their control by
insulin. Biochem J 344: 433–441, 1999.
10. Charlton M and Nair KS. Protein metabolism in insulin-
dependent diabetes. J Nutr 128: S323–S327, 1998.
11. Charlton MR, Balagopal P, and Nair KS. Skeletal muscle
myosin heavy chain synthesis in type 1 diabetes. Diabetes 46:
1336–1340, 1997.
12. Cheng KN, Dworzak F, Ford GC, Rennie MJ, and Halliday
D. Direct determination of leucine metabolism and protein
breakdown in humans using L-[1–13C, 15N]-leucine and the
forearm model. Eur J Clin Invest 15: 349–354, 1985.
13. Cooney RN, Kimball SR, and Vary TC. Regulation of skeletal
muscle protein turnover during sepsis: mechanisms and media-
tors. Shock 7: 1–16, 1997.
14. Coppack S, Persson M, and Miles J. Phenylalanine kinetics
in human adipose tissue. J Clin Invest 98: 692–697, 1996.
15. De Feo P, Gaisano MG, and Haymond MW. Differential
effects of insulin deficiency on albumin and fibrinogen synthesis
in humans. J Clin Invest 88: 833–840, 1991.
16. Elahi D, McAloon-Dyke M, Fukagawa NK, Sclater AL,
Wong GA, Shannon RP, Minaker KL, Miles JM, Ruben-
stein AH, Vandepol CJ, Guler HP, Good WR, Seaman JJ,
and Wolfe RR. Effects of recombinant human IGF-I on glucose
and leucine kinetics in men. Am J Physiol Endocrinol Metab 265:
E831–E838, 1993.
17. Fryburg DA. Insulin-like growth factor I exerts growth hor-
mone- and insulin-like actions on human muscle protein metab-
olism. Am J Physiol Endocrinol Metab 267: E331–E336, 1994.
283 • DECEMBER 2002 • www.ajpendo.org
E1112 INVITED REVIEW
18. Fryburg DA, Gelfand RA, and Barrett EJ. Growth hormone
acutely stimulates muscle protein synthesis in normal humans.
Am J Physiol Endocrinol Metab 260: E499–E504, 1991.
19. Fryburg DA, Louard RJ, Gerow KE, Gelfand RA, and
Barrett EJ. Growth hormone stimulates skeletal muscle pro-
tein synthesis and blunts insulin’s antiproteolytic action in hu-
mans. Diabetes 41: 424–429, 1992.
20. Garlick PJ, McNurlan MA, and Preedy VR. A rapid and
convenient technique for measuring the rate of protein synthesis
in tissues by injection of [3H]phenylalanine. Biochem J 192:
719–723, 1980.
21. Gelfand RA and Barrett EJ. Effect of physiologic hyperinsu-
linemia on skeletal muscle protein synthesis and breakdown in
man. J Clin Invest 80: 1–6, 1987.
22. Hillier T, Long W, Jahn L, Wei L, and Barrett EJ. Physio-
logical hyperinsulinemia stimulates p70S6K phosphorylation in
human skeletal muscle. J Clin Endocrinol Metab 85: 4900–4904,
2000.
23. Horber FF and Haymond MW. Human growth hormone pre-
vents the protein catabolic side effects of prednisone in humans.
J Clin Invest 86: 265–272, 1990.
24. Horber FF, Horber-Feyder CM, Krayer S, Schwenk WF,
and Haymond MW. Plasma reciprocal pool specific activity
predicts that of intracellular free leucine for protein synthesis.
Am J Physiol Endocrinol Metab 257: E385–E399, 1989.
25. Jacob R, Hu X, Niederstock D, Hasan S, McNulty PH,
Sherwin RS, and Young LH. IGF-I stimulation of muscle
protein synthesis in the awake rat: permissive role of insulin and
amino acids. Am J Physiol Endocrinol Metab 270: E60–E66,
1996.
26. Jefferson LS, Rannels DE, Munger BL, and Morgan HE.
Insulin in the regulation of protein turnover in heart and skel-
etal muscle. Fed Proc 33: 1098–1104, 1974.
27. Jorgensen JOL, Pedersen SA, Thuesen L, Jorgensen J,
Ingemann-Hansen T, Skakkebaek NE, and Christiansen
JS. Beneficial effects of growth hormone treatment in GH-defi-
cient adults. Lancet 1: 1221–1225, 1989.
28. Kimball SR, Horetsky RL, and Jefferson LS. Signal trans-
duction pathways involved in the regulation of protein synthesis
by insulin by L6 myoblasts. Am J Physiol Cell Physiol 274:
C221–C228, 1998.
29. Kimball SR, Vary TC, and Jefferson LS. Regulation of pro-
tein synthesis by insulin. Ann Rev Physiol 56: 321–348, 1994.
30. Le Roith D, Bondy C, Yakar S, Liu JL, and Butler A. The
somatomedin hypothesis: 2001. Endocr Rev 22: 53–74, 2001.
31. Lecker SH, Solomon V, Price SR, Kwon YT, Mitch WE, and
Goldberg AL. Ubiquitin conjugation by the N-end rule pathway
and mRNAs for its components increase in muscles of diabetic
rats. J Clin Invest 104: 1411–1420, 1999.
31a.Liu Z, Jahn LA, Wei L, Long W, and Barrett EJ. Amino acids
stimulate translation initiation and protein synthesis through
an Akt-independent pathway in human skeletal muscle. J Clin
Endocrinol Metab. In press.
32. Liu Z, Jahn LA, Long W, Fryburg DA, Wei L, and Barrett
EJ. Branched chain amino acids activate messenger ribonucleic
acid translation regulatory proteins in human skeletal muscle,
and glucocorticoids blunt this action. J Clin Endocrinol Metab
86: 2136–2143, 2001.
33. Liu Z, Long W, Hillier T, Saffer L, and Barrett EJ. Insulin
regulation of protein metabolism in vivo. Diabetes Nutr Metab
12: 421–429, 1999.
34. Liu Z, Miers WR, Wei L, and Barrett EJ. The ubiquitin-
proteasome proteolytic pathway in heart vs skeletal muscle:
effects of acute diabetes. Biochem Biophys Res Commun 276:
1255–1260, 2000.
35. Louard RJ, Fryburg DA, Gelfand RA, and Barrett EJ.
Insulin sensitivity of protein and glucose metabolism in human
forearm skeletal muscle. J Clin Invest 90: 2348–2354, 1992.
36. Matthews DE, Motil KJ, Rohrbaugh DK, Burke JF, Young
VR, and Bier DM. Measurement of leucine metabolism in man
AJP-Endocrinol Metab • VOL 283 • DECEMBER 2002 •
from a primed, continuous infusion of L-[1-13C]leucine. Am J
Physiol Endocrinol Metab 238: E473–E479, 1980.
37. McNulty PH, Louard RJ, Deckelbaum LI, Zaret BL, and
Young LH. Hyperinsulinemia inhibits myocardial protein deg-
radation in patients with cardiovascular disease and insulin
resistance. Circulation 92: 2151–2156, 1995.
38. McNurlan MA, Essen P, Heys SD, Buchan V, Garlick PJ,
and Wernerman J. Measurement of protein synthesis in hu-
man skeletal muscle: further investigation of the flooding tech-
nique. Clin Sci (Colch) 81: 557–564, 1991.
39. Mendez R, Welsh G, Kleijn M, Myers MG, White MF, Proud
CG, and Rhoads RE. Regulation of protein synthesis by insulin
through IRS-1. Prog Mol Subcell Biol 26: 49–93, 2001.
40. Mitch WE, Bailey JL, Wang X, Jurkovitz C, Newby D, and
Price SR. Evaluation of signals activating ubiquitin-protea-
some proteolysis in a model of muscle wasting. Am J Physiol Cell
Physiol 276: C1132–C1138, 1999.
41. Mortimore GE, Khurana KK, and Miotto G. Amino acid
control of proteolysis in perfused livers of synchronously fed rats.
J Biol Chem 266: 1021–1028, 1991.
42. Nair KS, Ford GC, Ekberg K, Fernqvist-Forbes E, and
Wahren J. Protein dynamics in whole body and in splanchnic
and leg tissues in type I diabetic patients. J Clin Invest 95:
2926–2937, 1995.
43. Nair KS, Halliday D, and Griggs RC. Leucine incorporation
into mixed skeletal muscle protein in humans. Am J Physiol
Endocrinol Metab 254: E208–E213, 1988.
44. Owen OE, Felig P, Morgan AP, Wahren J, and Cahill GF
Jr. Liver and kidney metabolism during prolonged starvation.
J Clin Invest 48: 574–583, 1969.
45. Patti ME, Brambilla E, Luzi L, Landaker EJ, and Kahn
CR. Bidirectional modulation of insulin action by amino acids.
J Clin Invest 101: 1519–1529, 1998.
46. Pepato MT, Migliorini RH, Goldberg AL, and Kettelhut IC.
Role of different proteolytic pathways in degradation of muscle
protein from streptozotocin-diabetic rats. Am J Physiol Endocri-
nol Metab 271: E340–E347, 1996.
47. Proud CG and Denton RM. Molecular mechanisms for the
control of translation by insulin. Biochem J 328: 329–341, 1997.
48. Revkin JH, Young LH, Stirewalt WS, Dahl DM, Gelfand
RA, Zaret BL, and Barrett EJ. The in vivo measurement of
myocardial protein turnover using an indicator dilution tech-
nique. Circ Res 67: 902–912, 1990.
49. Rooyackers OE, Adey DB, Ades PA, and Nair KS. Effect of
age on in vivo rates of mitochondrial protein synthesis in human
skeletal muscle. Proc Natl Acad Sci USA 93: 15364–15369, 1996.
50. Russell-Jones DL, Umpleby AM, Hennessy TR, Bowes SB,
Shojaee-Moradie F, Hopkins KD, Jackson NC, Kelly JM,
Jones RH, and Sonksen PH. Use of a leucine clamp to dem-
onstrate that IGF-I actively stimulates protein synthesis in
normal humans. Am J Physiol Endocrinol Metab 267: E591–
E598, 1994.
51. Salomon F, Cuneo R, Hesp R, and Sonksen P. The effects of
treatment with recombinant human growth hormone on body
composition and metabolism in adults with growth hormone
deficiency. N Engl J Med 321: 1797–1803, 1989.
52. Schwenk WF, Beaufrere B, and Haymond MW. Use of re-
ciprocal pool specific activities to model leucine metabolism in
humans. Am J Physiol Endocrinol Metab 249: E646–E650, 1985.
53. Shigemitsu K, Tsujishita Y, Hara K, Nanahoshi M, Avruch
J, and Yonezawa K. Regulation of translational effectors by
amino acid and mammalian target of rapamycin signaling path-
ways. Possible involvement of autophagy in cultured hepatoma
cells. J Biol Chem 274: 1058–1065, 1999.
54. Tessari P, Trevisan R, Inchiostro S, Bioli G, Nosadini R,
De Kreutzenberg SV, Duner E, Tiengo A, and Crepaldi G.
Dose-response curves of effects of insulin on leucine kinetics in
humans. Am J Physiol Endocrinol Metab 251: E334–E342, 1986.
55. Wang X, Campbell LE, Miller CM, and Proud CG. Amino
acid availability regulates p70 S6 kinase and multiple transla-
tion factors. Biochem J 334: 261–267, 1998.
www.ajpendo.org