Regulation of Glycogen Synthase by Glucose and
Glycogen
A Possible Role for AMP-Activated Protein Kinase
Reza Halse,1 Lee G.D. Fryer,2 James G. McCormack,3 David Carling,2 and Stephen J. Yeaman1
We report here use of human myoblasts in culture to
study the relationships between cellular glycogen concentrations and the activities of glycogen synthase (GS)
and AMP-activated protein kinase (AMPK). Incubation
of cells for 2 h in the absence of glucose led to a 25%
decrease in glycogen content and a significant decrease
in the fractional activity of GS. This was accompanied
by stimulation of both the ␣1 and ␣2 isoforms of AMPK,
without significant alterations in the ratios of adenine
nucleotides. When glucose was added to glycogen-depleted cells, a rapid and substantial increase in GS
activity was accompanied by inactivation of AMPK back
to basal values. Inclusion of the glycogen phosphorylase
inhibitor, CP-91149, prevented the loss of glycogen
during glucose deprivation but not the activation of
AMPK. However, in the absence of prior glycogen breakdown, glucose treatment failed to activate GS above
control values, indicating the crucial role of glycogen
content. Activation of AMPK by either 5-aminoimidazole-4-carboxamide 1-␤-D-ribofuranoside (AICAR) or
hydrogen peroxide was also associated with a decrease
in the activity ratio of GS. AICAR treatment had no
effect on total cellular glycogen content but led to a
modest increase in glucose uptake. These data support a
role for AMPK in both stimulating glucose uptake and
inhibiting GS in intact cells, thus promoting glucose flux
through glycolysis. Diabetes 52:9 –15, 2003
From the 1School of Biochemistry and Genetics, Medical School, University of
Newcastle, Newcastle upon Tyne, U.K.; 2Cellular Stress Group, MRC Clinical
Sciences Centre, Imperial College School of Medicine, Hammersmith Hospital, London, U.K.; and 3Target Cell Biology, Novo Nordisk A/S, Bagsvaerd,
Denmark.
Address correspondence and reprint requests to Stephen J. Yeaman, School
of Biochemistry and Genetics, The Medical School, University of Newcastle,
Newcastle upon Tyne NE2 4HH, U.K. E-mail: [email protected].
Received for publication 28 March 2001 and accepted in revised form 24
September 2002.
J.G.M. is currently affiliated with OSI Pharmaceuticals Ltd., Oxford, U.K.
R.H. receives consulting fees from Xcellsyz, Ltd., a start-up company
engaged in deriving immortalized cell lines for study of diabetes. J.G.M. is
employed by and holds stock in Novo Nordisk A/S. D.C. is on the Scientific
Advisory Board for Xcellsyz. S.J.Y. holds stock in Xcellsyz and has received
honoraria from Novo Nordisk and Glaxo Wellcome.
AICAR, 5-aminoimidazole-4-carboxamide 1-␤-D-ribofuranoside; AMPK,
AMP-activated protein kinase; DME Glu⫺, glucose-free Dulbecco’s Modified
Eagle’s medium; DTT, diothiothreitol; GS, glycogen synthase; GSK3, GS kinase
3; PKA, cAMP-dependent protein kinase; PMSF, phenylmethylsulfonyl fluoride.
DIABETES, VOL. 52, JANUARY 2003
G
lycogen synthase (GS) catalyzes a crucial and
rate-limiting step in muscle nonoxidative glucose disposal (1). The regulation of GS activity
is complex. Enzyme activity is sensitive to
allosteric regulation by a number of metabolites (2), is
subject to reversible phosphorylation, which inactivates
the enzyme, (3) and is regulated by feedback inhibition by
glycogen (4 – 6) via an unknown mechanism. GS activity is
modulated by reversible phosphorylation of primarily
three specific serine residues, collectively termed site 3
(3). GS is maintained in a low-activity state under basal
conditions principally through the continual phosphorylation of site 3 by GS kinase 3 (GSK3) (7). Insulin is believed
to activate GS mainly through the inhibition of GSK3 (8,9);
however, some level of regulation may control glycogentargeted protein phosphatases (10). A number of other
kinases have been identified that can phosphorylate GS in
vitro (11), including AMP-activated protein kinase (AMPK),
which can phosphorylate serine 7 (termed site 2) of GS
(12). Phosphorylation of site 2, which can also be catalyzed by cAMP-dependent protein kinase (PKA), primes
GS for further phosphorylation at site 2a by casein kinase
I, which in turn leads to a decrease in GS activity (13).
AMPK is a metabolite-sensing enzyme that has been implicated in the mediation of exercise-induced glucose
uptake (14), although to date, little experimental evidence
has attributed a role for AMPK in the regulation of GS
activity in vivo.
GS becomes activated following glycogen depletion
(such as occurs during exhaustive exercise) in an insulinindependent manner (15). The molecular mechanism underlying this phenomenon has not been elucidated,
although a number of hypotheses have been proposed.
One possibility is the involvement of an insulin-independent pathway leading to GSK3 inactivation (16). There is
evidence that this is the case in rat muscle (14), although
a more recent study in humans has suggested a GSK3independent mechanism (17). Another possibility is that
decreased cellular glycogen content may directly lead to
GS activation. We have recently developed a model system
in vitro using cultured human muscle, in which glycogen
depletion is achieved by glucose deprivation (18). Following readdition of glucose, a dramatic and sustained increase in GS activity is observed, which is independent of
GSK3 inactivation. Indeed, the mechanism leading to GS
9
GLYCOGEN SYNTHESIS IN HUMAN MYOBLASTS
activation in this model is independent of that utilized by
insulin (18).
The purpose of the present study was to further define
the role of cellular glycogen content in determining GS
activity and determine what role, if any, AMPK plays in
controlling glycogen metabolism. Using a specific inhibitor
of glycogenolysis, we have established a direct requirement for prior glycogen breakdown in the subsequent
activation of GS by glucose. We have also provided
evidence that AMPK is involved in the regulation of GS
activity and the rate of glucose uptake in cultured human
muscle cells.
RESEARCH DESIGN AND METHODS
Materials. All tissue culture trays were from Costar (Cambridge, MA).
Culture media, penicillin/streptomycin, and trypsin-EDTA were from GibcoBRL (Paisley, U.K.). Chick embryo extract and [␥-32P]ATP (148 TBq/mmol)
was obtained from ICN (Costa Mesa, CA). Uridine diphospho-D-[6-3H]glucose
(814 GBq/mmol) and 2-deoxy-D-[1-3H]glucose (362 GBq/mmol) were from
Amersham Pharmacia Biotech (Buckinghamshire, U.K.). Glucose-6-phosphate
dehydrogenase, hexokinase, and amyloglucosidase were from Boehringer
Mannheim (Lewes, U.K.). 5-Aminoimidazole-4-carboxamide 1-␤-D-ribofuranoside (AICAR) was from Sigma (Poole, U.K.). CP-91149 was provided by Pfizer
(Pfizer Global Research & Development, Groton Laboratories, Groton, CT).
Antibodies used for immunoprecipitation of ␣1 and ␣2 isoforms of AMPK
were as described previously (19).
Cell culture. Human myoblasts were grown from needle biopsy samples
taken from the vastus lateralis muscle of healthy subjects with no family
history of type 2 diabetes and with normal glucose tolerance and normal
insulin sensitivity, as assessed using the short insulin tolerance test (20).
Myoblasts were maintained in growth medium consisting of HAMS F-10
nutrient mixture containing 20% FCS, 1% chick embryo extract, 100 units/ml
penicillin, and 100 ␮g/ml streptomycin. All experiments were performed using
cells between the 5th and 15th passage at greater than 80% confluence.
Myoblasts and myotubes respond equally well to insulin and other agonists at
the level of glucose metabolism, in many respects mirroring the situation in
whole muscle. Differentiation of myoblasts to myotubes induces significant
increases in GS expression and some lowering of the activity ratio of the
enzyme (21). To avoid variations in basal GS activity, as a result of innate
differences in the differentiation capacity of myoblasts from different subjects,
myoblasts were used throughout this study.
Cellular glycogen content determination. Total cellular glycogen content
was assessed by modification of a previous method (22). Following treatment,
cells were washed rapidly in ice-cold PBS and scraped into 100 ␮l of 0.2 mol/l
sodium acetate, pH 4.8. Extracts were briefly sonicated, using a Soniprep 150,
before addition of 250 mU amyloglucosidase per sample. Samples were
incubated for 2 h at 40°C and vortexed regularly to avoid sedimentation.
Sample was incubated with assay cocktail (0.1 mol/l Tris-HCl, pH 8.0, 0.3
mmol/l ATP, 6 mmol/l MgCl2, 5 mmol/l diothiothreitol [DTT], 60 ␮mol/l
NADP⫹, 2.5 units/ml hexokinase, and 1 ␮g/ml G6P-dehydrogenase) for 30 min
at room temperature. Changes in fluorescence, as a result of NADPH
production, were determined using an excitation wavelength of 355 nm and an
emission wavelength at 460 nm. Reaction blanks were determined as the
fluorescence of samples before enzymatic treatment with amyloglucosidase.
Estimation of glucose uptake. Glucose uptake was determined as the rate
of 2-deoxy-D-[6-3H]glucose uptake, using modification of a previous method
(23). Cells were maintained in the absence of serum for 5 h before the
replacement of media with glucose-free Dulbecco’s Modified Eagle’s medium
(DME Glu⫺) for 15 min at 37°C. For AICAR treatments, cells were incubated
in serum free for 2 h before the addition of AICAR for the times indicated. The
rate of 2-deoxyglucose uptake was determined during 5 min of incubation
with 50 ␮mol/l 2-deoxy-D-[6-3H]glucose (specific activity 0.4 kBq/pmol). Reaction blanks were determined as the rate of 2-deoxy-D-[6-3H]glucose uptake in
the presence of 0.1 mmol/l cytochalasin B.
Following incubation, cells were washed with ice-cold PBS several times
and solubilized in 0.05% SDS for 30 min at room temperature. Protein content
of samples was assayed using Coomassie Protein Assay Reagent, and uptake
of 2-deoxy-D-[6-3H]glucose was determined by liquid scintillation counting.
Assay of GS. Following the indicated treatments, cells were rapidly washed
three times with ice-cold PBS and collected, by scraping, into GS extraction
buffer (10 mmol/l Tris-HCl, pH 7.8, 150 mmol/l KF, 15 mmol/l EDTA, 60 mmol/l
sucrose, 1 mmol/l 2-mercaptoethanol, 10 ␮g/ml leupeptin, 1 mmol/l benzamidine, and 1 mmol/l phenylmethylsulfonyl fluoride [PMSF]). Cells were then
10
disrupted by briefly sonicating using a Soniprep 150. GS activity was determined in whole lysates as the incorporation of 3H-glucose from uridine-5⬘diphosphate-[U-3H]glucose into glycogen, as described by Guinovart et al.
(24). Samples were incubated with reaction cocktail (50 mmol/l Tris-HCl, pH
7.8, 20 mmol/l EDTA, 25 mmol/l KF, 1% glycogen, 0.4 mmol/l UDP-[3H]glucose
[specific activity 3 kBq/nmol]), containing either 0.1 mmol/l (active) or 10
mmol/l glucose-6-phosphate (total), for 30 min at 30°C. Results were expressed as fractional activities (active/total). This assay has been optimized to
detect the activity changes resulting from dephosphorylation of GS (24).
AMPK activity determinations. Following the indicated treatments, cells
were rapidly washed three times with ice-cold PBS and collected, by scraping,
into buffer A (50 mmol/l Tris-HCl, pH 7.5, 1 mmol/l EDTA, 50 mmol/l NaF, 5
mmol/l NaPPI, 1 mmol/l benzamidine, 10% glycerol, 1% Triton X-100, 1 mmol/l
DTT, and 0.1 mmol/l PMSF). Samples were briefly sonicated before centrifugation at 13,000g for 5 min (4°C). Immunoprecipitations of ␣1, ␣2, and total
AMPK were performed on aliquots of supernatants containing 30 ␮g protein.
In each case, immunoprecipitations were carried out over 2 h at 4°C, using
appropriate antibodies and protein A or protein G immobilized on Sepharose.
The immune complexes were recovered by brief centrifugation and washed
twice with buffer A and twice with buffer B (50 mmol/l Hepes, pH 7.4, 1 mmol/l
EDTA, 10% glycerol, and 1 mmol/l DTT).
AMPK activity in immunoprecipitates was assayed in a final volume of 25
␮l containing 50 mmol/l Hepes, pH 7.4, 1 mmol/l EDTA, 10% glycerol, 1 mmol/l
DTT, 0.2 mmol/l SAMS peptide substrate (25), 200 ␮mol/l [␥-32P]ATP (specific
activity ⬃1.1 kBq/nmol), 5 mmol/l MgCl2, and 0.2 mmol/l AMP. After incubation for 30 min at 30°C, samples were centrifuged briefly, and 20 ␮l of the
supernatant containing the radiolabelled peptide product was spotted onto 1
cm2 Whatman P81 phosphocellulose paper squares. After washing in 1%
phosphoric acid with four changes, the papers were dried and phosphate
incorporation was determined by liquid scintillation counting. Enzyme activity
(U) was defined as that which catalyzes the incorporation of 1 pmol of
phosphate into peptide substrate in 1 min.
Statistics. All results are expressed as means ⫾ SE. Statistical analysis was
made using a two-tailed unpaired Student’s t test, following one-way ANOVA.
RESULTS
Activation of GS by glucose following time-dependent
decrease in cellular glycogen in response to glucose
deprivation has been reported in human muscle cells in
culture (18), suggesting that induced changes in cellular
glycogen content might be responsible for alterations in
the activity of GS. However, it was unclear whether
changes in the activity state of GS were a direct result of
alterations in the concentration of cellular glycogen or of
other metabolic consequences of glucose deprivation.
Therefore, a well-characterized inhibitor of liver glycogen
phosphorylase, CP-91149 (26), was used to dissociate
changes in intracellular glycogen levels from other experimental variables. The ability of CP-91149 to affect intracellular glycogen levels was assessed in human myoblasts
in culture (Table 1). Incubation of myoblasts in the absence of glucose for 2 h caused an ⬃25% decrease in
intracellular glycogen concentrations, consistent with an
earlier work (18). This decrease was essentially blocked
by 10 and 100 ␮mol/l CP-91149, indicating that this compound can also inhibit the human muscle isoform of
glycogen phosphorylase. At an inhibitor concentration of
10 ␮mol/l, no change in glycogen content was observed in
cells maintained in normal glucose (6.1 mmol/l); however,
100 ␮mol/l CP-91149 caused a small but significant increase in glycogen accumulation in these cells. For further
studies, 10 ␮mol/l CP-91149 was used to inhibit glycogenolysis in order to avoid alterations in basal glycogen levels.
Glycogen levels were not depleted in the presence of 10
␮mol/l CP-91149 following up to 7 h of glucose deprivation, whereas in the absence of inhibitor, glycogen levels
fell by ⬎50% in 5 h (Table 1).
The effect of CP-91149 on starvation-induced changes in
DIABETES, VOL. 52, JANUARY 2003
R. HALSE AND ASSOCIATES
TABLE 1
Concentration and time-dependent effects of glycogen phosphorylase inhibitor CP-91149 on glycogen content
Control
Glucose-starved
0
Concentration of CP-91149 (␮mol/l)
10
100
100
76.4 ⫾ 3.5*
102.9 ⫾ 1.9
95.5 ⫾ 3.0
108.8 ⫾ 2.9*
94.3 ⫾ 5.0
Time of treatment (h)
Glucose-starved
Glucose-starved
⫹ CP-91149
0
2
5
7
100
76.8 ⫾ 4.1*
45.6 ⫾ 6.2*
—
108.6 ⫾ 24.1
109.8 ⫾ 19.9
133.5 ⫾ 19.7*
115.5 ⫾ 17.5
Myoblasts were incubated in media containing (Control) or lacking glucose (6.1 mmol/l) for 2 h, during which time media was supplemented
or not with CP-91149 at the concentrations indicated (Upper table). Alternatively, cells were incubated for the times indicated in media
lacking glucose, with the inclusion or not of 10 ␮mol/l CP-91149 (Lower table). Cells were then extracted and the total cellular glycogen
content was determined. Results are expressed as a % of glycogen content in control cells (2.69 ⫾ 0.25 ␮mol glucose/mg protein) and
represent the mean ⫾ SE of n ⫽ 6 in three subjects. *P ⬍ 0.05 vs. control values in the absence of inhibitor.
2-deoxyglucose uptake was then examined (Fig. 1). Glucose withdrawal from myoblasts for 5 h caused a 1.6-fold
increase in the rate of 2-deoxyglucose uptake, as compared with cells maintained in glucose-containing media.
This is consistent with an earlier work (18). The basal rate
of 2-deoxyglucose uptake was unaffected by the presence
of CP-91149 during glucose deprivation; however, a slight
decrease was observed in the rate of uptake following
glucose-deprivation (644.7 ⫾ 24.1 pmol 䡠 min⫺1 䡠 mg⫺1 in
untreated cells vs. 574.1 ⫾ 15.2 pmol 䡠 min⫺1 䡠 mg⫺1 in
CP-91149 treated cells; P ⬍ 0.05). In the presence of
CP-91149, and therefore in the absence of glycogen breakdown (Table 1), a 1.5-fold increase in the rate of glucose
uptake following glucose deprivation persisted, indicating
that in this system the stimulation of glucose uptake
during glucose deprivation is not dependent on cellular
glycogen content.
A direct relationship between intracellular glycogen
concentration and GS activity has been suggested (5,6,27).
We have also previously reported a dramatic and sustained
increase in the fractional activity of GS following glucose
treatment of glucose-deprived, glycogen-depleted, human
myoblasts (18). CP-91149 was used to dissociate changes
in the activity of GS resulting from changes in cellular
glycogen concentration from other variables (Fig. 2).
Glucose-deprivation of cells for 2 h caused a significant
decrease in the fractional activity of GS before re-admission of glucose (0.020 in untreated cells to 0.004 in
glucose-starved cells) without altering the activity of GS in
the presence of saturating G6P (10 mmol/l) concentrations
(46.14 ⫾ 5.3 nmol 䡠 min⫺1 䡠 mg⫺1 in untreated cells to
54.14 ⫾ 6.4 nmol 䡠 min⫺1 䡠 mg⫺1 in glucose-starved cells).
Glucose treatment (5.5 mmol/l) of previously glucosestarved cultures increased GS activity approximately fivefold over that observed in control cultures. The inclusion
of 10 ␮mol/l CP-91149 had no effect on GS activity in either
control or glucose-starved cells; however, the fivefold
increase in GS activity over control cultures observed
during glucose re-admission was completely inhibited by
prior treatment with CP-91149, indicating that this effect is
totally dependent on prior depletion of glycogen. It is
noteworthy that the fractional activity of GS in glucosedeprived cells was restored to levels observed in control
cells following glucose re-admission of cultures mainDIABETES, VOL. 52, JANUARY 2003
tained in the presence of CP-91149 (0.007 before glucose
re-admission vs. 0.03 ⫾ 0.02 following glucose re-admission). Therefore, in the absence of glycogen breakdown,
glucose deprivation induces a decrease in GS activity that
can be reversed by re-admission of glucose. However,
glycogen depletion is required for full activation of GS
observed following the treatment of previously glucosestarved cells with glucose.
Although AMPK has been shown to phosphorylate site 2
of GS in vitro (12), to date no evidence has been offered to
suggest that GS is a physiological substrate of AMPK.
However, a reduction in the intracellular ATP-to-AMP
ratio following glucose-deprivation has been demonstrated in a pancreatic cell line (28). In addition, in view of
the fact that the intracellular ratio of ATP to AMP plays a
major role in the regulation of AMPK activity (29), we
wished to explore whether AMPK mediated the effects of
glucose withdrawal on GS activity. The activity of ␣1 and
␣2 AMPK isoforms was therefore examined in human
myoblasts following glucose-withdrawal and glucose readmission (Fig. 3). Significant levels of both the ␣1 and ␣2
AMPK isoforms were detected in control cells. An approximate fourfold increase in the activity of both the ␣1 and
FIG. 1. Effect of glycogen-depletion on 2-deoxyglucose uptake. Myoblasts were incubated in media lacking (Glu-) or containing 6.1 mmol/l
glucose (Control) for 5 h, in the absence (Basal) or presence of 10
␮mol/l CP-91149. Following incubations, the rate of 2-deoxyglucose
uptake was determined. Results represent the mean ⴞ SE of n ⴝ 4 in
three subjects. *P < 0.05 vs. control values.
11
GLYCOGEN SYNTHESIS IN HUMAN MYOBLASTS
FIG. 2. Dependence of glucose starvation/readmission-induced increases in GS activity on glycogen depletion. Myoblasts were incubated
in the presence (Control) or absence (Glu-) of 6.1 mmol/l glucose for
2 h. Where indicated, CP-91149 (10 ␮mol/l) was included for the
duration of incubations. Cells were then exposed (Re-Fed), or not, to
glucose-containing media for 10 min, before extraction. The fractional
activity of GS was determined in these extracts. Results represent the
mean ⴞ SE of n ⴝ 6 in three subjects. *P < 0.05 vs. values obtained in
control cells in the absence of CP-91149.
␣2 AMPK isoforms from 7.04 ⫾ 1.4 and 1.8 ⫾ 0.7 units/mg
to 26.8 ⫾ 6.1 and 7.5 ⫾ 2.0 units/mg, respectively, was
observed in myoblasts following 2 h of glucose withdrawal, during which time the fractional activity of GS fell
significantly (Fig. 2). Somewhat surprisingly, no significant
alteration in the ratio of ATP to ADP was apparent
following glucose starvation (glucose-fed control ATP/
ADP 3.6 ⫾ 0.3 compared with glucose-starved ATP/ADP
3.6 ⫾ 0.2; n ⫽ 6). AMPK activity returned to control values
following treatment of cells with 5.5 mmol/l glucose for 10
min, conditions associated with dramatic activation of GS
but again without significant change in the ATP/ADP ratio
(4.7 ⫾ 0.5). It is noteworthy that AMPK activity returned to
control values after 10 min of glucose re-administration, a
time at which glycogen levels are still significantly depleted (18). In the presence of CP-91149, a greater increase
in AMPK activity was observed in glucose-starved cells,
with a dramatic decrease in the ATP-to-ADP ratio (0.53).
Again, glucose re-admission caused AMPK levels to return
to control values, which was associated with a restoration
of the fractional activity of GS to control value. However,
the dramatic reactivation of GS was not observed, indicating that an additional glycogen-dependent mechanism is
involved.
AICAR, the cell-permeable precursor of ZMP, has been
shown to selectively activate AMPK in a number of model
systems (30,31). To assess further the involvement of
AMPK in the regulation of GS and glucose uptake, human
myoblasts were treated with AICAR (2 mmol/l) (Fig. 4).
Incubation of cells with AICAR caused a time-dependent
increase in AMPK activity, with stimulation being observed within 30 min and reaching a maximum of approximately twofold after 90 min. AICAR treatment also caused
a time-dependent decrease in the activity of GS (0.063 ⫾
0.01 in control cells vs. 0.029 following 2 h of AICAR
treatment) and an increase in the rate of 2-deoxyglucose
uptake (1.3-fold over basal levels following 2 h of AICAR
treatment), again consistent with a role for AMPK in
controlling both parameters. Total cellular glycogen con12
FIG. 3. Effects of glucose starvation and glycogen depletion on the
activities of AMPK isoforms. Myoblasts were incubated in the presence
(Control) or absence (Glu-) of 6.1 mmol/l glucose for 2 h. Where
indicated, CP-91149 (10 ␮mol/l) was included for the duration of
incubations. Cells were then exposed (Re-Fed), or not, to glucosecontaining media for 10 min, before extraction. The activity of ␣1 and
␣2 AMPK isoforms was determined in resulting extracts. Results
represent the mean ⴞ SE of n ⴝ 4 in three subjects. *P < 0.05 vs. values
obtained in control cells, in the absence of CP-91149.
tent was unaffected following AICAR treatment, suggesting that observed decreases in GS activity were not a
result of glycogen accumulation. In comparison to maximal treatments of AICAR (2 h at 2 mmol/l), glucosedeprivation of myoblasts (2 h) caused a greater activation
of AMPK (19.2 ⫾ 3 nmol 䡠 min⫺1 䡠 mg⫺1 in AICAR-treated
cells vs. 27.9 ⫾ 4 nmol 䡠 min⫺1 䡠 mg⫺1 in glucose-starved
cells) and inhibition of GS (0.029 ⫾ 0.0 in AICAR-treated
cells vs. 0.017 ⫾ 0.01 in glucose-starved cells). Cells in
normal glucose conditions were also treated with hydrogen peroxide, another known activator of AMPK (32), to
further examine the role of this enzyme in modulating GS
activity (Fig. 5). H2O2 rapidly activated AMPK in a transient manner, and this was mirrored by a transient decrease in GS activity.
DISCUSSION
Glycogen levels are depleted during exhaustive exercise
(15). Restoration of these levels is in part independent of
insulin action (15) and involves increases in both glucose
uptake (33,34) and GS activity (35). However, to date the
molecular mechanisms underpinning these events have
been poorly understood. A role for glycogen in influencing
the metabolic steps involved in glycogen repletion has
been described (4). We have previously reported the use of
human myoblasts as an experimental system to study the
relationship between glycogen content and subsequent
glucose metabolism (18).
DIABETES, VOL. 52, JANUARY 2003
R. HALSE AND ASSOCIATES
FIG. 4. Effects of AICAR treatment on the rate of 2-deoxyglucose
uptake, GS activity, total cellular glycogen content, and AMPK activity
in myoblasts. Myoblasts were incubated in glucose-containing media
with the addition, or not, of AICAR (2 mmol/l) for the times indicated.
The total cellular glycogen content was then determined (A). Results
represent the mean ⴞ SE of n ⴝ 9 in three subjects. In separate
cultures, myoblasts were treated in an identical manner, but extracts
prepared for the determination of either GS (B), total AMPK activity
(C), or 2-deoxyglucose uptake (D). Results represent the mean ⴞ SE of
n ⴝ 3 in three subjects. For comparative purposes, the rate of
DIABETES, VOL. 52, JANUARY 2003
Glucose deprivation of human muscle cells in culture
induces a decrease in both cellular glycogen content and
GS fractional activity. Subsequent glucose treatment of
cells causes a dramatic increase in the fractional activity
of GS, reflecting a decrease in the phosphorylation state of
the enzyme. A significant increase in the rate of glucose
uptake was also observed in glucose-starved cells. It could
be argued that the observed changes in glucose uptake and
GS activity in this model are a result of the glucose
withdrawal/re-admission protocol per se. To address this
concern, a specific inhibitor of glycogen phosphorylase
(CP-91149) was used to discriminate effects of glycogen
depletion from glucose deprivation. CP-91149 treatment of
human myoblasts completely inhibited glycogen breakdown during glucose deprivation and, at maximal concentrations, caused some slight glycogen accumulation. The
increased rate of glucose uptake in glucose-starved cells
was largely unaffected by inhibition of glycogenolysis,
implying that cellular glycogen content was not controlling
metabolite entry into the cell, following glucose-deprivation. In contrast, the large increase in GS activity observed
following re-admission of glucose to cells was severely
blunted by CP-91149, suggesting that glycogen depletion is
necessary for the superactivation of GS in this model. It is
noteworthy, however, that neither the decrease in GS
fractional activity during glucose deprivation nor the recovery of fractional activity to basal values following
glucose re-admission was affected by inhibition of glycogen breakdown. This suggests that two separate mechanisms are responsible for modulating GS activity in this
system, one that is dependent on glycogen depletion and
one that is not. The dramatic activation of GS by glucose
is clearly dependent on the preexisting glycogen content
of the cell. This provides direct evidence that intracellular
glycogen concentration can affect GS activity in the presence of physiological concentrations of glucose. Moreover, the low fractional activity of GS in human myoblasts,
as compared wth human muscle in vivo, may be explained
by the elevated intracellular glycogen concentration reported here and observed by others in human myotubes
(27).
A potential candidate for mediating the glycogen-independent effects on glucose uptake and GS fractional
activity is AMPK. In rat epitrochlearis muscles, AMPK
activity was strongly correlated with the rate of glucose
uptake following challenge with a variety of fuel-depleting
stimuli (14). In addition, AMPK has been implicated in
mediating contraction-induced increases in glucose uptake (31,35). It is worth noting that in a recent study of
transgenic mice where a dominant inhibitory AMPK mutant was overexpressed in muscle, exercise-stimulated
glucose uptake and translocation of Glut 4 to the cell
surface was only partly inhibited, implicating other AMPKindependent pathways in this process (36). Furthermore,
reported effects of AICAR on GS activity in rats are
confusing, apparently being dependent on muscle type and
experimental design (37,38).
AMPK is acutely sensitive to changes in cellular energy
2-deoxyglucose uptake and the activities of GS and AMPK were also
determined in myoblasts deprived of glucose for 2 h. (Results obtained
fall within the shaded areas, in each case.) *P < 0.05 vs. values
obtained in the absence of AICAR.
13
GLYCOGEN SYNTHESIS IN HUMAN MYOBLASTS
FIG. 5. Effect of hydrogen peroxide treatment on the activities of GS and AMPK. Myoblasts were exposed to H2O2 (50
␮mol/l) for the times indicated. Following treatments, the
total and fractional activities of AMPK and GS, respectively,
were determined in separate cultures. Results represent the
mean ⴞ SE of n ⴝ 4 in four subjects.
balance and is therefore a candidate for mediating the
inhibitory effects on GS. Indeed, inhibition of GS as a
result of fuel depletion in glucose-starved cells would be a
sensible energy-preserving response in order to meet
changing energy demands. AMPK has already been implicated in inhibition of other energy-consuming processes,
including fatty acid and cholesterol synthesis (29). In
human myoblasts, both AMPK ␣1 and ␣2 isoforms were
stimulated in response to glucose withdrawal in both the
absence and presence of the glycogen phosphorylase
inhibitor. AMPK activity rapidly returned to control values
in both instances following glucose re-administration. The
reverse correlation between GS fractional activity and
AMPK activity are consistent with a causal relationship
between the two.
The focus of the current article is the role of the AMPK
in controlling glucose metabolism, not its mechanism of
activation. However, several aspects of the nucleotide
ratio measurements merit discussion. Firstly, the ATP-toADP ratio is lower than in most other culture cell models,
although it is consistent with the value in neonatal cardiomyocytes (D.C., unpublished data). The absolute values
of the nucleotides is also low, making measurements
difficult, particularly that of AMP. As ATP, ADP, and AMP
are maintained in equilibrium, where AMP concentrations
are difficult to detect, the ratio of ATP to ADP can be used
as an indicator of AMP levels (39). In the present study, the
main observation is that the ATP-to-ADP ratio does not
change during glucose deprivation—presumably, glycogen
breakdown provides the necessary energy. Consistent
with this is the dramatic drop in ratio when glycogen
breakdown is inhibited by CP-9149. However, there are
now several reports where the activation of AMPK is
apparently independent of changes in the nucleotide ratios. These include activation of AMPK in hepatocytes (40)
and skeletal muscle (41) in response to the glucoselowering drug metformin, activation of AMPK in skeletal
muscle by leptin (42), and the effects of a number of
metabolites in perfused rat hearts (43). It is apparent from
these studies that other mechanisms regulate AMPK in
addition to the nucleotide ratio.
AICAR, a relatively selective activator of AMPK (30,31),
was used to further substantiate the role of AMPK in the
regulation of glucose uptake and GS activity. AICAR is an
intermediate in de novo purine biosynthesis and, once
metabolized (to ZMP), is a potent activator of AMPK (30).
14
In cultured human myoblasts, AICAR was a less potent
stimulator of AMPK activity than has been reported in
some other cell systems (19). This may be due to a lower
rate of AICAR metabolism and, thus, production of ZMP.
Despite relatively low levels of AMPK activation, AICAR
treatment of cultured human muscle cells stimulated
glucose uptake and inhibited GS activity in a time-dependent manner; also, there was a strong correlation between
GS and AMPK activity during both AICAR treatment and
glucose starvation. Furthermore, H2O2 treatment of cells
potently activated AMPK and inhibited GS activity. Therefore, several lines of evidence indicate that AMPK can
regulate GS in vivo, although whether AMPK is directly
phosphorylating GS remains to be established.
ACKNOWLEDGMENTS
We thank Dr. Dennis J. Hoover and Dr. Judith L. Treadway
for generously providing the glycogen phosphorylase inhibitor CP-91149 (Pfizer Inc., Groton Laboratories). This
work was supported in part by Diabetes U.K., Xcellsyz
Ltd., and Medical Research Council, U.K. R.H. held a CASE
studentship from the Biotechnology and Biological Sciences Research Council, U.K., partly funded by Novo
Nordisk A/S. We thank Dr. Mark Walker for his continued
assistance in obtaining muscle biopsies.
REFERENCES
1. Lawrence JC Jr, Roach PJ: New insights into the role and mechanism of
glycogen synthase activation by insulin. Diabetes 46:541–547, 1997
2. Piras R, Rothman LB, Cabib E: Regulation of muscle glycogen synthetase
by metabolites. Biochemistry 7:56 – 66, 1968
3. Roach PJ: Control of glycogen synthase by hierarchal protein phosphorylation. FASEB J 4:2961–2968, 1990
4. Laurent D, Hundal RS, Dresner A, Price TB, Vogel SM, Petersen KF,
Shulman GL: Mechanism of muscle glycogen autoregulation in humans.
Am J Physiol 278:E663–E668, 2000
5. Danforth WH, Harvey P: Glycogen synthetase and control of glycogen
synthesis in muscle. Biochem Biophys Res Commun 16:466 – 471, 1964
6. Piras R, Staneloni R: In vivo regulation of rat muscle glycogen synthetase
activity. Biochemistry 8:2153–2160, 1969
7. Parker PJ, Caudwell FB, Cohen P: Glycogen synthase from rabbit skeletal
muscle: effect of insulin on the state of phosphorylation of the seven
phosphoserine residues in vivo. Eur J Biochem 130:227–234, 1983
8. Welsh GI, Proud CG: Glycogen synthase kinase-3 is rapidly inactivated in
response to insulin and phosphorylates eukaryotic initiation factor eIF-2B.
Biochem J 294:625– 629, 1993
9. Borthwick AC, Wells AM, Rochford JJ, Hurel SJ, Turnbull DM, Yeaman SJ:
Inhibition of glycogen synthase kinase-3 by insulin in cultured human
DIABETES, VOL. 52, JANUARY 2003
R. HALSE AND ASSOCIATES
skeletal muscle myoblasts. Biochem Biophys Res Commun 210:738 –745,
1995
10. Dent P, Lavoinne A, Nakielny S, Caudwell FB, Watt P, Cohen P: The
molecular mechanism by which insulin stimulates glycogen synthesis in
mammalian skeletal muscle. Nature 348:302–308, 1990
11. Cohen P: Dissection of the protein phosphorylation cascades involved in
insulin and growth factor action. Biochem Soc Trans 21:555–567, 1993
12. Carling D, Hardie DG: The substrate and sequence specificity of the
AMP-activated protein kinase: phosphorylation of glycogen synthase and
phosphorylase kinase. Biochim Biophys Acta 1012:81– 86, 1989
13. Zhang W, DePaoli-Roach AA, Roach PJ: Mechanisms of multisite phosphorylation and inactivation of rabbit muscle glycogen synthase. Arch
Biochem Biophys 304:219 –225, 1993
14. Hayashi T, Hirshman MF, Fujii N, Habinowski SA, Witters LA, Goodyear
LJ: Metabolic stress and altered glucose transport: activation of AMPactivated protein kinase as a unifying coupling mechanism. Diabetes
49:527–531, 2000
15. Price TB, Rothman DL, Taylor R, Avison MJ, Shulman GI, Shulman RG:
Human muscle glycogen resynthesis after exercise: insulin-dependent and
-independent phases. J Appl Physiol 76:104 –111, 1994
16. Markuns JF, Wojtaszewski JFP, Goodyear LJ: Insulin and exercise decrease glycogen synthase kinase-3 activity by different mechanisms in rat
skeletal muscle. J Biol Chem 274:24896 –24900, 1999
17. Wojtazsewski JF, Nielsen P, Kiens B, Richter EA: Regulation of glycogen
synthase kinase-3 in human skeletal muscle: effects of food intake and
bicycle exercise. Diabetes 50:265–269, 2001
18. Halse R, Bonavaud SM, Armstrong JL, McCormack JG, Yeaman SJ: Control
of glycogen synthesis by glucose, glycogen, and insulin in cultured human
muscle cells. Diabetes 50:720 –726, 2001
19. Fryer LGD, Hajduch E, Rencurel F, Salt IP, Hundal HS, Hardie DG, Carling
D: Activation of glucose transport by AMP-activated protein kinase via
stimulation of nitric oxide synthase. Diabetes 49:1978 –1985, 2000
20. Halse R, Rochford JJ, McCormack JG, Vandenheede JR, Hemmings BA,
Yeaman SJ: Contol of glycogen synthesis in cultured human muscle cells.
J Biol Chem 274:776 –780, 1999
21. Halse R, Pearson SL, McCormack JG, Yeaman SJ, Taylor R: Effects of
tumor necrosis factor-␣ on insulin action in cultured human muscle cells.
Diabetes 50:1102–1109, 2001
22. Lust WD, Passonneau JV, Crites SK: The measurement of glycogen in
tissues by amylo-alpha-1,4-alpha-1,6-glucosidase after the destruction of
pre-existing glucose. Anal Biochem 68:328 –331, 1975
23. Sarabia V, Ramlal T, Klip A: Glucose uptake in human and animal muscle
cells in culture. Biochem Cell Biol 68:536 –542, 1990
24. Guinovart JJ, Salavert A, Massague J, Ciudad CJ, Salsas E, Itarte E:
Glycogen synthase: a new activity ratio assay expressing a high sensitivity
to the phosphorylation state. FEBS Lett 106:284 –288, 197
25. Davies SP, Carling D, Hardie DG: Tissue distribution of the AMP-activated
protein kinase, and lack of activation by cyclic-AMP-dependent protein
kinase, studied using a specific and sensitive peptide assay. Eur J Biochem
186:123–128, 1989
26. Martin WH, Hoover DJ, Armento SJ, Stock IA, McPherson RK, Danley DE,
Stevenson RW, Barrett EJ, Treadway JL: Discovery of a human liver
glycogen phosphorylase inhibitor that lowers blood glucose in vivo. Proc
Natl Acad Sci U S A 95:1776 –1781, 1998
DIABETES, VOL. 52, JANUARY 2003
27. Montell E, Arias A, Gomez-Foix AM: Glycogen depletion rather than
glucose 6-P increments controls early glycogen recovery in human cultured muscle. Am J Physiol 276:R1489 –R1495, 1999
28. Salt IP, Johnson G, Ashcroft SJH, Hardie DG: AMP-activated protein kinase
is activated by low glucose in cell lines derived from pancreatic beta cells,
and may regulate insulin release. Biochem J 335:533–539, 1998
29. Hardie DG, Carling D: The AMP-activated protein kinase: fuel gauge of the
mammalian cell? Eur J Biochem 246:259 –273, 1997
30. Corton JM, Gillespie JG, Hawley SA, Hardie DG: 5-Aminoimidazole-4carboxamide ribonucleoside: a specific method for activating AMP-activated protein kinase in intact cells? Eur J Biochem 229:558 –565, 1995
31. Russell IR, Bergeron R, Shulman GI, Young LH: Translocation of myocardial GLUT-4 and increased glucose uptake through activation of AMPK by
AICAR. Am J Physiol 277: H643–H649, 1999
32. Choi SL, Kim SJ, Lee KT, Kim J, Mu J, Birnbaum MJ, Soo Kim S, Ha J: The
regulation of AMP-activated protein kinase by H2O2. Biochem Biophys Res
Commun 287:92–97, 2001
33. Hayashi T, Wojtaszewski JFP, Goodyear LJ: Exercise regulation of glucose
transport in skeletal muscle. Am J Physiol 273:E1039 –E1051, 1997
34. Derave W, Lund S, Holman GD, Wojtaszewski J, Pedersen O, Richter EA:
Contraction-stimulated muscle glucose transport and GLUT-4 surface
content are dependent on glycogen content. Am J Physiol 277:E1103–
E1110, 1999
35. Franch J, Aslesen R, Jensen J: Regulation of glycogen synthesis in rat
skeletal muscle after glycogen-depleting contractile activity: effects of
adrenaline on glycogen synthesis and activation of glycogen synthase and
glycogen phosphorylase. Biochem J 344:231–235, 1999
36. Mu J, Brozinick JT Jr, Valladares O, Bucan M, Birnbaum MJ: A role for
AMP-activated protein kinase in contraction- and hypoxia-regulated glucose transport in skeletal muscle. Mol Cell 7:1085–1094, 2001
37. Wojtaszewski JF, Jorgensen SB, Hellsten Y, Hardie DG, Richter EA:
Glycogen-dependent effects of 5-aminoimidazole-4-carboxamide (AICA)riboside on AMP-activated protein kinase and glycogen synthase activities
in rat skeletal muscle. Diabetes 51:284 –292, 2002
38. Aschenbach WG, Hirshman MF, Fujii N, Sakamoto K, Howlett KF, Goodyear LJ: Effect of AICAR treatment on glycogen metabolism in skeletal
muscle. Diabetes 51:567–573, 2002
39. Hardie DG, Carling D, Carlson M: The AMP-activated/SNF1 protein kinase
subfamily: metabolic sensors of the eukaryotic cell? Annu Rev Biochem
67:821– 855, 1998
40. Zhou G, Myers R, Li Y, Chen Y, Shen X, Fenyk-Melody J, Wu M, Ventre J,
Doebber T, Fujii N, Musi N, Hirshman MF, Goodyear LJ, Moller DE: Role
of AMP-activated protein kinase in mechanism of metformin action. J Clin
Invest 108:1167–1174, 2001
41. Fryer LG, Parbu-Patel A, Carling D: The anti-diabetic drugs rosiglitazone
and metformin stimulate AMP-activated protein kinase through distinct
pathways. J Biol Chem 277:25226 –32, 2002
42. Minokoshi Y, Kim YB, Peroni OD, Fryer LG, Muller C, Carling D, Kahn BB:
Leptin stimulates fatty-acid oxidation by activating AMP-activated protein
kinase. Nature 415:268 –269, 2002
43. Frederich M, Balschi JA: The relationship between AMP-activated protein
kinase activity and AMP concentration in the isolated perfused rat heart.
J Biol Chem 277:1928 –1932, 2002
15