Insulin Action on Gene Expression in Skeletal Muscle
J. Clin. Invest.
© The American Society for Clinical Investigation, Inc.
Volume 98, Number 1, July 1996, 43–49
Acute Regulation by Insulin of Phosphatidylinositol-3-kinase, Rad, Glut 4, and
Lipoprotein Lipase mRNA Levels in Human Muscle
M. Laville, D. Auboeuf, Y. Khalfallah, N. Vega, J.P. Riou, and H. Vidal
Institut National de la Santé et de la Recherche Médicale U 449, Faculté de Médecine Alexis Carrel, 69373 Lyon Cedex 08, and Centre de
Recherche en Nutrition Humaine Hsp. E. Herriot 69003 Lyon, France
We have investigated the acute regulation by insulin of the
mRNA levels of nine genes involved in insulin action, in
muscle biopsies obtained before and at the end of a 3-h eugly-
cemic hyperinsulinemic clamp. Using reverse transcription-
competitive PCR, we have measured the mRNAs encoding
the two insulin receptor variants, the insulin receptor sub-
strate-1, the p85
subunit of phosphatidylinositol-3-kinase,
Ras associated to diabetes (Rad), the glucose transporter
Glut 4, glycogen synthase, 6-phosphofructo-1-kinase, lipo-
protein lipase, and the hormone-sensitive lipase. Insulin in-
fusion induced a significant increase in the mRNA level of
Glut 4 (
13%), Rad (
in the lipoprotein lipase mRNA level (
abundance of the other mRNAs was unaffected. The rela-
tive expression of the two insulin receptor variants was not
modified. These results demonstrate an acute coordinated
regulation by insulin of the expression of genes coding key
proteins involved in its action in human skeletal muscle and
suggest that Rad and the p85
phatidylinositol-3-kinase can be added to the list of the
genes controlled by insulin. (
J. Clin. Invest.
Key words: reverse transcription–competitive polymerase
25%), the p85
18%) and a decrease
5%), while the
regulatory subunit of phos-
A major action of insulin is the regulation of metabolic path-
ways of protein, glycogen, and fat synthesis. These anabolic ef-
fects are initiated by the stimulation of the insulin receptor ty-
rosine kinase after insulin binding to cell membranes and are
dependent on the regulation of key controlling enzymes (1).
The action of insulin involves modifications of the activity of
these critical enzymes and changes in their expression. Insulin
can control specific protein amount, in part by acting at the
level of mRNA translation, and mainly at the level of their
gene expression (2). This last action is certainly a major effect
of insulin, and the list of insulin-regulated genes is rapidly
Skeletal muscle is the main site for insulin-dependent glu-
cose disposal in humans (5). Insulin stimulates glucose uptake
and glucose use in oxidative and storage pathways. While insu-
lin action on the activity of the main enzymes that control me-
tabolism in muscle has been well investigated, its role on the
expression of genes involved in insulin metabolic action has re-
ceived little attention. Moreover, when gene expression has
been investigated in vivo, the studies generally involve dia-
betic, insulin-treated, or diet-controlled patients, and therefore
brought only indirect evidence on the role of insulin. In pathol-
ogies with insulin resistance, the expression of only a few genes
has been found altered in skeletal muscle. The expression of
the protein Ras associated to diabetes (Rad)
(6) and the expression of hexokinase II (7) and glycogen syn-
thase (GS) (8) was decreased in patients with non–insulin-
dependent diabetes mellitus (NIDDM). On the other hand,
the expression of the genes coding Glut 4 (9) and 6-phospho-
fructo-1-kinase (PFK-1) (8) seemed unaltered in obese or
NIDDM patients, and intensive insulin treatment did not mod-
ify the expression of the genes for GS and PFK-1 in type I dia-
betic subjects (10). Using the euglycemic hyperinsulinic glucose
clamp, a method that allows the individual effect of insulin to
be studied, it was found that Glut 4, GS, and hexokinase I
mRNA levels were unaltered in normal muscle; whereas, hex-
okinase II mRNA increased three times in response to insulin
(11). At higher insulin levels, other workers found that Glut 4
mRNA increased in skeletal muscle of normal subjects, but
not in muscle from type I or type II diabetic patients (12, 13).
All these data suggest that some important genes involved
in glucose metabolism could be regulated by insulin in human
muscle. Until now, the study of the specific effects of insulin on
the coordinated expression of genes has been limited by the
difficulty of measuring the abundance of several mRNAs in
percutaneous muscle biopsies due to the small amount of ma-
Address correspondence to M. Laville, INSERM U 449, Faculté de
Médecine Alexis Carrel, Rue G. Paradin, 69373 Lyon Cedex 08,
France. Phone: 33 78 77 86 29; FAX: 33 78 77 87 62.
Received for publication 6 February 1996 and accepted in revised
form 16 April 1996.
GS, glycogen synthase; HSL, hormone-sensitive lipase; IR Ex11, in-
sulin receptor mRNA variant with exon 11; IRS-1, insulin receptor
substrate-1; LPL, lipoprotein lipase; NIDDM, non–insulin-dependent
diabetes mellitus; PFK-1, 6-phosphofructo-1-kinase; PFK-2, 6-phos-
phofructokinase-2-kinase/fructose-2,6-bisphosphatase; PI-3K, phos-
phatidylinositol-3-kinase; Rad, Ras associated to diabetes; RT, re-
Abbreviations used in this paper:
M. Laville, D. Auboeuf, Y. Khalfallah, N. Vega, J.P. Riou, and H. Vidal
terial. Using a sensitive and quantitative method of reverse
transcription (RT) reaction, followed by competitive poly-
merase chain reaction (RT-competitive PCR), we investigated
the acute regulation by insulin of the mRNA level of nine key
genes involved in the metabolic pathways and in insulin action.
The mRNA abundances of the target genes were determined
in normal human muscle biopsies, obtained before and at the
end of a 3-h euglycemic hyperinsulinic clamp. We have devel-
oped a multispecific internal standard (14, 15) that allowed us
to measure, by RT-competitive PCR, the levels of the mRNAs
encoding the insulin receptors (total mRNA and mRNA vari-
ant with exon 11), the insulin receptor substrate-1 (IRS-1), the
regulatory subunit of phosphatidylinositol-3-kinase (PI-
3K), Glut 4, Rad, GS, PFK-1, lipoprotein lipase (LPL), hor-
mone-sensitive lipase (HSL), and a reference mRNA encoding
the constitutively-expressed beta-2-microglobulin (
protein. The choice of these genes was made according to the
role of their respective products in the metabolic action of in-
sulin, but was limited by the availability of the human cDNA
sequences at the time we constructed the multispecific stan-
dard. Using this approach, we have revealed the expression
profile of the nine target genes and investigated their acute co-
ordinate regulation by insulin in normal human muscle.
Ten lean healthy volunteers (six men/four women) with a mean
SEM) age of 24
1 yr and a body mass index of 21.4
participated in the study. None of the subjects had a familial or per-
sonal history of diabetes, obesity, dyslipidemia, or hypertension or
was taking any medication except for oral contraceptives. They were
on their usual diet before the study and none were engaged in heavy
exercise. All subjects gave their written consent after being informed
of the nature, purpose, and possible risks of the study. The experi-
mental protocol was approved by the ethics committee of Hospices
Civils de Lyon and performed according to the French legislation
All studies were performed in the postabsorptive state at least 12
hours after the last meal and initiated between 0800 and 0900 hours,
after at least 30 min of bed rest. Intravenous catheters were inserted
into veins of one forearm for insulin and glucose infusions. To obtain
arterialized blood samples, another catheter was inserted, in a retro-
grade manner, in a vein of the opposite hand kept at 55
A percutaneous biopsy of the vastus lateralis
muscle was obtained with Weil Blakesley pliers. Approximately 5 ml
of xylocaïne (1%) without epinephrine was injected into the skin and
superficial tissue before the biopsy. The procedure involved a small
incision (5 mm) through the skin and muscle sheath 15–20 cm above
the knee. Muscle biopsies were taken at the beginning of the experi-
ment and after 3 h of euglycemic hyperinsulinemic clamp. The clamp
study began 2 h after the first biopsy to avoid the metabolic effect of
biopsy-related stress. Average muscle samples of 57
(range 35–87) were obtained, immediately frozen in liquid nitrogen,
and stored at
C until analyzed.
Euglycemic hyperinsulinemic clamp.
Copenhagen, Denmark) was infused at 12 pmol/kg per min for 3 h.
The insulin infusion rate was primed according to Rizza et al. (16).
Any decrease in blood glucose was prevented by the infusion of 20%
glucose (Aguettant, Lyon, France) as described previously (17). After
3 h of euglycemic clamp, a second muscle biopsy was obtained from
the opposite thigh. At this time, the insulin infusion was discontinued
and the study ended.
C in a warm-
4 mg wet wt
Insulin (Actrapid; Novo,
Muscle samples were pulverized in liquid nitrogen and total RNA
was isolated using guanidium thiocyanate, phenol-chloroform extrac-
tion, and alcohol precipitation (18). The average yield of total RNA
Table I. Oligonucleotides Used to Quantify the Target mRNAs by RT-competitive PCR
Size of PCR product
mRNA speciesSense primers Antisense primers mRNA Competitor
The primer sequences were selected as indicated in Methods. Glut 4, insulin-responsive glucose transporter (23); PFK-2, 6-phosphofructo-2-kinase
(24); Rad, protein Ras associated to diabetes (6); GS, glycogen synthase (25); IRS-1, insulin receptor substrate-1 (22); HSL, hormone-sensitive lipase
(26); PFK-1, 6-phosphofructo-1-kinase (27);
2-microglobulin (28); PI-3K, p85
BZR, peripheral benzodiazepine receptor (30); IR EX11, insulin receptor variant with exon 11 (31), total IR, total insulin receptor (31); LPL, lipopro-
tein lipase (32). The quantification of PFK-2 mRNA could not be performed because of a mutation in the sequence of the multispecific competitor.
The reference BZR mRNA was not amplified efficiently from human muscle total RNA preparations and thus was not measured in the present
regulatory subunit of phosphatidylinositol-3-kinase (29);
Insulin Action on Gene Expression in Skeletal Muscle
was 0.27?0.03 and 0.31?0.02 ?g/mg muscle (wet wt) for the biopsies
taken before and at the end of the hyperinsulinemic clamp, respec-
tively (no significant difference). The absorption ratio (260:280 nm) was
between 1.7 and 2.0 for all preparations and the integrity of the RNA
was verified on agarose gel colored with ethidium bromide. Total
RNA was stored at ?80?C until quantification of the target mRNAs.
Quantification of mRNAs
Specific mRNAs were quantified by reverse transcription reaction
followed by RT-competitive PCR, which consists of the coamplifica-
tion of a known amount of standard DNA with the target cDNA in
the same tube. The standard is designed to use the same PCR primers
as the target, but yields a PCR product of different size, so that the
two amplicons can be separated by gel electrophoresis and quantified
Selection of the primers and construction of the multispecific stan-
dard. The standard, or competitor, was a 525-bp long synthetic gene,
the sequence of which corresponded to the juxtaposition of 13 spe-
cific sense-primer sequences (sense-primer box), followed by the jux-
taposition of the complementary sequences of 12 specific antisense
primers in the same order (antisense primer box) (Fig. 1, Table I).
Total and Ex11 insulin receptor mRNAs were measured with the
same antisense primer (located in exon 13) and with specific sense
primers (located in exons 12 and 11). The multispecific standard was
constructed by a technique of oligonucleotide overlap extension and
amplification by PCR (21), starting with four purified long oligonu-
cleotides (139 to 151 bases; Eurogentec, Seraing, Belgium) that suc-
cessively covered all the sequence of the standard, with small over-
laps. A base substitution, at position ? 309, appeared during the
synthesis of the competitor so that we could not perform the compet-
itive PCR assay of the 6-phosphofructokinase-2-kinase/fructose-2,6-
bisphosphatase (PFK-2) mRNA. The phagemid containing the com-
petitor cDNA was purified in large amount, quantified by absorption
at 260 nm, and stored at ?20?C as concentrated stock solution. Work-
ing solutions at defined concentrations (25 amol/?l to 10?3 amol/?l)
were prepared by serial dilution in 10 mM Tris-HCl (pH 8.3), and 1
mM EDTA buffer, aliquoted under small volume and stored at
Reverse transcription reaction. For each mRNA, a specific first-
strand cDNA synthesis was performed from 0.5 to 1 ?g of total RNA
with 2.5 U of thermostable reverse transcriptase (Tth DNA poly-
merase; Promega Corp., Charbonnière, France) in 10 mM Tris-HCl,
pH 8.3, 90 mM KCl, 1 mM MnCl2, 0.2 mM deoxynucleoside triphos-
phates, and 15 pmol of the specific antisense primer, in a final volume
of 20 ?l. The medium was overlaid with mineral oil and subjected to
incubations for 3 min at 60?C followed by 15 min at 70?C in the ther-
mal cycler (Omnigene; Hybaid, Teddington, UK). The reaction was
stopped by heating at 99?C for 5 min. After chilling on ice, 4 ?l of wa-
ter was added to the RT medium, from which 20 ?l was sampled for
cDNA quantification by PCR.
Competitive PCR. The 20-?l sample of the RT reaction was
added to 80 ?l of PCR mix (10 mM Tris–HCl, pH 8.3, 100 mM KCl,
0.75 mM EGTA, 5% glycerol) containing 0.2 mM deoxynucleoside
triphosphates, 5 U of Taq polymerase (Life Technologies, Cergy Pon-
toise, France), 45 pmol of the corresponding sense primer, and 30
pmol of the antisense primer. Four 20-?l aliquots were then trans-
ferred to microtubes containing 5 ?l of defined working solutions of
the competitor cDNA. The microtubes were kept in ice until the ther-
mal cycler was up to 90?C. After 120 s at 94?C, the PCR mixtures
were subjected to 40 cycles of PCR amplification with a cycle profile
including denaturation for 40 s at 95?C, hybridation for 50 s at 55?C,
and elongation for 50 s at 72?C.
Analysis of PCR products. The amplification products of 15 ?l
of each PCR were separated in a 3 or 4% agarose gel, stained with
ethidium bromide, and photographed (665 film; Polaroid Corp.,
Cambridge, MA). The band densities were evaluated from the nega-
tive film with a Vernon photometer-integrator. After correction for
the difference in nucleotide number, the logarithm of the density ra-
tio of the competitor band to the target mRNA band was plotted vs
the logarithm of the initial amount of competitor cDNA. The initial
concentration of target cDNA in the PCR was determined at the
competition equivalence point as previously described (20).
Preparation of RNA by in vitro translation. Parts of the coding
sequence of Glut 4, the insulin receptor, and ?-?glob mRNAs, which
included the sequences targeted in the RT-competitive PCR assay,
were amplified by RT-PCR from a total RNA preparation obtained
from a normal human muscle biopsy, and subcloned in the phagemid
pGEM-T (Promega Corp.). The three partial cDNAs were used to
produce RNA by in vitro translation (Riboprobe System, Promega
Corp.) and the synthetic RNAs were quantified by RT-competitive
PCR as describe above.
Control experiments and the presentation of the results. The absence
of contamination was verified periodically by control experiments
that omitted RNA or cDNA in the reactions. For the mRNA of IRS-1,
the coding sequence is contained within a single exon (22) and, in the
cases of PI-3K and Rad mRNAs, the organization of the genes was
Figure 1. Organization of the multispe-
cific internal standard. The construction
of the multispecific standard is explained
in Methods. The 525-bp-long standard
was subcloned in the pBluescript KS?
phagemid. The T3 RNA polymerase pro-
moter was located upstream of the sense
primer box that corresponded to the jux-
taposition of the 13 specific-sense primer
sequences and was followed by the anti-
sense primer box with the complementary
sequences of the 12 antisense primers in
the same order. The abbreviations used
are defined in Table I.
M. Laville, D. Auboeuf, Y. Khalfallah, N. Vega, J.P. Riou, and H. Vidal
not reported when we selected the primer sequences, so that they
could be located in the same exon. Therefore, when these mRNAs
were measured, we checked for the absence of genomic DNA ampli-
fication by performing control RT-PCR without the reverse tran-
scriptase in the RT step.
To accurately compare the target mRNA abundance between
samples, total RNA from the two biopsies from the same subject (be-
fore and at the end of the clamp) were prepared at the same time. In
addition, the different target mRNAs were always measured in paral-
lel in the same run of PCR and with amounts of competitor taken
from the same serial dilution so that the level of the different mRNAs
could be compared. The results were then normalized and presented
by reference to the mRNA level of the constitutively expressed ?-?glob
gene (15, 22). This presentation had the advantage of erasing errors
in total RNA measurement between samples.
The results are expressed as means?SEM. Statistical compari-
sons were performed with a Wilcoxon nonparametric test for paired
values. The threshold for significance was set at P ? 0.05.
Validation of the RT-competitive PCR assay. To validate the
RT-competitive PCR assay, in vitro–synthesized RNAs (see
Methods) were quantified in five mixtures that contained a
constant amount of ?-?glob RNA and different amounts of
Glut 4 RNA (ranging from 0.1- to 10-fold the amount of
?-?glob RNA) and insulin receptor RNA (ranging from 0.02-
to 2-fold the amount of ?-?glob RNA). Each mixture was pre-
pared from the stock solutions of the synthetic RNAs and di-
luted to achieve a concentration of about 20 amol/?l (20 ?
10?18 mol) of ?-?glob RNA. The amount of ?-?glob RNA was
found to be 121.8?8.6 amol (mean?SD) in 5 ?l of the differ-
ent mixtures. Fig. 2 shows the RNA ratio (Glut 4 or insulin re-
ceptor RNA/?-?glob RNA) measured by RT-competitive
PCR plotted against the initial RNA ratio in the mixtures. The
slope of the standard curve was 1.18 for the ratio Glut 4/?-
?glob (r ? 0.997) and 1.05 for the ratio insulin receptor/?-
?glob (r ? 0.999). In addition, since the in vitro–synthesized
insulin-receptor RNA contained the sequence of the exon 11,
it could be measured with the primers specific for the insulin
receptor mRNA variant with exon 11 (IR EX11). Similar re-
sults were obtained when the IR EX11 or the total IR primers
were used (data not shown) and the slope of the standard
curve for the ratio IR Ex11/?-?glob was 0.95 (r ? 0.999). The
standard curve (Fig. 2) shows that the RT-competitive PCR as-
say was linear from 2 to 1,000 amol of RNA in the RT reac-
Euglycemic hyperinsulinemic clamp. Ten normal subjects
were submitted to a 3-h euglycemic hyperinsulinemic clamp to
achieve supraphysiological plasma insulin concentrations. In-
sulinemia increased from 42?3 pmol/liter in the basal state, to
816?28 pmol/liter during the last hour of insulin infusion. The
fasting plasma glucose concentration was 4.4?0.1 mmol/liter
and glycemia was clamped during the insulin infusion at
4.3?0.2 mmol/liter. The rate of glucose infusion required to
maintain euglycemia during the last hour of insulin infusion
was 51.1?2.7 mmol/kg per min.
Effect of insulin infusion on the abundance of the target
mRNAs in muscle. Fig. 3 shows the mRNA level profile of
the target genes in normal skeletal muscle, before and at the
end of the hyperinsulinemic clamp. The results were expressed
by reference to the level of ?-?glob mRNA. The absolute val-
ues of ?-?glob mRNA were 130?10 amol/?g of total RNA in
the basal state and 123?8 amol/?g of total RNA at the end of
the clamp (no significant difference). Before the clamp, the
abundance of the different target mRNAs varied in a broad
range from 2.4?0.5 (HSL) to 93.4?6.1% (Glut 4) of ?-?glob
Figure 2. Validation of the RT-competitive PCR. In vitro–produced
RNAs containing part of the coding sequence of ?-?glob, Glut 4, and
insulin receptor were measured in solutions containing a constant
amount of ?-?glob RNA and different amounts of Glut 4 and insulin
receptor RNAs. The results were expressed as ratio Glut 4 (open cir-
cles) or insulin receptor (filled circles) RNA/?-?glob RNA. The ini-
tial ratio Glut4 RNA/?-?glob RNA in the solutions were 0.1, 0.2, 1, 5,
and 10 and the initial ratio insulin receptor RNA/?-?glob RNA were
0.02, 0.04, 0.2, 1, and 2.
Figure 3. Profiles of the mRNA levels of the nine target genes in hu-
man muscle before and at the end of the euglycemic hyperinsulinemic
clamp. Specific mRNA levels were determined by RT-competitive
PCR and expressed as a percentage of the ?-?glob mRNA level. The
values measured before the euglycemic hyperinsulinemic clamp are
represented by the open boxes, while the hatched boxes corre-
sponded to the values obtained at the end of the 3-h clamp. Data are
means?SEM (n ? 10), *P ? 0.05 using the nonparametric test of
Wilcoxon for paired series.
Insulin Action on Gene Expression in Skeletal Muscle
mRNA. With the exception of the mRNA for HSL, which is
not a typical muscle enzyme, the least abundant mRNA en-
coded the insulin receptor (3.2?0.4% of ?-?glob mRNA). The
amount of the mRNA variant with exon 11 was 1.3?0.2% of
?-?glob mRNA, thus representing 43?5% of the total insulin
receptor mRNA. Rad mRNA was in the same range as the
mRNAs for Glut 4 and GS.
After 3 h of hyperinsulinemia, the mRNAs encoding Glut
4, Rad, and PI-3K significantly increased; whereas, LPL
mRNA decreased (Fig. 3). No change was observed in the
level of the mRNAs for GS, PFK-1, IRS-1, insulin receptor,
and HSL (Fig. 3). Fig. 4 shows the individual variations of the
mRNA levels. Glut 4 mRNA increased by 56?13% (P ?
0.007), Rad mRNA by 96?25% (P ? 0.093) and PI-3K
mRNA by 92?18% (P ? 0.005), while LPL mRNA decreased
by 49?5% (P ? 0.011). Insulin infusion affected neither the
amount (1.4?0.4% of ?-?glob mRNA), nor the percentage
(40?5% of total insulin-receptor mRNA) of the mRNA en-
coding the isoform of insulin receptor with exon 11. One sub-
ject had no response to insulin on both Glut 4 and PI-3K
mRNA levels (Fig. 4). The two subjects with the highest level
of Rad mRNA and no response to insulin were from the same
family (brother and sister). None of these subjects presented
abnormal glucose metabolism during the clamp.
Investigation of the coordinated regulation of a number of tar-
get genes by insulin in human muscle was made possible by the
development of a sensitive and reliable RT-competitive PCR
method. In this study, we report the expression profile of sev-
eral of the key genes that code proteins involved in the insulin-
sensitive metabolic pathways and study their acute regulation
by insulin. We demonstrate that insulin infusion changes the
mRNA level of Glut 4, Rad, PI-3K, and LPL in normal human
skeletal muscle, without affecting the level of GS, PFK-1, IRS-1,
insulin receptor, or HSL mRNAs.
The RT-competitive PCR developed in this study is a pow-
erful method of determining the absolute quantity of a given
mRNA. However, due to possible imprecisions in total RNA
concentration determinations and in the dilutions of the inter-
nal standard molecule, the results were presented by reference
to the level of a reference mRNA that was measured using the
same method (14, 15). We have validated this procedure using
RNAs synthesized by in vitro translation. Nevertheless, the ab-
solute values found by RT-competitive PCR in this study were
in accordance with the mRNA levels already reported using
other methods. We found that Glut 4 mRNA was the most
abundant of the selected targets and varied from 77 to 146
amol/?g total RNA (72 to 115% of ?-?glob mRNA), values
that were similar to those reported by Schalin-Jäntti et al. (13)
using a quantitative dot blot method (68 to 100 pg/?g total
RNA corresponding to 78 to 115 amol/?g total RNA, taking
3.5 kb for Glut 4 mRNA, reference 23). Similarly, the level of
insulin-receptor mRNA determined by solution hybridization
method (33) (? 2.5 amol/?g of total RNA, assuming that one
copy corresponds to one molecule of mRNA and using the
Avogadro constant) was in agreement with the values ob-
tained by RT-competitive PCR in this study (3.2?0.4% of the
?-?glob mRNA level that corresponded to values between 1.6
and 6.3 amol/?g of total RNA). Finally, we found that IRS-1
mRNA level was low in human muscle, but more abundant
than the insulin-receptor mRNA, in agreement with a recent
report using a semiquantitative PCR method (22). To our
knowledge, quantitative data are not available for the other
mRNAs. The use of a multispecific standard in the RT-com-
petitive PCR assay allowed us to compare the levels of ten dif-
ferent mRNAs in a single muscle biopsy and thus to propose
an expression profile of the mRNAs of nine genes involved in
insulin action. This profile reveals large differences in the basal
state, suggesting differences in gene expression. However, the
variety of the mRNA levels could also be attributed to differ-
ences in mRNA stability.
Insulin acutely increased the mRNA level of Glut 4, Rad,
and PI-3K and decreased the level of LPL mRNA. The most
striking effects were the increases in Rad and PI-3K mRNAs.
The Rad mRNA encodes a recently discovered protein related
to the Ras/GTPase superfamily and it has been shown to be
overexpressed in the muscle of NIDDM patients (6). Our find-
ing that insulin induced a twofold increase in Rad mRNA level
brings the first evidence that Rad expression is acutely regu-
lated in vivo and argues for a role for Rad in the mechanism of
action of insulin. It could also suggest that the overexpression
of Rad in NIDDM muscle might be related to chronic hyperin-
sulinemia. One of the earliest postreceptor events in insulin
signaling on glucose metabolism is the activation of phosphati-
dylinositol-3-kinase (1). This activation results from the bind-
Figure 4. Effect of insulin infusion on mRNA levels of Glut 4,
Rad, glycogen synthase, and p85? subunit of phosphatidylinositol-3-
kinase. The mRNA levels, expressed as a percentage of ?-?glob
mRNA, were measured before (open circles) and at the end (filled
circles) of the euglycemic hyperinsulinemic clamp in vastus lateralis
muscle biopsies of the ten subjects. Individual variations of PFK-1, of
the two variants of insulin receptor, and of HSL mRNA levels that
did not change significantly during the clamp (Fig. 3), were not pre-
M. Laville, D. Auboeuf, Y. Khalfallah, N. Vega, J.P. Riou, and H. Vidal
ing of the p85?-regulatory subunit of phosphatidylinositol-3-
kinase to specific phosphorylated tyrosine residues of IRS-1
(1). We show in this study that the mRNA level of the phos-
phatidylinositol-3-kinase regulatory subunit was increased by
insulin, thus demonstrating that insulin acts not only on the ac-
tivity of the enzyme, but also participates in the control of its
Direct effects of insulin on human muscle Glut 4 mRNA
have already been investigated. With an insulinemia main-
tained at around 390 pmol/liter, Mandarino et al. (11) did not
observe change in Glut 4 mRNA level. However, using supra-
physiological insulin concentrations (? 750 pmol/liter), other
investigators found a significant (35 to 50%) increase in Glut 4
mRNA (12, 13). This discrepancy probably results from the
differences in insulin levels, since in our study with insulinemia
maintained at 816?28 pmol/liter, Glut 4 mRNA increased by
56% at the end of the 3-h hyperinsulinemic clamp. For LPL,
insulin infusion has been shown to decrease its enzyme activity
by 35% during a 2-h clamp (34), which could be attributed to
an insulin-induced inhibition of LPL expression. Indeed, we
did find a decrease in LPL mRNA level in muscle under our
conditions of hyperinsulinemia.
The abundance of the mRNAs for GS, PFK-1, IRS-1, and
insulin receptor was not significantly affected during the
clamp. The absence of alterations of GS and PFK-1 mRNAs in
the presence of physiological levels of hyperinsulinemia has al-
ready been reported (8, 10, 11). The lack of change in the pres-
ence of high insulin concentrations suggests that the expres-
sion of these genes is not regulated by insulin. It cannot be
excluded, however, that insulin effects on the expression of
these genes require a longer time to become apparent. Alter-
native splicing of insulin receptor mRNA, leading to two re-
ceptor isoforms differing by the presence or the absence of 12
amino acids encoded by the exon 11, has been proposed to be
regulated by insulin in vivo (35) and in vitro (36). Our results
show that neither the total amount of insulin receptor nor the
relative expression of the mRNA variants encoding the two
isoforms varied during the hyperinsulinemic clamp. This is in
agreement with our recent report suggesting that insulin is not
the principal regulatory factor of insulin receptor mRNA splic-
ing in rat tissues in vivo (20).
Recent reviews dealing with insulin action in liver pointed
out that insulin effects on gene expression parallel its meta-
bolic actions. Insulin produces a coordinate stimulation of the
expression of genes coding key glycolytic enzymes and inhibi-
tion of the expression of genes coding key gluconeogenic en-
zymes (3, 37). Similarly, in the muscle, insulin increases Glut 4,
PI-3K, and hexokinase II mRNAs (11), and decreases LPL
mRNA, in agreement with the insulin stimulation of glucose
uptake and use and the inhibition of fatty acid oxidation in this
As to the mechanism of the modification of the mRNA lev-
els, insulin could affect the rate of the gene transcription and/
or the stability of the mRNAs. Little is known about the hor-
monal regulation of the last process, but it is probably an im-
portant control step (2, 3). On the other hand, insulin could ex-
ert a transcriptional control of gene expression, either directly
or through the stimulation of glucose metabolism in the cell
(2–4, 37). Insulin response sequences and glucose response ele-
ments have been identified in the promoter sequence of sev-
eral genes. Among the genes that seem to be regulated by in-
sulin in our work, the promoter regions of the genes coding
Rad and the p85? subunit of phosphatidylinositol-3-kinase
have not been characterized. In addition, putative insulin regu-
latory elements in the promoters of Glut 4 and LPL genes
have not been clearly identified, while transcriptional effects
of insulin on the Glut 4 gene have been demonstrated in ro-
dent models and cell lines (38).
In conclusion, our work demonstrated an acute and coordi-
nated regulation by insulin of the mRNA levels of key proteins
involved in insulin action in the skeletal muscle of normal hu-
mans. We have identified two important genes, (namely, Rad
and the p85 ? regulatory subunit of phosphatidylinositol-3-
kinase) that can be added to the growing list of the genes con-
trolled by insulin. Since the products of these genes, in particu-
lar PI-3K, play a crucial role in the insulin mechanism of action,
our findings strengthen the possibility that altered regulation
of gene expression by insulin could result in insulin resistance
(39). The methodology developed here allowing a wide screen-
ing of mRNA levels represents a valuable tool to test this hy-
The authors thank Dr. M.H. Rider for critical reading of the manu-
This study was supported in part by grant MESR 94G0268 from
the French Ministère de l’Enseignement Supérieur et de la Recherche.
1. Cheatham, B., and R.C. Kahn. 1995. Insulin action and the insulin signal-
ing network. Endocr. Rev. 16:117–142.
2. O’Brien, R.M., and D.K. Granner. 1991. Regulation of gene expression
by insulin. Biochem. J. 278:609–619.
3. Lemaigre, F.P., and G.G. Rousseau. 1994. Transcriptional control of
genes that regulate glycolysis and gluconeogenesis in adult liver. Biochem. J.
4. Vaulont, S., and A. Kahn. 1994. Transcriptional control of metabolic reg-
ulation genes by carbohydrates. FASEB J. 8:28–35.
5. DeFronzo, R.A. 1988. The triumvirate: ?-cell, muscle, liver. A collusion
responsible for NIDDM. Diabetes. 37:667–687.
6. Reynet, C., and C.R. Kahn. 1993. Rad: a member of the Ras family over-
expressed in muscle of type II diabetic humans. Science (Wash. DC). 262:1441–
7. Vestergaard, H., C. Bjorbaek, T. Hansen, F.S. Larsen, D.K. Granner, and
O. Pedersen. 1995. Impaired activity and gene expression of hexokinase II in
muscle from non–insulin-dependent diabetes mellitus patients. J. Clin. Invest.
8. Vestergaard, H., S. Lund, F.S. Larsen, O.J. Bjerrum, and O. Pedersen.
1993. Glycogen synthase and phosphofructokinase protein and mRNA levels in
skeletal muscle from insulin-resistant patients with non–insulin-dependent dia-
betes mellitus. J. Clin. Invest. 91:2342–2350.
9. Garvey, T.W., L. Maianu, J.A. Hancock, A.M. Golichowski, and A.
Baron. 1992. Gene expression of Glut 4 in skeletal muscle from insulin-resistant
patients with obesity, IGT, GDM and NIDDM. Diabetes. 41:465–475.
10. Vestergaard, H., P.H. Andersen, S. Lund, P. Vedel, and O. Pedersen.
1994. Expression of glycogen synthase and phosphofructokinase in muscle from
type 1 (insulin-dependent) diabetic patients before and after intensive insulin
treatment. Diabetologia. 37:82–90.
11. Mandarino, L.J., R.L. Printz, K.A. Cusi, P. Kinchington, R.M.
O’Doherty, H. Osawa, C. Sewell, A. Consoli, D.K. Granner, and R.A. De-
Fronzo. 1995. Regulation of hexokinase II and glycogen synthase mRNA, pro-
tein, and activity in human muscle. Am. J. Physiol. 269:E701–E708.
12. Yki-Järvinen, H., H. Vuorinen-Markkola, L. Koranyi, R. Bourey, K.
Tordjman, M. Mueckler, A.M. Permutt, and V.A. Koivisto. 1992. Defect in in-
sulin action on expression of the muscle/adipose tissue glucose transporter gene
in skeletal muscle of type 1 diabetic patients. J. Clin. Endocrinol. & Metab. 75:
13. Schalin-Jänti, C., H. Yki-Järvinen, L. Koranyi, R. Bourey, J. Lindström,
P. Nikula-Ijäs, A. Franssila-Kallunki, and L.C. Groop. 1994. Effect of insulin on
Glut 4 mRNA and protein concentrations in skeletal muscle of patients with
NIDDM and their first-degree relatives. Diabetologia. 37:401.
14. Wang, A.M., M.V. Doyle, and D.F. Mark. 1989. Quantitation of mRNA
by the polymerase chain reaction. Proc. Natl. Acad. Sci. USA. 86:9717–9721.
Insulin Action on Gene Expression in Skeletal Muscle
15. Bouaboula, M., P. Legoux, B. Pességué, B. Delpech, X. Dumont, M.
Piechaczyk, P. Casellas, and D. Shire. 1992. Standardization of mRNA titration
using polymerase chain reaction method involving co-amplification with multi-
specific internal control. J. Biol. Chem. 267:21830–21838.
16. Rizza, R.A., L.J. Mandarino, and J.E. Gerich. 1981. Dose-response
characteristics for effects of insulin on production and utilization of glucose in
man. Am. J. Physiol. 240:E630–E639.
17. Laville, M., J.P. Riou, P.F. Bougnères, B. Canivet, M. Beylot, R. Cohen,
P. Serusclat, C. Dumontet, F. Berthezene, and R. Mornex. 1984. Glucose me-
tabolism in experimental hyperthyroidism: intact in vivo sensitivity to insulin
with abnormal binding and increased glucose turnover. J. Clin. Endocrinol. &
18. Chomczynski, P., and N. Sacchi. 1987. Single step method of RNA isola-
tion by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Bio-
19. Gilliland, G., S. Perrin, K. Blanchard, and H.F. Bunn. 1990. Analysis of
cytokine mRNA and DNA: detection and quantitation by competitive poly-
merase chain reaction. Proc. Natl. Acad. Sci. USA. 87:2725–2729.
20. Vidal, H., D. Auboeuf, M. Beylot, and J.P. Riou. 1995. Regulation of in-
sulin receptor mRNA splicing in rat tissues: effect of fasting, aging and diabe-
tes. Diabetes. 44:1196–1201.
21. Higuchi, R. 1989. Using PCR to engineer DNA. In PCR Technology:
Principles and Application for DNA Amplification. H.A. Erlich, editor. Stock-
ton Press, New York. 61–70.
22. Araki, E., X.J. Sun, B.L. Haag, L.M. Chuang, Y. Zhang, T.L. Yang-
Feng, M.F. White, and C.R. Kahn. 1993. Human skeletal muscle insulin recep-
tor substrate-1: characterization of the cDNA, gene, and chromosomal localiza-
tion. Diabetes. 42:1041–1054.
23. Fukumoto, H., T. Kayano, J.B. Buse, Y. Edwards, P.F. Pilch, G.I. Bell,
and S. Seino. 1989. Cloning and characterization of the major insulin-respon-
sive glucose transporter expressed in human skeletal muscle and other insulin-
responsive tissues. J. Biol. Chem. 264:7776–7779.
24. Lange, A.J., and S.J. Pilkis. 1990. Sequence of the human 6-phospho-
fructo-2-kinase/fructose-2,6-bisphosphatase. Nucleic Acids Res. 18:3652–3655.
25. Browner, M.F., K. Nakano, A.G. Bang, and R.J. Fletterick. 1989. Hu-
man muscle glycogen synthase cDNA sequence: a negatively charged protein
with an asymmetric charge distribution. Proc. Natl. Acad. Sci. USA. 86:1443–
26. Langin, D., H. Laurell, L. Stenson-Holst, P. Belfrage, and C. Holm.
1993. Gene organization and primary structure of human hormone-sensitive li-
pase: possible significance of a sequence homology with the lipase of Moraxella
TA 144, an antarctic bacterium. Proc. Natl. Acad. Sci. USA. 90:4897–4901.
27. Nakajima, H., T. Noguchi, T. Yamasaki, N. Kono, T. Tanaka, and S.
Tarui. 1987. Cloning of human muscle phosphofructokinase cDNA. FEBS Lett.
28. Guessow, D., R. Rein, I. Ginjaar, F. Hochstenbach, G. Seemann, A.
Kottman, and H.L. Ploegh. 1987. The human beta-2-microglobulin gene: pri-
mary structure and definition of the transcription unit. J. Immunol. 139:3132–
29. Skolnik, E.Y., B. Margolis, M. Mohammadi, E. Lowenstein, R. Fischer,
A. Drepps, A. Ullrich, and J. Schlessinger. 1991. Cloning of PI3 kinase-associ-
ated p85 utilizing a novel method for expression/cloning of target proteins for
receptor tyrosine kinases. Cell. 65:83–90.
30. Riond, J., M.G. Mattei, M. Kaghad, X. Dumont, J.C. Guillemot, G. Le
Fur, D. Caput, and P. Ferrara. 1991. Molecular cloning and chromosomal local-
ization of a human peripheral-type benzodiazepine receptor. Eur. J. Biochem.
31. Ebina, Y., L. Ellis, K. Jarnagin, M. Edery, L. Graf, E. Clauser, J. Ou, F.
Massiarz, Y.W. Kan, I.D. Goldfine, et al. 1985. The human insulin receptor
cDNA: the structural basis for hormone-activated transmembrane signalling.
32. Wion, K.L., T.G. Kirchgessner, A.J. Lusis, M.C. Schotz, and R.M.
Lawn. 1987. Human lipoprotein lipase complementary DNA sequence. Science
Wash. DC 235:1638–1641.
33. Norgren, S., P. Arner, and H. Luthman. 1994. Insulin receptor ribonu-
cleic acid levels and alternative splicing in human liver, muscle, and adipose tis-
sue: tissue specificity and relation to insulin action. J. Clin. Endocrinol. &
34. Kiens, B., H. Lithell, K.J. Mikines, and E.A. Richter. 1989. Effects of in-
sulin and exercise on muscle lipoprotein lipase activity in man and its relation to
insulin action. J. Clin. Invest. 84:1124–1129.
35. Norgren, S., J. Zierath, A. Wedell, H. Wallberg-Henriksson, and H.
Luthman. 1994. Regulation of human insulin receptor RNA splicing in vivo.
Proc. Natl. Acad. Sci. USA. 91:1465–1469.
36. Sell, S.M., D. Reese, and V.M. Ossowski. 1994. Insulin-inducible
changes in insulin receptor mRNA splice variants. J. Biol. Chem. 269:30769–
37. Granner, D.K., and S. Pilkis. 1990. The genes of hepatic glucose metab-
olism. J. Biol. Chem. 265:10173–10176.
38. Stephens, J.M., and P.F. Pilch. 1995. The metabolic regulation and vesic-
ular transport of Glut 4, the major insulin-responsive glucose transporter. En-
docr. Rev. 16:529–546.
39. O’Brien, R.M., and D.K. Granner. 1995. Why there is an IRS. J. Clin.