| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Comments |
Department of Internal Medicine, Division of Endocrinology, University of Virginia School of Medicine, Charlottesville, Virginia 22908
Address all correspondence and requests for reprints to: Eugene J. Barrett, M.D., Ph.D., Department of Internal Medicine, MR4-5116, University of Virginia Health Sciences Center, Charlottesville, Virginia 22908. E-mail: ejb8x{at}virginia.edu
Abstract
Using tracer methods, insulin stimulates muscle protein synthesis in vitro, an effect not seen in vivo with physiological insulin concentrations in adult animals or humans. To examine the action of physiological hyperinsulinemia on protein synthesis using a tracer-independent method in vivo and identify possible explanations for this discrepancy, we measured the phosphorylation of ribosomal protein S6 kinase (P70S6k) and eIF4E-binding protein (eIF4E-BP1), two key proteins that regulate messenger ribonucleic acid translation and protein synthesis. Postabsorptive healthy adults received either a 2-h insulin infusion (1 mU/min·kg; euglycemic insulin clamp; n = 6) or a 2-h saline infusion (n = 5). Vastus lateralis muscle was biopsied at baseline and at the end of the infusion period. Phosphorylation of P70S6k and eIF4E-BP1 was quantified on Western blots after SDS-PAGE. Physiological increments in plasma insulin (42 ± 13 to 366 ± 36 pmol/L; P = 0.0002) significantly increased p70S6k (P < 0.01), but did not affect eIF4E-BP1 phosphorylation in muscle. Plasma insulin declined slightly during saline infusion (P = 0.04), and there was no change in the phosphorylation of either p70S6k or eIF4E-BP1. These findings indicate an important role of physiological hyperinsulinemia in the regulation of p70S6k in human muscle. This finding is consistent with a potential role for insulin in regulating the synthesis of that subset of proteins involved in ribosomal function. The failure to enhance the phosphorylation of eIF4E-BP1 may in part explain the lack of a stimulatory effect of physiological hyperinsulinemia on bulk protein synthesis in skeletal muscle in vivo.
AN INTERESTING and unexplained dichotomy arises from studies examining insulin’s actions on muscle protein synthesis in vitro and in vivo. Abundant in vitro data clearly indicate that insulin strongly stimulates bulk cellular protein synthesis in a variety of cells and perfused organs (1, 2, 3) and that insulin acts primarily to enhance messenger ribonucleic acid (mRNA) translation (4, 5). However, using steady state tracer infusion methods (6, 7, 8, 9, 10, 11), the flooding-bolus tracer technique (12, 13), and more recently even measurements of aminoacyl transfer RNA labeling (14, 15), investigators have almost without exception (16) reported that insulin at physiological concentrations does not stimulate either whole body or muscle protein synthesis in vivo. This contrasts with the prominent synthetic effect seen in vivo when a balanced mixture of amino acids (6, 11), GH (17), insulin-like growth factor (11, 18), or pharmacological doses of insulin (19) are infused.
Recent data suggest that insulin stimulates mRNA translation in vitro at least in part by enhancing the activity of the translation initiation step that begins with eIF4E forming the preinitiation complex (20, 21). Insulin increases eIF4E availability by phosphorylating the 4E-binding protein (eIF4E-BP1), which then dissociates from 4E, thereby increasing the availability of the latter to form the eIF4F cap binding complex of mRNA. In addition to eIF4E-BP1, p70S6k is phosphorylated in a variety of tissues in response to insulin (22). Phosphorylation of p70S6k and subsequently ribosomal protein S6 regulates the synthesis of a selected subset of proteins that includes ribosomal proteins, translation initiation and elongation factors, and other proteins with oligopyrimidine sequences at the transcriptional start site (23, 24). Whether it has a major role in the translation of the bulk of cellular proteins is not clear.
In the current study we examined whether physiological hyperinsulinemia in vivo affects the phosphorylation state of these two proteins in an effort to use a nontracer based method to identify how physiological hyperinsulinemia regulates protein synthesis in vivo in human muscle.
Subjects and Methods
Subjects
Eleven (nine men and two women) healthy, normal weight (body mass index, 24 ± 1 kg/m2) adult (age, 30 ± 2 yr) volunteers were admitted to the University of Virginia General Clinical Research Center the evening before the study. No subject was taking any medication, and all female participants had a negative serum pregnancy test 1–2 days before the study. The study protocol was approved by the University of Virginia human investigation committee, and each subject gave written consent.
Experimental protocol
After an overnight (12-h) fast, an iv catheter was placed in an antecubital vein for infusion of either insulin and glucose or saline, and a second catheter was placed retrograde in a dorsal vein of the contralateral hand for sampling of glucose and insulin. Beginning 0.5 h before and throughout the infusion period the catheterized hand was placed in a warming box (55 C) to arterialize venous blood. Duplicate baseline samples of plasma glucose and insulin were obtained, and a baseline vastus lateralis muscle biopsy was taken (see muscle biopsy) from each subject. One group (n = 6) was given a continuous insulin infusion (1 mU/min·kg) for 120 min with a 20% dextrose solution infused at a variable rate to maintain plasma glucose within 5% of postabsorptive values (25). The control group (n = 5) received a 0.9% saline infusion for 120 min. A second muscle biopsy was obtained in all subjects in the contralateral leg during the last 30 min of the study (90–120 min).
Muscle biopsy
Before the muscle biopsy, the anterior thigh was shaved and
washed with iodine. The patient was prepped and draped in a sterile
fashion. Anesthesia was obtained with approximately 5 cc 1% xylocaine
in the overlying skin and muscle fascia of the vastus lateralis. Once
adequate anesthesia was obtained, the skin and underlying tissue was
incised (
5 mm long x 2 cm deep) with a no. 11 scalpel blade
and a Bergstrom biopsy needle (od, 4 mm; Popper and Sons, New Hyde
Park, NY) advanced through the incision tract into the vastus lateralis
muscle. Approximately 60 cc of suction were applied, and three muscle
samples were rapidly obtained by rotating the needle clockwise. The
biopsy needle was then removed, needle plus specimen were placed
directly into liquid N2 (<5 s), and the frozen
sample was removed and stored in liquid N2 until
analyzed. To minimize bruising, firm pressure was maintained at the
biopsy site for 10 min. The site was subsequently dressed, and 2–3 lb
of pressure were placed on the site for 30 min before wrapping with an
Ace bandage. Subjects were instructed to avoid vigorous leg
exercise for 48 h after the study.
Analytic methods
Plasma glucose was measured using the glucose oxidase method,
and insulin was determined using a double antibody RIA
(Diagnostic Products, Los Angeles, CA). For Western
blotting, pieces (20 mg) of frozen vastus lateralis muscle were
weighed and powdered in frozen 25 mmol/L Tris-HCl buffer (26 mmol/L
potassium fluoride and 5 mmol/L ethylenediamine tetraacetate, pH
7.5). Muscle was then disrupted by sonication using a microtip probe,
0.5 s on/0.5 s off, for 45 s at a 3.0 power setting on the
Fisher XL2020 sonicator (Fisher Scientific, Pittsburgh, PA). The
homogenate was centrifuged at 2000 rpm for 2 min, and the protein
concentration was measured in the supernatant. For
p70S6k, one aliquot of the supernatant (50 µg
protein) was diluted with an equal volume of SDS sample buffer and run
on an 8% SDS PAGE. For eIF4E-BP1, another aliquot of supernatant (60
µg protein) was diluted with an equal volume of SDS sample buffer and
electrophoresed on a 15% polyacrylamide gel. Proteins on both gels
were electrophoretically transferred to nitrocellulose membranes
(Schleicher & Schuell, Inc., Keene, NH). After blocking
with 5% low fat milk in TBS-T, membranes were incubated with
rabbit anti-p70S6k developed against the
C-terminus of rat p70s6k (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) or with rabbit antirat
eIF4E-BP1 (PHAS-I) developed against intact recombinant rat PHAS-I
(gift from Dr. J. Lawrence) for 1 h at room temperature. This was
followed by a donkey antirabbit IgG coupled to horseradish peroxidase,
and the blot was developed using ECL Western blotting kit
(Amersham Pharmacia Biotech, Piscataway, NJ).
Autoradiographic film was scanned densitometrically (Molecular Dynamics, Inc., Piscataway, NJ) and quantitated using ImageQuant
3.3. eIF4E-BP1, a 12-kDa protein, migrates in this system anomalously
at approximately 20 kDa. One to three bands were seen on Western
blotting that corresponded to the hyperphosphorylated and
hypophosphorylated forms of the protein from top to bottom
(26). Likewise, p70S6k in extracts
from unstimulated muscle migrated as a single band. However, with
successive phosphorylations electrophoretic mobility was retarded, and
several discrete bands were observed. This method of examining the
phosphorylation state of p70s6k and eIF4E-BP1 has been used
successfully by several investigators (20, 21) to track
the roles of these proteins in the stimulation of protein synthesis by
hormones or nutrients. For eIF4E-BP1 we quantified the ratio of the
most rapidly migrating (hypophosphorylated) band (
) to the total
immunoreactive material. The -band was selected because this form of
the protein binds to eIF4E and limits formation of the active
initiation complex. Conversely, for p70S6k we
quantified the ratio of the more heavily phosphorylated (more slowly
migrating forms) to the total immune reactivity, as it is the
hyperphosphorylated forms that possess kinase activity. Use of the
ratio among the phosphorylated species can facilitate quantitation by
minimizing the effects of variation in loading of the SDS gels. In
separate experiments we verified that loading 40–80 µg protein on
the 15% polyacrylamide gels and 60–120 µg on the 8% gels did not
affect the ratio of the several phosphorylated forms of eIF4E-BP1 to
p70S6k gels (data not shown).
Data presentation and statistical analysis
All data are presented as the mean ± SEM. Data for the average rate of glucose infusion (milligrams per kg/min) and for insulin concentrations are averaged over the final 60 min of the study period. Stochastic comparisons were made using ANOVA, and post-hoc comparison between groups was performed with two-tailed Student’s t test. Comparisons within groups were made with two-tailed paired t tests.
Results
Study subjects
The baseline characteristics of the two study groups as well as
their plasma insulin concentrations before and during the infusion
study are given in Table 1
. There were no
significant differences between the two study groups in mean age, body
mass index, or baseline insulin concentrations. For the insulin-infused
group, baseline glucose concentrations averaged 5.3 ± 0.1 mmol/L
and remained within 5% of the basal value over the last 60 min of the
study period (mean, 5.1 ± 0.1 mmol/L). The average rate of
glucose infused over the final hour of the study was 5.5 ± 0.9
mg/kg·min. Baseline insulin concentrations averaged 42 ± 13
pmol/L and rose significantly to average 366 ± 36 pmol/L over the
final hour of the study (P = 0.0002; see Table 1). In
the saline group, baseline insulin values averaged 39 ± 7 pmol/L
and decreased slightly to 28 ± 5 pmol/L by the final hour of the
study (P = 0.04).
|
Figure 1 illustrates typical
patterns for p70S6k observed on Western blots of
vastus lateralis muscle at 2 h of either insulin or saline
infusion. When insulin was infused, there were at least three bands
that reacted with the anti-p70S6k antibody. The
two uppermost bands represent the more highly phosphorylated forms of
p70S6k and generally correspond to species with
greater kinase activity. To quantify the extent of phosphorylation of
p70S6k, we measured the ratio of the intensity of
the more slowly migrating species (the top two bands) to the total
intensity. The mean densitometry results for
p70S6k (Fig. 1) indicated that physiologicsl
hyperinsulinemia had a highly significant effect to increase the
fraction of p70S6k that migrated more slowly
(P < 0.001, by ANOVA).
|
Figure 2 illustrates a typical
pattern of eIF4E-BP1 migration on Western blots of vastus lateralis
muscle obtained basally and during the last 30 min of the 2-h infusion
in each group. Two distinct bands were evident. To quantify the extent
of phosphorylation of eIF4E-BP1, we measured the ratio of the intensity
of the more rapidly migrating species to that of the total integrated
intensity. The rapidly migrating species represents the least
phosphorylated form of eIF4E-BP1 and is the form associated with eIF-4E
(26). A decline in the quantity of this form would
correspond to an increase in phosphorylation of eIF4E-BP1 and a greater
amount of eIF-4E available to initiate translation. In the blots shown
there was no change from basal in the fraction of eIF4E-BP1 that
migrates as a hypophosphorylated form with insulin infusion and no
clear difference between saline- and insulin-infused subjects. Mean
data obtained from the entire set of subjects are also shown in Fig. 2.
These data further indicated that physiological doses of insulin did
not affect the phosphorylation state of eIF4E-BP1 compared with the
baseline level of phosphorylation in the same subject or in the
saline-infused group.
|
Two hours of euglycemic hyperinsulinemia in young, healthy adults significantly increased the phosphorylation of p70S6k compared with the effect of saline, but did not affect the phosphorylation of eIF4E-BP1. These findings suggest that physiological regulation of p70S6k by insulin may be important for the effect of insulin to maintain the synthesis of a specific subset of proteins involved in mRNA translation. A 2-h infusion period was selected based on previous studies demonstrating that in serial samples of rectus muscle in rats receiving a 3-h high dose insulin infusion (10 mU/min·kg) eIF4E-BP1 and p70s6k were phosphorylated within 30–60 min, and this was maintained for up to 3 h. The lack of change in the phosphorylation of eIF4E-BP1 in the current study may in part explain the absence of a stimulatory effect of physiological hyperinsulinemia on bulk protein synthesis in skeletal muscle in vivo.
Phosphorylation of p70S6k correlates strongly with increases in kinase activity (22). Ribosomal S6 protein is a principal physiological substrate for p70S6k, and phosphorylation of S6 enhances the translation of a restricted subset of proteins with oligopyrimidine sequences at the transcriptional start site (24). This includes transcripts for many ribosomal proteins as well as translation initiation and elongation factors (24). In amino acid-starved cells p70S6k activity declines, and adding serum rapidly restores p70S6k activity (27). Animals with deletions of p70S6k are small, and expression of a highly homologous kinase (S6K2) is up-regulated in these animals (28). In intact rats, induction of diabetes leads to a rapid decline in polysome number in skeletal muscle as well as a decline in protein synthetic rates and ribosomal protein content (29). Insulin treatment reconstitutes ribosome and polysomal complement and synthetic activity (30). Thus, it appears that p70S6k is required for maintaining the apparatus required for ongoing protein synthesis. These previous results together with the current findings are consistent with the suggestion that physiological concentrations of insulin play a significant role in maintaining the protein synthetic apparatus.
The initiation factor eIF4E-BP1 acutely regulates protein synthesis by binding to eIF4E, thereby inhibiting its association with eIF-4G, which is needed for formation of the translation initiation complex (20). Phosphorylation of eIF4E-BP1 promotes its dissociation from 4E, which is then available for formation of the initiation complex. This appears a necessary step for the translation of mRNAs with m7GTP at the 5'-Cap. This includes the bulk of cellular protein (heat shock proteins and some viral proteins are notable exceptions) (31). In that sense, eIF4E-BP1 acts more directly than p70S6k at the translational level to regulate ongoing synthesis of a large number of proteins.
It is generally considered that p70S6k and eIF4E-BP1 are phosphorylated in parallel when the mTOR kinase cascade, which is part of the insulin signaling network that includes phosphoinositol 3-kinase, is activated (32). However, there are now several circumstances in which it has been possible to dissociate the activation of p70S6k and eIF4E-BP1. In HEK 293 cells constitutively active mutants of protein kinase B enhance phosphorylation of PHAS-I, but fail to activate p70S6k if they lack a membrane-targeting sequence (33). In studies in rat skeletal muscle, we have observed that pharmacological doses of insulin enhance the phosphorylation of both p70S6k and eIF4E-BP1, whereas physiological hyperinsulinemia differentially phosphorylated p70S6k, but not eIF4E-BP1 (34).
The lack of phosphorylation of eIF4E-BP1 with physiological concentrations of insulin in the current study may provide some explanation for the otherwise unexpected observation that insulin does not affect the rate of whole body (6, 9) or skeletal muscle protein synthesis in vivo in adult animals or humans (12, 13, 35, 36, 37, 38, 39, 40). In these in vivo studies, insulin concentrations are typically raised from basal to high physiological concentrations using either systemic or local (for skeletal muscle) insulin infusions. Most recently, using measurements of the specific activity or enrichment of aminoacyl transfer RNA, a very direct method for estimating protein synthesis, studies in rats (14) and humans (15) confirm that physiological hyperinsulinemia has no apparent effect on protein synthesis. These in vivo findings in mature animals and humans contrast with abundant excellent studies clearly demonstrating that insulin stimulates protein synthesis in perfused cardiac and skeletal muscle (2, 3, 4). However, insulin concentrations well above the physiological range (2–25 mU/mL) were typically used in those in vitro studies, and insulin-like growth factor I receptor-mediated effects cannot be excluded. Likewise, in two previous studies in which the phosphorylations of p70S6k and eIF4E-BP1 were measured after insulin treatment in vivo, supraphysiological insulin doses were used (41, 42). It is important to consider that even the physiological insulin infusion given in the current study decreases the circulating concentration of most amino acids. Amino acids themselves have a stimulatory effect on the phosphorylations of p70S6k and eIF4E-BP1 (27), and the decline in plasma amino acid or amino acid concentrations in muscle tissue may have blunted potential effects of insulin on eIf4E-BP1 phosphorylation. If this were a factor, it would suggest different sensitivities of p70S6k and 4E-BP1 to this action of amino acids.
In conclusion, we believe that the current results suggest an important physiological action of insulin to regulate the phosphorylation and thereby the activity of p70S6k. This action suggests that physiological doses of insulin will have a second order effect to promote muscle protein synthesis by preserving synthesis of ribosomal proteins and other factors involved in translation regulation. The loss of this effect of insulin could contribute to the net loss of ribosomes and polysomes seen in animal models of type 1 diabetes. Conversely, the lack of effect of insulin on eIF4E-BP1 phosphorylation may in part explain the absence of an acute first order effect of insulin to stimulate protein synthesis in vivo.
Footnotes
1 This work was supported by USPHS Grants DK-54058, DK-38578,
and RR-00847 to the University of Virginia General Clinical Research
Center. 
Received March 9, 2000.
Revised August 11, 2000.
Accepted August 25, 2000.
References
This article has been cited by other articles:
 |
|
 |
|
  P. L. Greenhaff, L. G. Karagounis, N. Peirce, E. J. Simpson, M. Hazell, R. Layfield, H. Wackerhage, K. Smith, P. Atherton, A. Selby, et al. Disassociation between the effects of amino acids and insulin on signaling, ubiquitin ligases, and protein turnover in human muscle Am J Physiol Endocrinol Metab, September 1, 2008; 295(3): E595 - E604. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Fujita, H. C. Dreyer, M. J. Drummond, E. L. Glynn, J. G. Cadenas, F. Yoshizawa, E. Volpi, and B. B. Rasmussen Nutrient signalling in the regulation of human muscle protein synthesis J. Physiol., July 15, 2007; 582(2): 813 - 823. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Krebs, B. Brunmair, A. Brehm, M. Artwohl, J. Szendroedi, P. Nowotny, E. Roth, C. Furnsinn, M. Promintzer, C. Anderwald, et al. The Mammalian Target of Rapamycin Pathway Regulates Nutrient-Sensitive Glucose Uptake in Man Diabetes, June 1, 2007; 56(6): 1600 - 1607. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Liu, G. Li, S. R. Kimball, L. A. Jahn, and E. J. Barrett Glucocorticoids modulate amino acid-induced translation initiation in human skeletal muscle Am J Physiol Endocrinol Metab, August 1, 2004; 287(2): E275 - E281. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Long, E. J. Barrett, L. Wei, and Z. Liu Adrenalectomy enhances the insulin sensitivity of muscle protein synthesis Am J Physiol Endocrinol Metab, January 1, 2003; 284(1): E102 - E109. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Liu and E. J. Barrett Human protein metabolism: its measurement and regulation Am J Physiol Endocrinol Metab, December 1, 2002; 283(6): E1105 - E1112. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. McDaniel, C. A. Marshall, K. L. Pappan, and G. Kwon Metabolic and Autocrine Regulation of the Mammalian Target of Rapamycin by Pancreatic {beta}-Cells Diabetes, October 1, 2002; 51(10): 2877 - 2885. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Greiwe, G. Kwon, M. L. McDaniel, and C. F. Semenkovich Leucine and insulin activate p70 S6 kinase through different pathways in human skeletal muscle Am J Physiol Endocrinol Metab, September 1, 2001; 281(3): E466 - E471. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Liu, L. A. Jahn, W. Long, D. A. Fryburg, L. Wei, and E. J. Barrett Branched Chain Amino Acids Activate Messenger Ribonucleic Acid Translation Regulatory Proteins in Human Skeletal Muscle, and Glucocorticoids Blunt This Action J. Clin. Endocrinol. Metab., May 1, 2001; 86(5): 2136 - 2143. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
)
forms as a fraction of the total S6 kinase.