Originally published In Press as doi:10.1074/jbc.M107001200 on October 12, 2001
J. Biol. Chem., Vol. 277, Issue 2, 1514-1523, January 11, 2002
Glycogen-targeting Subunits and Glucokinase Differentially
Affect Pathways of Glycogen Metabolism and Their Regulation in
Hepatocytes*
Ruojing
Yang
§,
Liwei
Cao¶,
Rosa
Gasa§,
Matthew J.
Brady
,
A. Dean
Sherry¶, and
Christopher B.
Newgard§**
From the Departments of Biochemistry and Internal
Medicine, the § Touchstone Center for Diabetes Research,
and the ¶ Rogers NMR Center, University of Texas Southwestern
Medical Center, Dallas, Texas 75390 and the Department of
Internal Medicine, Section on Endocrinology, University of Chicago,
Chicago, Illinois 60637
Received for publication, July 24, 2001, and in revised form, September 28, 2001
 |
ABSTRACT |
Overexpression of the glucose-phosphorylating
enzyme glucokinase (GK) or members of the family of glycogen-targeting
subunits of protein phosphatase-1 increases hepatic glucose disposal
and glycogen synthesis. This study was undertaken to evaluate the functional properties of a novel, truncated glycogen-targeting subunit
derived from the skeletal muscle isoform
GM/RGl and to compare pathways of
glycogen metabolism and their regulation in cells with overexpressed
targeting subunits and GK. When overexpressed in hepatocytes, truncated
GM/RGl (GM
C) was approximately
twice as potent as full-length GM/RGl in
stimulation of glycogen synthesis, but clearly less potent than GK or
two other native glycogen-targeting subunits, GL and PTG.
We also found that cells with overexpressed GMC are
unique in that glycogen was efficiently degraded in response to
lowering of media glucose concentrations, stimulation with forskolin,
or a combination of both maneuvers, whereas cells with overexpressed
GL, PTG, or GK exhibited impairment in one or both of these
glycogenolytic signaling pathways. 2H NMR analysis of
purified glycogen revealed that hepatocytes with overexpressed GK
synthesized a larger portion of their glycogen from triose phosphates
and a smaller portion from tricarboxylic acid cycle
intermediates than cells with overexpressed glycogen-targeting subunits. Additional evidence for activation of distinct pathways of
glycogen synthesis by GK and targeting subunits is provided by the
additive effect of co-overexpression of the two types of proteins upon
glycogen synthesis and a much larger stimulation of glucose
utilization, glucose transport, and lactate production elicited by GK.
We conclude that overexpression of the novel targeting subunit
GMC confers unique regulation of glycogen metabolism. Furthermore, targeting subunits and GK stimulate glycogen synthesis by
distinct pathways.
| |
INTRODUCTION |
Hepatic glucose production is poorly controlled in type II
diabetes due to increased gluconeogenesis and impaired glycogen storage
(1-3). The glucose-phosphorylating enzyme glucokinase (GK)1 plays a key role in
determining the balance between glucose disposal and production in the
liver. Increased expression of GK in hepatoma cells (4), isolated
hepatocytes (5, 6), and livers of intact animals (7-10) potently
affects glucose disposal and glycogen deposition. Conversely,
overexpression of key components of the glucose-6-phosphatase enzyme
complex, which catalyzes glucose 6-phosphate hydrolysis, causes a sharp
reduction in glycogen storage in liver cells (11-13).
Although the control strength of GK in hepatic glucose metabolism is
substantial, it has become clear that it is also possible to stimulate
glycogen synthesis by expression of proteins that function distal to
the glucose phosphorylation step. In particular, recent studies have
highlighted an important role for glycogen-targeting subunits of
protein phosphatase-1 in spatial organization and regulation of
glycogen metabolism (14). Four members of a gene family encoding these
proteins are known. GM or RGl (hereafter referred to as GM/RGl) is expressed primarily
in striated skeletal muscle (15); GL is expressed primarily
in liver (16); and PTG (protein targeting to
glycogen) (17, 18) and PPPR6 (19) are expressed in a wide
range of tissues. These proteins bind to glycogen and protein
phosphatase-1 and have differential capacities for binding to glycogen
synthase, glycogen phosphorylase, and phosphorylase kinase (14-19).
Overexpression of these proteins in mammalian cells results in
stimulation of glycogen synthesis (17, 20), but with differential
potency and response to regulatory factors, as recently demonstrated in
a study comparing the effects of overexpressed PTG, GL, and
GM/RGl in isolated hepatocytes (21). Overexpressed GL was the most effective targeting subunit
for stimulation of glycogen synthesis, consistent with its superior capacity to activate glycogen synthase, and cells with overexpressed PTG were least responsive to forskolin as a glycogenolytic stimulus. Interestingly, cells with overexpressed GM/RGl
exhibited a modest increase in glycogen storage, but also responded to
forskolin by lowering glycogen to levels similar to those of control
cells. The relatively weak glycogenic effect of overexpressed
GM/RGl may be related to structural differences
between this targeting subunit and other family members, most notably
its long C-terminal tail containing a putative sarcoplasmic
reticulum-binding domain that is absent in other isoforms. The
brisk response to forskolin in
GM/RGl-overexpressing cells may occur via
protein kinase A-mediated phosphorylation of a serine in its protein
phosphatase-1-binding site; this protein kinase A consensus site is
lacking in targeting subunits other than
GM/RGl (14).
Increased expression of GK in livers of normal rats results in lowering
of blood glucose levels and increased glycogen deposition, but these
are accompanied by a large increase in circulating triglycerides and
fatty acids (8). PTG overexpression in livers of normal rats also
improves glucose tolerance, but in contrast to GK, does so without
perturbing lipid homeostasis (22). However, animals with increased
hepatic PTG expression exhibit very high liver glycogen levels after an
overnight fast (22), consistent with the poor response of
PTG-overexpressing hepatocytes to forskolin and glucagon (20, 21).
Furthermore, oral delivery of [13C]glucose to
PTG-overexpressing animals and analysis of glycogen/glucose by NMR
revealed that the majority of glycogen is synthesized by an indirect
pathway (e.g. a pathway other than glucose 
glucose-6-P glucose-1-P UDP-glucose glycogen).
This study was designed to address two fundamental questions raised by
the foregoing work. First, is it possible to design a
glycogen-targeting subunit of protein phosphatase-1 that has a potent
stimulatory effect on glucose disposal and glycogen deposition while
still allowing normal regulation of glycogenolysis in response to
catabolic signals? Second, are there differences in the metabolic fate
of glucose in cells with overexpressed glycogen-targeting subunits
relative to cells with overexpressed GK? In answer to these questions,
we show that a truncated form of GM/RGl lacking its long C-terminal tail has a significantly enhanced glycogenic effect
compared with native GM/RGl, but with retention
of effective glycogenolytic signaling. We have also evaluated pathways
of glycogen synthesis in living cells via administration of
2H2O and application of NMR (23) to identify
the 2H-labeled carbon atoms in the glucose molecules of
purified glycogen. This analysis revealed that overexpression of
glycogen-targeting subunits or GK in hepatocytes activates discrete and
complementary pathways of glucose disposal.
| |
MATERIALS AND METHODS |
Plasmids and Recombinant Adenoviruses--
To express a
truncated version of GM/RGl, PCR was used to
clone a 1.1-kb fragment of the GM/RGl cDNA
encoding amino acids 1-375 and lacking the 735 C-terminal amino acids
from a rabbit skeletal muscle cDNA library
(CLONTECH). PCR products were cloned into the pAMP1
vector (Life Technologies, Inc.), restricted with BamHI, and
subcloned into BamHI-restricted pGEX-KG (a gift from Dr.
Kun-Liang Guan, University of Michigan). The oligonucleotides used for amplification of the GM/RGl
fragment were 5'-CAUCAUCAUCAUGGATCCATGGACCCTTCTGAA-3' and
3'-CUACUACUACUAGAGCTCCAAGTCAGTGCATGC-5'. This fragment was designated
GMC and cloned into the adenoviral vector pACCMV-pLpA (24) to generate plasmid pYN1. Plasmid pYN10 containing
GMC-FLAG was generated by PCR amplification of the
rabbit GM/RGl cDNA (15) using
oligonucleotide primers
5'-GCTCTCTAGACCAGAGAGCCCAATGGAGCCTTCTGAAGTACCTGGTCAGAACAGCAAA-3' (5'-primer) and
5'-GCGCAAGCTTCTACTTGTCATCGTCGTCCTTGTAGTCAAGCCTTTGGGACAAGTCAGT-3' (3'-primer). The 3'-primer contains an XbaI site followed by
a stop codon and 24 bases encoding the FLAG sequence (25); the remaining 21 bases are homologous to the GM/RGl
cDNA sequence from nucleotides 1105 to 1125. The 5'-primer contains
a HindIII site followed by 47 bases homologous to the rabbit
GM/RGl cDNA sequence, including 12 bases of
5'-untranslated sequence. The resultant 1.1-kb PCR fragment was
restricted with HindIII and XbaI, subcloned
into plasmid pACCMV-pLpA to generate plasmid pYN10, and verified by
sequencing. Plasmid pYN11 containing GL-FLAG was generated
by PCR amplification of the rat GL cDNA (16, 21) using
oligonucleotide primers
5'-GCTCTCTAGAAACTTCTCTTCAGGCTCTCCCATGAGGCCAGCGAGCAGCGACCCCGGC-3' (5'-primer) and
5'-GCGCAAGCTTCTACTTGTCATCGTCGTCCTTGTAGTCGTAATAGGGCCCCAGCTTTTC-3' (3'-primer). The 3'-primer contains an XbaI site
followed by a stop codon and 24 bases encoding the FLAG sequence (25);
the remaining 21 bases are homologous to the GL cDNA
sequence from nucleotides 923 to 943. The 5'-primer contains a
HindIII site followed by 47 bases homologous to the rat
GL cDNA sequence from nucleotides 1 to 47 in the
5'-untranslated region of the cDNA (16). The resultant 1.1-kb PCR
fragment was restricted with HindIII and XbaI,
subcloned into plasmid pACCMV-pLpA to generate plasmid pYN11, and
verified by sequencing. Recombinant adenoviruses containing
GMC (AdCMV-GMC), GMC-FLAG
(AdCMV-GMC-FLAG), and GL-FLAG
(AdCMV-GL-FLAG) were generated by cotransfecting pYN1, pYN10, or pYN11 and plasmid pJM17 into 293 cells as described previously (26). Preparation and testing of recombinant adenoviruses expressing rat GL (AdCMV-GL), mouse PTG-FLAG
(AdCMV-PTG-FLAG), wild-type rabbit GM/RGl
(AdCMV-GM/RGl), rat liver glucokinase (AdCMV-GKL), and the Escherichia coli 
-galactosidase gene
(AdCMV-gal) have been described elsewhere (17, 20-22, 27, 28).
Cell Culture and Viral Treatment--
Primary hepatocytes were
isolated from 200-250-g Wistar rats after an overnight fast by
collagenase perfusion as described (29). Cells were suspended in
attachment medium (Dulbecco's modified Eagle's medium with 25 mM glucose, 10% fetal bovine serum, 100 nM
insulin, 100 nM dexamethasone, 2 mM sodium
pyruvate, and antibiotics) and plated onto collagen-coated plastic at a
density of 2.4 × 106 cells/60-mm culture dish,
1.2 × 106 cells/6-well dish, or 2 × 107 cells/150-mm dish at 37 °C for 1-3 h. Viruses were
diluted in culture medium (Dulbecco's modified Eagle's medium
containing 0.07% bovine serum albumin, 1 nM insulin, 10 nM dexamethasone, 2 mM sodium pyruvate,
antibiotics, and 2 mM glucose) and incubated at 37 °C
for 70 min. The virus-containing medium was removed, and the cells were
washed once with phosphate-buffered saline. Cells were incubated
further in culture medium containing various concentrations of glucose
for varying time periods as specified under "Results" and in the
figure legends. The titer of viral preparations was determined as
described (26), and a range of 2-200 × 106 pfu/ml
was used for hepatocyte gene transfer experiments as specified under
"Results" and in the figure legends.
Immunoblot Analysis--
Hepatocytes were homogenized in lysis
buffer (50 mM Tris-HCl, pH 7.5, 5 mM EDTA, 1%
Triton X-100, and proteinase inhibitors) by two rounds of
freeze-thawing. Cell lysates were centrifuged at 3000 × g for 5 min, and total protein concentration was measured by
the method of Bradford (30). Samples were resolved on
SDS-polyacrylamide gels and transferred to nitrocellulose membranes.
The membranes were incubated in blocking buffer (50 mM
Tris-HCl, pH 7.5, 150 mM NaCl, and 1% bovine serum
albumin) for 1 h and treated overnight at 4 °C with rabbit
polyclonal serum specific for GM/RGl (15) or
glucokinase (27) at dilutions of 1:1000 and 1:5000, respectively. The
membranes were washed and subsequently treated with horseradish peroxidase-labeled anti-rabbit IgG secondary antibody at 4 °C for
2 h. The FLAG epitope-tagged proteins were detected with a mouse
monoclonal antibody recognizing the FLAG peptide (Stratagene) diluted
1:1000 in blocking buffer, followed by treatment with horseradish
peroxidase-labeled anti-mouse IgG secondary antibody. The
protein-antibody complexes were visualized using an enhanced chemiluminescence detection kit (ECL, PerkinElmer Life Sciences).
Glycogen, Glucose, and Lactate Assays--
Glycogen was measured
by extraction in 10% trichloroacetic acid, precipitation with
methanol, and digestion of glycogen to free glucose by incubation with
0.4 mg/ml amyloglucosidase as previously described (31). Media glucose
and lactate levels were measured using kits from Sigma and normalized
to cell number.
Glucose Uptake--
Primary hepatocytes were treated
with the various recombinant adenoviruses or were left untreated for 70 min. Cells were incubated in tissue culture medium containing 12 mM glucose for 42 h and washed prior to assay of
glucose uptake. To measure the rate of glucose uptake, cells were
incubated for 3-360 min in culture medium containing 12 mM
glucose and 1 µCi/ml [2-3H]deoxyglucose (PerkinElmer
Life Sciences). Reactions were terminated by three washes with ice-cold
phosphate-buffered saline. Cells were collected and lysed using lysis
buffer as described above. A portion of the lysate was used for
measurement of total protein levels, and the remainder was used for
scintillation counting, with results expressed as cpm in the cell
lysate/mg of total protein. In another set of experiments, cells were
incubated in tissue culture medium containing 12 mM glucose
for 24 h and switched to medium containing 12 mM
glucose and 1 µCi/ml [2-3H]deoxyglucose for 18 h.
2-Deoxyglucose uptake was then measured as described above.
Analysis of Pathways of Glycogen Synthesis by 2H
NMR--
To analyze pathways of glycogen synthesis by application of
2H NMR, hepatocytes were prepared and treated with various
recombinant adenoviruses as described above. Following virus removal,
cells were incubated in medium containing various glucose
concentrations and 4% 2H2O for 42 h.
Following this incubation, glycogen was purified as described (31) and
converted to glucose by amyloglucosidase treatment. NMR analysis of
2H labeling of glucose was achieved by conversion of
glucose to monoacetone glucose by a modification of the method of
Landau et al. (32). Briefly, glucose samples (2-10 mg) were
vacuum-dried overnight and dissolved in 5 ml of acetone containing 4%
H2SO4. After stirring for 5 h at room
temperature, the pH was adjusted to 2.0 by addition of 1 N
NaOH, and the solution was stirred overnight at room temperature. The
samples were then further neutralized to pH 8.0 by addition of
Na2CO3 and lyophilized.
Monoacetone glucose was dissolved in 150 µl of 90% acetonitrile and
10% deuterium-depleted water. 2H NMR spectra were obtained
with a Varian Inova 14.1T spectrometer operating at 92.1 MHz and
equipped with a 3-mm broadband probe. 2H NMR spectra were
gathered at 50 °C using a 90° pulse and a 1-s acquisition time.
Typically, between 4000 and 50,000 scans were averaged, and the
resulting free induction decay was zero-filled to 4096 points and multiplied by a 1-Hz exponential prior to Fourier transformation. Peak areas were derived by a bayesian analysis of the
raw free induction decay (software provided by Varian). The relative
contributions of glucose, triose phosphates, and trichloroacetic acid
cycle intermediates to glycogen synthesis were calculated from
the relative H2, H5, and H6(S)
resonance areas using the following formulas: glycogen from glucose
(direct pathway): (H2 
H5)/H2; glycogen from the trichloroacetic
acid cycle (indirect pathway):
H6(S)/H2; and glycogen from triose
phosphates (indirect pathway): (H5 H6(S))/H2
| |
RESULTS |
Truncation of the C-terminal Tail of GM/RGl
Increases Its Glycogenic Effect in
Hepatocytes--
GM/RGl is highly homologous
to other glycogen-targeting subunits such as GL and PTG in
its N-terminal region, but is distinct from the other forms in that it
contains a long C-terminal tail that includes a putative sarcoplasmic
reticulum-binding domain (14, 15). It is possible that this unique C
terminus could contribute to the limited glycogenic effect of
overexpressed GM/RGl in hepatocytes relative to
that of other glycogen-targeting subunits. This idea was tested by
constructing a recombinant adenovirus encoding a truncated form of
GM/RGl (AdCMV-GMC) consisting of the 375 N-terminal amino acids with direct homology to other targeting subunit isoforms, but lacking the entire 735-amino acid C-terminal domain of native GM/RGl. Immunoblot analysis of
hepatocyte extracts following treatment with AdCMV-GMC
or AdCMV-GM/RGl revealed expression of the
truncated and wild-type GM/RGl proteins of the
anticipated molecular masses (Fig.
1A). Interestingly, the
truncated protein was expressed more efficiently than the full-length
targeting subunit, with approximately six times more
AdCMV-GM/RGl virus required than
AdCMV-GMC virus to achieve comparable levels of protein
expression. When normalized for protein expression, glycogen accumulation was consistently higher in
AdCMV-GMC-treated cells than in
AdCMV-GM/RGl-treated cells. For example,
glycogen content in cells treated with 5 × 106 pfu/ml
AdCMV-GMC was 76 µg/mg of protein compared with 30 µg/mg of protein in cells treated with 30 × 106
pfu/ml AdCMV-GM/RGl (Fig. 1A). These
results suggest that the presence of an intact C terminus in the
GM/RGl molecule inhibits its capacity to
stimulate glycogenesis in hepatocytes.
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 1.
Differential glycogenic potency of
overexpressed glycogen-targeting subunits and GK in hepatocytes.
Hepatocytes were prepared from fasted rats and treated with the
indicated titers of recombinant adenoviruses encoding various
glycogen-targeting subunits (PTG-FLAG, GL-FLAG,
GMC-FLAG, and GM/RGl) or GK.
After exposure to these viruses for 70 min, cells were cultured for
42 h in the presence of 20 mM glucose and then
harvested for immunoblot analysis of transgene expression and
measurement of glycogen content. Control hepatocytes (No
Virus) were cultured under the same conditions without adenoviral
treatment. A, immunoblot analysis with a polyclonal antibody
that detects both full-length GM/RGl and
GMC (15); B, immu- noblot analysis with an anti-FLAG antibody; C,
immunoblot analysis with an antibody that detects glucokinase (27). In
A-C, the lower panels show the glycogen content
of the same cells used for immunoblot analysis. Each panel is
representative of three independent experiments.
|
|
We next compared the glycogenic effect of the newly constructed
GMC molecule with that of other glycogen-targeting
subunits and GK. To compare expression levels of GMC,
GL, and PTG, FLAG-tagged versions of these proteins were
expressed using adenoviral vectors (Fig. 1B). Treatment of
hepatocytes with identical titers of AdCMV-GMC and
AdCMV-GMC-FLAG adenoviruses caused the same amount of
glycogen accumulation, showing that the presence of the FLAG tag had no influence on the metabolic properties of the molecule (data not shown).
When expressed at similar low levels, PTG was at least 10 times as
effective, and GL was nearly 20 times as effective at
stimulating glycogen accumulation as GMC
(e.g. 2 × 106 pfu/ml
AdCMV-GMC-FLAG versus 25 × 106 pfu/ml AdCMV-PTG-FLAG and 15 × 106
pfu/ml AdCMV-GL-FLAG, yielding 29, 370, and 580 µg of
glycogen/mg of protein, respectively) (Fig. 1B). At higher
doses of these viruses, glycogen accumulation became saturated, and
GMC-FLAG was expressed more effectively than PTG-FLAG or
GL-FLAG; but cells with the latter two constructs still
accumulated two to three times as much glycogen (Fig. 1B).
Finally, incubation of hepatocytes with increasing doses of AdCMV-GKL
caused a gradual increase in immunodetectable GK that was correlated
with increases in glycogen accumulation (Fig. 1C). Glycogen
content at the highest dose of GK virus was 350 µg/mg of protein,
which was less than the highest level achieved in
GL-overexpressing cells, equivalent to the highest level
achieved in PTG-overexpressing cells, and higher than the highest level
attained in GMC-overexpressing cells (200 µg/mg of
protein) (Fig. 1). In all subsequent studies, titers of the various
recombinant adenoviruses shown to cause maximum levels of glycogen
accumulation in Fig. 1 were used to minimize variability in results.
Hepatocytes with Overexpressed GMC Exhibit Enhanced
Glycogenic Potency with Retention of Forskolin-stimulated
Glycogenolysis--
We next compared the capacity of hepatocytes with
overexpressed PTG-FLAG, GK, GMC-FLAG, or
GM/RGl to respond to the glycogenolytic agent
forskolin. After treatment with the various viruses, cells were
cultured in 20 mM glucose and 1 nM insulin for
42 h, followed by an additional incubation in 20 mM
glucose without insulin in the presence or absence of 50 µM forskolin for 6 h. Cells cultured without
forskolin accumulated glycogen in an order consistent with the
data of Fig. 1 (PTG-FLAG, GK > GMC-FLAG > GM/RGl > -galactosidase; no-virus controls) (Fig.
2). Cells with overexpressed PTG-FLAG or
GK exhibited very small decreases in glycogen content in response to
forskolin (11 and 20%, respectively), whereas cells with overexpressed GMC-FLAG or GM/RGl degraded 55 and 78% of their glycogen, respectively, in response to the same
treatment. Furthermore, the absolute amount of glycogen degraded in
forskolin-treated GMC-overexpressing cells was higher
than that in cells overexpressing other targeting subunits (82 µg of
glycogen degraded in 6 h in GMC-overexpressing cells compared with 34, 50, and 47 µg/6 h in PTG-, GK-, and
GM/RGl-overexpressing cells, respectively).
These results indicate that the novel GMC construct has
improved glycogenic potency relative to native
GM/RGl, but with retention of sensitivity to
glycogenolytic effectors.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 2.
Differential activation of glycogenolysis by
forskolin in hepatocytes with overexpressed glycogen-targeting subunits
or GK. Hepatocytes were prepared from fasted rats, treated with
recombinant adenoviruses as described in the legend to Fig. 1, and
cultured for 42 h in the presence of 20 mM glucose.
Control hepatocytes were treated with a recombinant adenovirus
containing the -galactosidase (-Gal) gene
(AdCMV-gal) or in the absence of viral treatment (No
Virus). After the 42-h culture period, cells were switched either
to fresh culture medium containing 20 mM glucose and
lacking forskolin or to medium containing 20 mM glucose and
50 µM forskolin for a period of 6 h. Thereafter,
cells were collected for measurement of glycogen content. Data
represent the means ± S.E. for five independent experiments, each
performed in duplicate. * and **, significant differences between
forskolin-treated and untreated cells at p < 0.05 and
p < 0.01, respectively.
|
|
Glycogen-targeting Subunits and GK Alter Glycogenolytic Responses
to Hypoglycemia--
We also evaluated the capacity of cells
expressing glycogen-targeting subunits or GK to respond to a decrease
in glucose concentration. This was accomplished by culturing
hepatocytes treated with the various recombinant adenoviruses in 20 mM glucose for 42 h, allowing glycogen stores to
accumulate to high levels, followed by switching of the cells to medium
lacking glucose or containing 20 mM glucose for an
additional 24 h. Different batches of cells were collected before
and after the 24-h incubations for measurement of glycogen content. As
shown in Fig. 3A, all of the
cell groups increased their glycogen content with an additional 24 h of culture in 20 mM glucose relative to levels after
42 h. However, cells with overexpressed PTG or GL
exhibited very limited depletion of cellular glycogen stores in
response to removal of glucose from the medium (8 and 11%,
respectively). In sharp contrast, cells with overexpressed GMC or GK degraded 60 and 82% of their glycogen stores
when cultured in the absence of glucose. Additional experiments shown
in Fig. 3B revealed that GMC is unique with
regard to its responses to forskolin and low glucose. This is manifest
in two ways. First, the AdCMV-GMC-treated cells were the
only group in which glycogen decreased during the 24-h incubation in 20 mM glucose and 50 µM forskolin. Second,
GMC-overexpressing cells treated with the combined
glycogenolytic stimuli of forskolin and 0 mM glucose had
60% less glycogen than cells incubated in 0 mM glucose and in the absence of forskolin (Fig. 3, compare B and
A). In contrast, forskolin did not enhance the
glycogenolytic effect of 0 mM glucose in GK-overexpressing
cells. These experiments separated the molecules under study into three
categories. PTG and GL are powerful stimulators of glycogen
synthesis, but their overexpression results in poor glycogenolytic
signaling in response to forskolin or a fall in glucose. GK is also a
potent glycogenic agent and allows a robust glycogenolytic response to
a fall in glucose, but renders the cells insensitive to forskolin.
Finally, GMC is unique in that it combines the features
of moderate stimulation of glycogen synthesis with robust
responsiveness to both types of glycogenolytic signaling. The data in
Fig. 3 pertaining to PTG- and GL-overexpressing cells are
generally consistent with a previous study from our laboratory (21),
except that forskolin was found to be less effective as a
glycogenolytic signal in GL-overexpressing cells in this
work. The most likely explanation for this discrepancy is the longer time of preincubation in 20 mM glucose in this study (42 h
versus 24 h in the previous study), allowing a higher
level of GL overexpression and more complete saturation of
glycogen stores, making the effects of forskolin less obvious.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 3.
Differential activation of glycogenolysis by
lowering of media glucose concentrations in hepatocytes with
overexpressed glycogen-targeting subunits or GK. Hepatocytes were
prepared from fasted rats, treated with recombinant adenoviruses as
described in the legend to Fig. 1, and cultured for 42 h in the
presence of 20 mM glucose. A, after the 42-h
period, one group of cells was collected for measurement of glycogen
content (Start Point). A second group of cells
was cultured for an additional 24 h in medium containing 20 mM glucose without insulin (20 mM
Glucose), and a third group of cells was cultured for an
additional 24 h in medium lacking glucose or insulin (0 mM Glucose). Glycogen content was measured in the
latter two groups of cells after the 24-h culture period. B,
the same procedure was used as described for A, except that
the cell groups cultured for 24 h in 20 or 0 mM
glucose were also exposed to 50 µM forskolin for the full
24-h period prior to measurement of glycogen content. Data represent
the means ± S.E. for three independent experiments, each
performed in duplicate. * and **, significant differences between cells
in 0 mM glucose versus cells at the start point
at p < 0.05 and p < 0.01, respectively. ND, not detectable.
|
|
Overexpression of Glycogen-targeting Subunits and GK Modulates the
Glucose Dependence of Glycogen Synthesis--
We have previously
demonstrated that overexpression of PTG or GL in
hepatocytes stimulates glycogen synthesis even in the complete absence
of glucose in the culture medium (20, 21). We therefore compared the
glycogenic effects of overexpressed PTG with those of GK,
GMC, and full-length GM/RGl as a
function of glucose concentration. This was accomplished by incubating hepatocytes overexpressing the different glycogen-targeting subunits or
GK for 42 h in medium lacking glucose or containing 1, 5, 12, or
20 mM glucose. Consistent with our previous finding,
overexpression of PTG caused a large increase in glycogen content even
in the absence of glucose (Fig. 4). In
the absence of glucose, overexpression of GMC, GK, or
GM/RGl caused glycogen to accumulate to levels greater than those in control cells, but these values were only 31, 7, and 7% of those in PTG-overexpressing cells, respectively. As glucose
concentrations were raised to a maximum concentration of 20 mM, PTG-, GMC-, and
GM/RGl-overexpressing cells exhibited 12-, 20-, and 26-fold increases in glycogen content, respectively. Much larger
increases in glycogen content were observed as glucose concentrations
were raised in GK-overexpressing cells, such that total glycogen
content became equal in PTG- and GK-overexpressing cells cultured in 20 mM glucose, representing a 160-fold increase in
GK-expressing cells relative to their levels in 0 mM
glucose (Fig. 4). Thus, glycogen synthesis in GK-overexpressing cells is much more sensitive to changes in glucose concentration than in
cells with overexpressed glycogen-targeting subunits.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 4.
Glucose dependence of glycogen synthesis in
hepatocytes with overexpressed glycogen-targeting subunits or GK.
Hepatocytes were isolated from fasted rats and treated with recombinant
adenoviruses as described in the legend to Fig. 1. After viral
treatment, cells were incubated for 42 h in culture medium
containing 0, 1, 5, 12, or 20 mM glucose, after which they
were harvested for measurement of glycogen content. Data represent the
means ± S.E. for four independent experiments, each performed in
duplicate. Note that results are plotted on a log scale to accommodate
the very large changes in glycogen content occurring as a function of
glucose concentration in GK-overexpressing cells. For control cells,
glycogen content was measurable only at 12 or 20 mM
glucose.
|
|
Effects of GK and Targeting Subunit Overexpression on Glucose
Utilization, Glucose Uptake, and Lactate Production in
Hepatocytes--
The different glucose dependences of glycogen
synthesis in GK-overexpressing compared with glycogen-targeting
subunit-overexpressing hepatocytes suggest that the two types of
proteins activate glycogen synthesis via distinct mechanisms. To
investigate this further, we evaluated the effects of co-overexpression
of GK and glycogen-targeting subunits. As shown in Fig.
5, overexpression of PTG,
GMC, or GK caused large increases in glycogen content
compared with control hepatocytes. Co-overexpression of PTG + GK or
GMC + GK had additive effects on glycogen accumulation
relative to expression of the glycogen-targeting subunits alone. In
contrast, co-overexpression of PTG + GMC did not
increase glycogen levels relative to expression of PTG alone.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 5.
Additive glycogenic effects of GK and
glycogen-targeting subunit overexpression. Hepatocytes were
isolated from fasted rats and treated with individual recombinant
adenoviruses or combinations thereof, as indicated. After viral
treatment, cells were incubated overnight in medium containing 2 mM glucose, followed by a switch to 12 mM
glucose for 24 h, after which glycogen content was measured. Data
represent the means ± S.E. for three independent experiments,
each performed in duplicate. **, significant differences between the
PTG + GK versus PTG-alone groups and the GMC + GK versus GMC-alone groups, respectively,
at p < 0.01.
|
|
Another clear difference in metabolic function in glycogen-targeting
subunit-overexpressing versus GK-overexpressing cells is
illustrated in Fig. 6A. When
cultured in 12 mM glucose for 24 h, hepatocytes with
overexpressed PTG or GMC caused a modest (
1.2
mM) but significant decrease in media glucose
concentrations. In sharp contrast, cells with overexpressed GK lowered
media glucose concentrations from 12 mM to between 6 and 7 mM. The effect of GK was dominant in that cells with
co-overexpression of GK and either PTG or GMC exhibited
the same decrease in media glucose concentrations as cells with GK
overexpression alone. Consistent with these findings, cells with
overexpressed GK also increased their lactate production by 6-fold
relative to control cells or cells with overexpressed PTG or
GMC (Fig. 6B). Lactate output was slightly
but significantly lower in PTG-overexpressing cells relative to
controls and showed a trend toward being lower in cells with
overexpressed GK + PTG or GK + GMC relative to GK alone,
possibly indicating a "pull" or diversion of a small fraction of
glucose 6-phosphate from the glycolytic pathway to glycogen synthesis
in targeting subunit-overexpressing cells.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 6.
Differential effects of glycogen-targeting
subunits and GK on glucose consumption and lactate production.
Hepatocytes were isolated from fasted rats and treated with individual
recombinant adenoviruses or combinations thereof, as indicated. After
viral treatment, cells were incubated overnight in medium containing 2 mM glucose, followed by a switch to 12 mM
glucose for 24 h, after which media samples were taken for
measurement of glucose and lactate concentrations, normalized to cell
number. A, media glucose concentrations; B, media
lactate concentrations. Data represent the means ± S.E. for six
independent experiments, each performed in duplicate. * and **,
significant differences relative to cells not treated with virus
(None) at p < 0.002 and p < 0.001, respectively.
|
|
The decrease in media glucose concentrations caused by
glycogen-targeting subunit expression shown in Fig. 6A,
although significant, is modest relative to cells with overexpressed
GK. To compare the effects of targeting subunits and GK on glucose
transport in a more direct fashion, we measured uptake of
[2-3H]deoxyglucose in hepatocytes treated with the
various recombinant adenoviruses. As shown in Fig.
7A (left panel),
overexpression of GK caused a dramatic increase in the rate of
2-deoxyglucose uptake, such that by 360 min, it was 19-fold higher than
in untreated cells. As shown in Fig. 7A (right
panel), overexpression of PTG also caused an increase in the rate
of 2-deoxyglucose uptake relative to controls, but to a much smaller
extent. 2-Deoxyglucose uptake in control cells reached a plateau at 60 min, whereas it continued to increase to 360 min in PTG-overexpressing
cells, such that cumulative uptake was 70% higher in these cells at
this time point. Over a longer time period (18 h incubation with 12 mM [2-3H]deoxyglucose), overexpression of
GL, PTG, or GMC caused 6.6-, 4.4-, and
2.1-fold increases in 2-deoxyglucose uptake, respectively, relative to
AdCMV-gal-treated cells, an order consistent with the effects of
these molecules on glycogen synthesis (e.g. GL > PTG > GMC) (Fig. 7B). Again, GK
overexpression increased glucose uptake by a full order of magnitude
relative to any of the targeting subunits. In sum, the data presented
in Figs. 5-7 provide clear evidence for activation of distinct
pathways of glucose disposal in cells with overexpressed
glycogen-targeting subunits versus those with overexpressed
GK.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 7.
[2-3H]Deoxyglucose uptake in
hepatocytes with overexpressed glycogen-targeting subunits or GK.
Hepatocytes were isolated from fasted rats and treated with the
indicated recombinant adenoviruses for 70 min. A, cells were
incubated in tissue culture medium containing 12 mM glucose
for 42 h and then switched to medium containing 12 mM
glucose and 1 µCi/ml [2-3H]deoxyglucose for measurement
of glucose uptake over the indicated time periods, as described under
"Materials and Methods." Note the logarithmic scale used in the
left panel versus the linear scale used in the right
panel. Data represent the means ± S.E. for three independent
experiments. *, points at which 2-deoxyglucose uptake was greater in
PTG-overexpressing cells than in controls, with p < 0.02. Differences were significant at all time points in
GK-overexpressing cells versus controls. B, after
viral treatment, cells were incubated in tissue culture medium
containing 12 mM glucose for 24 h and switched to
medium containing 12 mM glucose and 1 µCi/ml
[2-3H]deoxyglucose for 18 h, over which time glucose
uptake was measured as described under "Materials and Methods."
Data represent the means ± S.E. for two to five independent
experiments per group. Note the logarithmic scale. All cell groups
expressing GK or targeting subunits were significantly different from
controls at p < 0.001.
|
|
2H NMR Discriminates between Pathways of Glycogen
Synthesis as a Function of Glucose Concentration--
Our finding that
GK overexpression simultaneously activates glycolytic flux and glycogen
deposition, whereas glycogen-targeting subunit overexpression appears
to have little effect on glycolysis, suggests that the pathways used
for glycogen synthesis may also be different in the two groups of
cells. This was investigated by incubation of hepatocytes in medium
containing 4% 2H2O, allowing the pathways by
which glycogen was formed to be traced in GK-overexpressing
versus glycogen-targeting subunit-overexpressing cells by
NMR analysis of glucose from purified glycogen. The use of isotopic
water to evaluate glucose metabolism is based on prior studies by
Landau et al. (32). The concept behind the method is that
2H2O will be incorporated into glucose at the
levels of the malate/fumarate equilibration step of the trichloroacetic
acid cycle (labeling carbon 6 of glucose), the triose-phosphate
isomerase step (labeling carbon 5 of glucose), or the hexose-phosphate
isomerase step (labeling carbon 2 of glucose). The relative labeling of
these carbons can be deduced by mass spectrometry (after chemical
degradation of the glucose) or, much more simply, from a single
2H NMR of glucose (23).
To test the method, hepatocytes were treated with AdCMV-PTG and
incubated in 0, 1, or 20 mM glucose for 40 h, at which
time glycogen was purified and degraded to glucose by incubation with amyloglucosidase, and the glucose was converted to monoacetone glucose
as described under "Materials and Methods." The initial trial was
performed with PTG-overexpressing hepatocytes because this facilitated
glycogen purification from cells incubated even in the absence of
glucose. 2H NMR spectra of monoacetone glucose derived from
glycogen isolated from the three groups of hepatocytes are shown in
Fig. 8A. Several points can be
made from simple inspection of these spectra. First, the 2H
labels on each of the six carbons of glucose were all well resolved and
clearly quantifiable. Second, the 2H resonances of
H6(R) and
H6(S) were also distinguishable; and
therefore, glucose derived from the level of fumarase in the trichloroacetic acid cycle could be quantified. Third, there were major
and obvious differences in the relative size of the peaks dependent
upon the media glucose concentration. For example, glycogen isolated
from cells incubated in 20 mM glucose had a large H2 peak
relative to all other positions, whereas the labeling at each position
was similar in glycogen isolated from cells incubated without glucose.
These data are consistent with a larger contribution of a "direct"
pathway of glycogen synthesis in 20 mM glucose, in which
glucose is converted to glycogen via glucose glucose-6-P glucose-1-P UDP-glucose glycogen. In this pathway, labeling at
H2 is expected as a result of rapid equilibration of glucose-6-P and
fructose-6-P in the hexose-phosphate isomerase reaction, whereas other
positions would remain unlabeled. In contrast, the relative increase in
labeling at H5 and H6(S) in cells
incubated in 0 or 1 mM glucose is consistent with a larger
contribution of gluconeogenic or "indirect" pathways of glycogen
synthesis in these cells, allowing 2H incorporation at the
triose phosphate level (H5) or in the trichloroacetic acid cycle
(H6(S)).
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 8.
2H NMR analysis of hepatocytes
overexpressing PTG at various glucose concentrations. Hepatocytes
were isolated from fasted rats and treated with AdCMV-PTG for 70 min.
Following viral treatment, cells were incubated with medium containing
4% 2H2O and 0, 1, or 20 mM glucose
for 42 h. At this point, glycogen was purified from extracted
cells and degraded to glucose by treatment with amyloglucosidase. The
glucose was then converted into monoacetone glucose, and 2H
NMR spectra were obtained as described under "Materials and
Methods." Data were further analyzed by the bayesian curve-fitting
program. A, representative 2H spectra for
glycogen/glucose purified from PTG-overexpressing hepatocytes incubated
in 0, 1, or 20 mM glucose. 2H labeling of
carbons 1-5 and the R and S positions of
carbon 6 are labeled H1, H2, H3,
H4, H5, H6(R), and
H6(S), respectively. B, relative
contributions of different pathways of glycogen synthesis (derived from
media glucose (black bars), triose phosphates (hatched
bars), and trichloroacetic acid cycle intermediates (open
bars)) in PTG-overexpressing cells cultured in 0, 1, or 20 mM glucose calculated as described under "Materials and
Methods." Data represent the means ± S.E. for 5, 7, and 11 independent experiments performed in 0, 1, and 20 mM
glucose, respectively.
|
|
Based on these spectra, the relative contributions of carbon coming
from the trichloroacetic acid cycle, the level of the triose
phosphates, or the level of hexose phosphates to glycogen synthesis can
be estimated using the formulas described under "Materials and
Methods." Fig. 8B provides such an estimate based on the
spectra of Fig. 8A for PTG-overexpressing hepatocytes
incubated in 0, 1, or 20 mM glucose. At 0 mM
glucose, >80% of the glycogen was synthesized from the level of
trichloroacetic acid cycle intermediates, whereas only a very small
portion came from the level of the triose phosphates or via the direct
pathway. As glucose was raised to 1 mM, the contribution of
the direct pathway increased by 4-fold, at the expense of a 14%
decrease in the contribution of trichloroacetic acid cycle
intermediates. Finally, at 20 mM glucose, a large increase in the relative contribution of the direct pathway was observed, such
that it contributed >60% of glycogen synthesis. Again, this increase
occurred at the expense of flux from the trichloroacetic acid cycle
intermediates, with little effect on the contribution of triose phosphates.
2H NMR Provides Direct Evidence of Activation of
Alternative Pathways of Glycogen Synthesis by Overexpressed GK and
Glycogen-targeting Subunits--
Having established the method, we
next compared the relative contributions of the different pathways of
glycogen synthesis in cells with overexpressed glycogen-targeting
subunits or GK. This analysis was performed in cells incubated in 1 or
20 mM glucose (Fig. 9,
A and B, respectively). As shown in Fig.
9A, incubation of PTG- or GMC-expressing
cells in 1 mM glucose resulted in virtually identical
labeling patterns, with the large majority of glycogen being derived
from trichloroacetic acid cycle intermediates. In contrast,
GK-overexpressing cells incubated in 1 mM glucose had larger contributions of the direct and triose phosphate pathways, with
a proportionally lower contribution of trichloroacetic acid cycle
precursors relative to cells with overexpression of either targeting
subunit. In all cell groups cultured in 20 mM glucose, the
direct pathway of glycogen synthesis was dominant (Fig. 9B). However, the remaining glycogen was derived in roughly equal proportion from trichloroacetic acid cycle intermediates and triose phosphates in
cells expressing glycogen-targeting subunits, whereas in
GK-overexpressing cells, the contribution of triose phosphates was more
than double that of trichloroacetic acid cycle intermediates. At the
higher glucose concentration (where more glycogen was made), it was
possible to include control cells for which no viruses were added.
Interestingly, these cells bore a closer resemblance to
GK-overexpressing cells than to the targeting subunit-overexpressing
groups, with a larger proportional contribution of triose phosphates
compared with trichloroacetic acid cycle intermediates to glycogen
synthesis. These data clearly show that glycogen-targeting subunits and
GK stimulate distinct pathways of glycogen synthesis, consistent with
the additive effects of co-overexpression of these two kinds of
proteins.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 9.
Comparison of 2H labeling
patterns of glycogen in hepatocytes with overexpressed
glycogen-targeting subunits or GK. Hepatocytes were prepared from
fasted rats, treated with various recombinant adenoviruses, and
cultured for 42 h in the presence of 4% deuterium and 1 mM (A) or 20 mM (B)
glucose. Relative contributions of different pathways of glycogen
synthesis were determined by 2H NMR analysis of purified
glycogen as described under "Materials and Methods" and in the
legend to Fig. 8. Data represent the means ± S.E. for four
independent experiments for each cell group studied in 1 mM
glucose and for three to five independent experiments for each cell
group studied in 20 mM glucose. *, significant differences
in the contribution of the indicated pathways of glycogen synthesis
between GK-overexpressing cells and all groups of glycogen-targeting
subunit-overexpressing cells, with p < 0.002.
|
|
| |
DISCUSSION |
Hepatic glycogen stores are depleted in all major forms of
diabetes (1-3). For example, activity-lowering mutations in the glucokinase gene cause maturity-onset diabetes of youth type II, and
patients experience a clear decline in liver glycogen stores (3).
Animals with experimental suppression of GK expression in liver exhibit
a similar deficit (33). The potential role of glycogen-targeting
subunits of protein phosphatase-1 in the pathogenesis of diabetes is
less clear (see Ref. 14 for review), in part because genetic analysis
of all of the different members of the gene family has not yet been
accomplished. However, overexpression of one family member, PTG,
stimulates glycogen synthesis and improves glucose disposal without
perturbation of lipid homeostasis in rodents (22). Thus, this study
continues a line of investigation in our laboratory in which GK and
glycogen-targeting subunits are being evaluated as potential
therapeutic targets for enhancing hepatic glucose disposal and lowering
blood glucose levels in diabetes (14, 20-22).
This study had two major goals. The first was to compare the metabolic
and regulatory properties of hepatocytes overexpressing a truncated
form of GM/RGl (GMC) with those
of cells with overexpressed GK or native glycogen-targeting subunit
isoforms. The second was to investigate the distinct pathways by which
GK and glycogen-targeting subunits stimulate glycogen synthesis in
liver cells. The accomplishment of these goals has led to the following
key findings. 1) Cells with overexpressed GMC accumulate
more glycogen than cells with overexpression of wild-type
GM/RGl, but not as much as cells with overexpressed PTG, GL, or GK. 2) The molecules tested can
be divided into three classes with regard to responses to
glycogenolytic signals. Cells with overexpressed PTG or GL
are poorly responsive to lowering of media glucose concentrations or
addition of forskolin. GK-overexpressing cells exhibit an intermediate
response, with brisk glycogenolysis in response to lowering of media
glucose concentrations, but with a poor response to forskolin and no
additive response to a combination of low glucose and forskolin.
Finally, cells with overexpressed GMC are unique in that
they activate glycogenolysis in response to lowering of media glucose
concentrations, stimulation with forskolin, or a combination of both
maneuvers. Thus, overexpression of GMC causes
significant stimulation of glycogen deposition, but with retention of
exquisite sensitivity to nutritional and pharmacological catabolic
signals. 3) The difference in sensitivity to glycogenolytic agents of
cells with overexpressed GMC relative to cells with
overexpressed PTG is not reflective of stimulation of different
pathways of glycogen synthesis. Thus, analysis of
2H2O labeling of glycogen/glucose shows
essentially identical patterns for cells with overexpressed
GMC or PTG at either glucose concentration tested, with
major contributions made by the direct pathway and flux from the level
of the trichloroacetic acid cycle and a smaller contribution from the
triose phosphates. 4) The labeling pattern just described for cells
with overexpressed targeting subunits is clearly different from that
for cells with overexpressed GK, with a larger contribution of flux
from the triose phosphates apparent in GK-overexpressing cells at
either glucose concentration. Additional evidence for activation of
distinct and complementary pathways of glycogen synthesis by GK and
targeting subunit overexpression is provided by the additive effect of
the two types of proteins on glycogen synthesis and the much more
dramatic stimulation of glucose utilization, glucose uptake, and
lactate production caused by overexpression of GK.
The mechanism underlying the enhanced glycogenic potency of
GMC versus native
GM/RGl is unknown. Possible explanations
include the following. 1) Removal of the C-terminal domain alters the nature of interactions of GM/RGl with protein
phosphatase-1 and/or with the protein phosphatase-1 substrates,
glycogen synthase, phosphorylase kinase, and glycogen phosphorylase. 2)
Removal of the C-terminal domain prevents interaction of
GM/RGl with cellular membranes, allowing more
of the protein to bind to glycogen. 3) A combination of these
possibilities may be the explanation. Consistent with possibility 2, the C-terminal region of GM/RGl contains a patch of hydrophobic amino acids that may mediate binding of the native
protein to membranes of the sarcoplasmic reticulum in muscle (34).
Glycogen particles accumulate around the sarcoplasmic reticulum in
muscle, suggesting that targeting of key enzymes to this site may be
facilitated/guided by GM/RGl. When the native GM/RGl protein is overexpressed in hepatocytes,
which lack sarcoplasmic reticulum, this may cause "mistargeting" of
GM/RGl to membranes or other cellular sites
that are non-ideal environments for glycogen synthesis. In this model,
removal of the C terminus of GM/RGl may allow
the molecule to target mainly to glycogen, enhancing its stimulatory
effect on deposition of the glucose polymer.
Another interesting feature of cells with overexpressed
GMC is their enhanced capacity for response to catabolic
stimuli relative to cells with overexpression of other targeting
subunits or GK. Native GM/RGl differs from
other targeting subunits in that it has two serine-containing consensus
sequences for protein kinase A-mediated phosphorylation. Indeed,
phosphorylation of one of these serines, residing within the known
protein phosphatase-1-binding site (serine 67 in the human
GM/RGl sequence), has been reported to be
stimulated by glycogenolytic agents such as forskolin, resulting in
disassociation of protein phosphatase-1 (35, 36). Of note is the fact
that truncated GMC and native
GM/RGl share the consensus protein kinase A
sites, and their overexpression in hepatocytes results in the same
brisk responses to forskolin (Fig. 2). Whether phosphorylation of these
sites is sufficient to explain the responses to both the nutritional
and pharmacological glycogenolytic signals remains to be tested.
This work clearly demonstrates that glycogen-targeting subunits and GK
stimulate glycogen accumulation by distinct mechanisms. In the
2H2O labeling experiments, the fundamental
difference in the glycogen labeling pattern between these groups of
cells was the relatively large contribution of carbon from the level of
triose phosphates (indicated by labeling of carbon 3 of glucose) in
GK-overexpressing cells, which was replaced by a larger portion of
labeling from the level of trichloroacetic acid cycle intermediates
(indicated by labeling of carbon 6) in cells with overexpressed
targeting subunits. Based on these findings and the data of Figs. 6 and 7 that demonstrate increased glucose consumption, glucose uptake, and
lactate output in GK-overexpressing cells, we propose the following
model to explain our results (see Fig.
10 for schematic summary). In cells
with overexpressed glycogen-targeting subunits, activation of glycogen
synthase exerts a pull on the glucose 6-phosphate pool, which in
turn activates gluconeogenesis from 2- and 3-carbon precursors and
uptake of glucose. Since the medium contained little or no glycerol,
fructose, or other immediate precursors of the triose phosphates, the
main precursors used for glycogen synthesis under these conditions are
likely to be those that pass through the mitochondrial compartment and
the fumarase reaction on their way to glucose 6-phosphate,
e.g. pyruvate, lactate, and amino acids, all of which are
abundant in the culture medium. In cells with overexpressed GK, the
glucose 6-phosphate pool is expanded by the increase in glucose
phosphorylation (6). A large portion of this glucose 6-phosphate is
pushed directly into glycogen, facilitated in part by the known
allosteric activation of glycogen synthase by this metabolic
intermediate. However, an even larger percentage of the pool enters
glycolysis, making the major contribution to the fall in media glucose
concentration and the rise in lactate. A portion of this flux
apparently represents metabolism of glucose 6-phosphate to the level of
the triose phosphates, where labeling at carbon 3 occurs, followed by
incorporation of these labeled precursors into glycogen. With the major
contributions of the direct (hexose phosphates glycogen) and
modified direct (hexose phosphates triose phosphates hexose
phosphates glycogen) pathways to glycogen synthesis in
GK-overexpressing cells, gluconeogenesis from the level of
trichloroacetic acid cycle intermediates is not stimulated.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 10.
Model for explaining activation of distinct
pathways of glycogen synthesis by GK and glycogen-targeting subunit
overexpression. A, effects of glycogen-targeting
subunit overexpression; B, effects of GK overexpression. See
"Discussion" for an explanation. G-6-P, glucose
6-phosphate; TCA, trichloroacetic acid.
|
|
What, then, are the implications of our findings for whole animal
physiology and diabetes therapy? First, our work establishes that most
of the glucose 6-phosphate produced by GK overexpression enters
glycolysis, with a far smaller portion contributing to glycogen
synthesis. Thus, although GK overexpression will clearly lower
circulating glucose concentrations, there will also be undesirable increases in lactate production and lipogenesis, consistent with our
previous findings (8). In previous studies, adenovirus-mediated overexpression of PTG in normal rats resulted in improved disposal of
an oral glucose load relative to controls, with no perturbation of
lipid homeostasis (22). However, these animals did not lower their
liver glycogen levels in response to fasting, in effect presenting with
a phenotype of glycogen storage disease. In this study, we have
described a novel, modified glycogen-targeting subunit
(GMC) that appears to provide substantial stimulation of
glycogen accumulation while also allowing the cells to remain sensitive
to two different kinds of catabolic signals. It will be of interest to
test this targeting isoform in the in vivo setting.
| |
ACKNOWLEDGEMENTS |
We are grateful to Kimberly Jones-Ross and
Paul Anderson for expert technical assistance, Drs. Guoxun Chen and
Danhong Lu for helpful discussion, and Dr. Anna A. DePaoli-Roach
(University of Indiana School of Medicine) for provision of the
anti-GM/RGl antibody used in this study.
| |
FOOTNOTES |
*
This work was supported by Grants P01-DK58398 (to
C. B. N. and A. D. S.) and P41-RR02584 (to A. D. S.) from the
National Institutes of Health.The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement" in
accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
**
To whom correspondence should be addressed: Touchstone Center for
Diabetes Research, Rm. Y8.212, University of Texas Southwestern Medical
Center, 5323 Harry Hines Blvd., Dallas, TX 75290. Tel.: 214-648-2930;
Fax: 214-648-9191; E-mail: newgard@utsw.swmed.edu.
Published, JBC Papers in Press, October 12, 2001, DOI 10.1074/jbc.M107001200
| |
ABBREVIATIONS |
The abbreviations used are:
GK, glucokinase;
pfu, plaque-forming units.
| |
REFERENCES |
| 1.
|
Magnusson, I.,
Rothman, D. L.,
Katz, L. D.,
Shulman, R. G.,
and Shulman, G. I.
(1992)
J. Clin. Invest.
90,
1323-1327
|
| 2.
|
Cline, G. W.,
Rothman, D. L.,
Magnusson, I.,
Katz, L. D.,
and Shulman, G. I.
(1994)
J. Clin. Invest.
94,
2369-2376
|
| 3.
|
Velho, G.,
Petersen, K. F.,
Perseghin, G.,
Hwang, J. H.,
Rothman, D. L.,
Pueyo, M. E.,
Cline, G. W.,
Froguel, P.,
and Shulman, G. I.
(1996)
J. Clin. Invest.
98,
1755-1761
|
| 4.
|
Valera, A.,
and Bosch, F.
(1994)
Eur. J. Biochem.
222,
533-539
|
| 5.
|
O'Doherty, R. M.,
Lehman, D. L.,
Seoane, J.,
Gomez-Foix, A. M.,
Guinovart, J. J.,
and Newgard, C. B.
(1996)
J. Biol. Chem.
271,
20524-20530
|
| 6.
|
Seoane, J.,
Gomez-Foix, A. M.,
O'Doherty, R. M.,
Gomez-Ara, C.,
Newgard, C. B.,
and Guinovart, J. J.
(1996)
J. Biol. Chem.
271,
23756-23760
|
| 7.
|
Ferre, T.,
Pujol, A.,
Efren, R.,
Bosch, F.,
and Valera, A.
(1996)
Proc. Natl. Acad. Sci. U. S. A.
93,
7225-7230
|
| 8.
|
O'Doherty, R. M.,
Lehman, D.,
Telemaque-Potts, S.,
and Newgard, C. B.
(1999)
Diabetes
48,
2022-2027
|
| 9.
|
Hariharan, N.,
Farrelly, D.,
Hagan, D.,
Hillyer, D.,
Arbeeny, C.,
Sabrah, T.,
Treloar, A.,
Brown, K.,
Kalinowski, S.,
and Mookhtiar, K.
(1997)
Diabetes
46,
11-16
|
| 10.
|
Niswender, K. D.,
Shiota, M.,
Postic, C.,
Cherrington, A. D.,
and Magnuson, M. A.
(1997)
J. Biol. Chem.
272,
22570-22575
|
| 11.
|
Seoane, J.,
Trinh, K.,
O'Doherty, R.,
Gomez-Foix, A. M.,
Lange, A. J.,
Newgard, C. B.,
and Guinovart, J. J.
(1997)
J. Biol. Chem.
272,
26972-26977
|
| 12.
|
Trinh, K.,
O'Doherty, R.,
Anderson, P.,
Lange, A. J.,
and Newgard, C. B.
(1998)
J. Biol. Chem.
273,
31615-31620
|
| 13.
|
An, J., Li, Y.,
van de Werve, G.,
and Newgard, C. B.
(2001)
J. Biol. Chem.
276,
10722-10729
|
| 14.
|
Newgard, C. B.,
Brady, M. J.,
O'Doherty, R. M.,
and Saltiel, A. R.
(2000)
Diabetes
49,
1967-1977
|
| 15.
|
Tang, P. M.,
Bondor, J. A.,
Swiderek, K. M.,
and DePaoli-Roach, A. A.
(1991)
J. Biol. Chem.
266,
15782-15789
|
| 16.
|
Doherty, M. J.,
Moorhead, G.,
Morrice, N.,
Cohen, P.,
and Cohen, P. T.
(1995)
FEBS Lett.
375,
294-298
|
| 17.
|
Printen, J. A.,
Brady, M. J.,
and Saltiel, A. R.
(1997)
Science
275,
1475-1478
|
| 18.
|
Doherty, M. J.,
Young, P. R.,
and Cohen, P. T.
(1996)
FEBS Lett.
399,
339-343
|
| 19.
|
Armstrong, C. G.,
Browne, G. J.,
Cohen, P.,
and Cohen, P. T.
(1997)
FEBS Lett.
418,
210-214
|
| 20.
|
Berman, H. K.,
O'Doherty, R. M.,
Anderson, P.,
and Newgard, C. B.
(1998)
J. Biol. Chem.
273,
26421-26425
|
| 21.
|
Gasa, R.,
Jensen, P. B.,
Berman, H. K.,
Brady, M. J.,
DePaoli-Roach, A. A.,
and Newgard, C. B.
(2000)
J. Biol. Chem.
275,
26396-26403
|
| 22.
|
O'Doherty, R. M.,
Jensen, P. B.,
Anderson, P.,
Jones, J. G.,
Berman, H. K.,
Kearney, D.,
and Newgard, C. B.
(2000)
J. Clin. Invest.
105,
479-488 |
TOP
ABSTRACT
INTRODUCTION
