0013-7227/02/$15.00/0
Printed in U.S.A.
Endocrinology 143(4):1512–1520
Copyright © 2002 by The Endocrine Society
Paradoxical Regulation of Sp1 Transcription Factor
by Glucagon
CHITHRA N. KEEMBIYEHETTY, ROSALIND P. CANDELARIA, GIPSY MAJUMDAR,
RAJENDRA RAGHOW, ANTONIO MARTINEZ-HERNANDEZ, AND SOLOMON S. SOLOMON
Veterans Affairs Medical Center, Research Services (C.N.K., R.P.C., G.M., S.S.S., A.M.-.H. and R.S.R.), and Departments of
Medicine (S.S.S.), Pharmacology (S.S.S., R.S.R.), and Pathology (A.M.-H.), University of Tennessee Health Science Center,
Memphis, Tennessee 38104
Insulin is a potent regulator of Sp1 transcription factor. To
examine if glucagon, which usually antagonizes insulin, regulates Sp1, we assessed the levels of Sp1 by Western blotting
from H-411E cells exposed to glucagon with or without insulin.
Glucagon alone (1.5 ⴛ 10ⴚ9 to 1.5 ⴛ 10ⴚ5 M) stimulated Sp1
accumulation but inhibited insulin’s (10,000 ␮U/ml) stimulatory effect on Sp1. We also assessed the effect of TNF-␣, wortmannin, a PI3K inhibitor, and cAMP-dependent protein kinase inhibitor on Sp1 accumulation. While TNF-␣ (5 ng/ml)
blocked insulin-stimulated Sp1, it failed to block stimulation
of Sp1 by glucagon (1.5 ⴛ 10ⴚ5 M). Similarly, wortmannin inhibited insulin- but not glucagon-stimulated Sp1, whereas
I
NSULIN HAS A broad range of physiological actions and
plays a major role in diabetes mellitus. Despite much
research, the signal transduction pathways involved in insulin action have not been completely elucidated (1). Past
studies from our laboratories have shown reduced amount
and activity of calmodulin (CaM), and low Km cAMP phosphodiesterase (PDE) activity in tissues from diabetic rats.
Treatments with insulin restored these activities to normal
(1a–3). Extension of these studies with a minimal deviant rat
hepatoma cell line (H-411E) showed that CaM gene transcription is regulated by insulin (4). Furthermore, we found
an obligatory role for the transcription factor Sp1 in both
basal and insulin stimulated CaM gene expression (5). A key
role for Sp1 in CaM gene transcription has been demonstrated by DNase I foot-printing, EMSA, Western blot, and
transfection experiments (6 – 8).
The interrelationship between insulin and Sp1 and its role
in CaM gene expression lead us to consider the role of glucagon in this system. Glucagon opposes insulin in most physiologic settings. Glucagon binds to a cell surface receptor and
initiates its effects, primarily through cAMP-dependent
mechanisms. For example, insulin and glucagon oppose each
other in the maintenance of blood glucose concentrations.
Similarly, glucagon raises intracellular cAMP, whereas insulin lowers it profoundly (9). Although insulin deficiency
remains the physiologic hallmark of diabetes, it is thought
that the remaining unopposed actions of glucagon contribute
significantly to the diabetic state by elevating the blood glucose and triggering both lipolysis and ketosis through cAMPmediated processes (9, 10). High circulating plasma glucagon
Abbreviations: CaM, Calmodulin; IRS-1, insulin receptor substrate 1;
PDE, phosphodiesterase; PKI, protein kinase inhibitor.
protein kinase inhibitor had an opposite effect. Thus, insulin
acts primarily via PI3K, and glucagon apparently stimulates
through a cAMP-dependent pathway. Insulin increased the
staining intensity of Sp1 seen exclusively in the nuclei of
H-411E cells. Sp1 was demonstrable in both nucleus and cytoplasm after glucagon treatment. Finally, as judged by immunoblotting to specific antibody, insulin but not glucagon,
stimulated O-glycosylation of Sp1. Thus, unique signaling
mechanisms mediate the response of Sp1 to glucagon in the
presence or absence of insulin. (Endocrinology 143: 1512–1520,
2002)
and cAMP levels are associated with diabetic ketoacidosis.
Considering the critical role of both hormones in the diabetic
process, we decided to systematically examine the regulation
of Sp1 in response to insulin and glucagon, alone and together, and try to elucidate mechanisms involved.
Materials and Methods
Chemicals
Insulin, glucagon, cAMP-dependent protein kinase inhibitor (PKI),
wortmannin, 8-Br ATP, and protein standard were obtained from
Sigma (St. Louis, MO). TNF-␣ was from R&D Systems (Minneapolis,
MN). Protease inhibitors were from Roche Molecular Biochemicals
(Mannheim, Germany). Rabbit polyclonal anti-Sp1 antibody was obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Mouse
monoclonal anti-O-linked GlcNAc antibody was obtained from
Affinity BioReagents, Inc. (Golden, CO). Mouse monoclonal antiactin
antibody was purchased from Chemicon Inc. (Piscataway, NJ). Protein A-Sepharose, Bio-Rad protein assay reagent and Kaleidoscope
protein molecular weight marker were obtained from Bio-Rad
Laboratories, Inc. (Hercules, CA).
Cell culture
Minimal deviant H-411E rat hepatoma cells were obtained from
ATCC (Manassas, VA) and were grown in Eagle’s MEM supplemented
with 1% glutamine, 1% nonessential amino acids, 1% streptomycin/
penicillin and serum (10% calf serum, 5% FBS). Cell cultures were
incubated at 37 C in 5% CO2 and 95% air using tissue culture flasks or
Petri dishes.
Experimental procedure for immunoblotting
Cells were cultured in 60 ⫻ 15-mm sterilized Petri dishes until 70 –
80% confluency. Before the experimental treatments, growth medium
was changed to serum-free media (Eagle’s MEM, 1% glutamine, nonessential amino acids, and antibiotics) for 36 – 40 h. Cell plates were
treated (in triplicate or duplicate) as outlined for each experiment, and
1512
Keembiyehetty et al. • Glucagon Regulates Sp1
experiments were repeated at least three times. At the end of the experimental treatments, total protein was extracted from the cells as
described previously (7, 8). Briefly, cells were washed twice with phosphate buffer, and RIPA buffer [1⫻ phosphate buffer, 1% igepal CA-230
(Sigma), 0.5% sodium deoxycholate, 0.1% SDS], 0.5 mm phenylmethylsulfonyl fluoride, 0.5 mm dithiothreitol, 1.0 mm sodium orthovanadate,
aprotinin, and protease inhibitor cocktail was added. Cells were scraped
and collected into an Eppendorf tube (Brinkman, Westbury, CT) and
passed through a 21G needle to disrupt the cells and the resultant
material homogenized. This procedure was conducted over ice. Extracts
were further kept on ice for 30 – 60 min and centrifuged at 10,000 ⫻ g for
10 min. The supernatant was collected and protein quantitated, using
Bio-Rad Laboratories, Inc. protein assay.
For immunoprecipitation reaction, 500 ␮g of protein were added to
4 ␮g of anti Sp1 antibodies in binding buffer (10 mm Tris HCl, pH 7.9,
2 mm MgCl2, 0.15 mm NaCl, 1 mm dithiothreitol, 10% glycerol, and 1 mm
phenylmethylsulfonyl fluoride) at a final concentration of 1 ␮g of protein
per ␮l and incubated at 4 C overnight. Protein A-Sepharose was then
added and the mixture was incubated at 4 C on a rocker platform for 2 h.
The antibody-protein A complexes were pelleted by centrifugation
(1,000 ⫻ g) and washed four times with binding buffer. The pellets were
resuspended in 1⫻ SDS sample buffer, boiled and analyzed by SDSPAGE (7, 8).
Proteins from the above procedures were separated using 7.5% SDSPAGE, loading equal amounts of protein from each sample. After electrophoresis, the protein was transferred to a Immobilon-P transfer membrane (Millipore Corp., Bedford, MA) using a Trans-Blot electrophoretic
transfer cell (Bio-Rad Laboratories, Inc.). Western-blot analyses were
conducted using either rabbit polyclonal anti-Sp1 antibody (1:5000) or
monoclonal anti-O-linked GlcNAc antibody (1:1000). Chemiluminescent
signal was developed using detection reagents from an enhanced chemiluminescence kit (ECL Plus, Amersham Pharmacia Biotech, Piscataway,
NJ) and exposed to x-ray film to obtain the signal. The blots probed with
anti-GlcNAc antibody were stripped and reprobed with anti-Sp1 antibody to confirm the presence of Sp1 protein and to validate that equal
amounts of total protein were loaded in each lane. The other blots were
stripped and probed again with antiactin antibody (1:10,000) to validate
the equivalent loading of proteins in each lane. After analysis of the
individual gels for Sp1, data from these experiments were collected and
statistical analysis performed with reference to actin controls.
Endocrinology, April 2002, 143(4):1512–1520 1513
scribed in previous experiments. In these experiments, the cells had been
treated for 24 h in serum-free media and then with hormone alone or
hormone plus selected inhibitor overnight (12–14 h). Concentrations of
agents used were: insulin 10,000 ␮U/ml; glucagon, 1.5 ⫻ 10⫺5 m; TNF-␣,
5 ng/ml; wortmannin, 1 ␮m; and PKI, 10 ␮g/ml. Cells were washed with
PBS and immediately fixed with 10% neutralized formalin. Immunocytochemical staining was conducted with Sp1 antibody (11, 12). Signal
was developed by horseradish peroxidase conjugated second antibody
employing 3,3⬘ diaminobenzidine tetrahydrochloride-plus kit from
Zymed Laboratories, Inc. (San Francisco, CA).
Statistical analysis
Bands in the x-ray films were scanned and quantified using the
Quantity One software program from Bio-Rad Laboratories, Inc. in a
Macintosh G-3 computer. Mean, sd, se, and t tests were calculated using
the Excel program. These data were then grouped and analyzed statistically as shown.
Results
Effects of insulin and glucagon on Sp1
H-411E cells in culture were exposed to various doses of
insulin for time intervals as indicated and cell extracts were
Dose- and time-response studies
Quiescent, 70 – 80% confluent, H-411E cells were treated overnight
(12–16 h) with different doses of insulin and glucagon. Insulin doses
were 100, 1,000, 10,000, and 100,000 ␮U/ml, whereas glucagon doses
were 1.5 ⫻ 10⫺9, 1.5 ⫻ 10⫺8, 1.5 ⫻ 10⫺7, 1.5 ⫻ 10⫺6, 1.5 ⫻ 10⫺5 m. Another
study was conducted with 10,000 ␮U/ml (0.06 ␮mol/liter) insulin together with each glucagon dose. Insulin time-response studies were
conducted at 1, 2, 4, and 6 h using 10,000 ␮U/ml.
Signaling pathway studies with inhibitors
Several experiments were conducted with quiescent H-411E cells
exposed to selected inhibitors. Cell plates were exposed to wortmannin
(0.01, 0.1 ␮m), TNF-␣ (5 ng/ml), or cAMP-dependent PKI (0.01, 0.1, 1
mg/ml) for 1 h. These samples were then incubated together with
respective inhibitor plus insulin (10,000 ␮U/ml) or glucagon (1.5 ⫻ 10⫺5
m) for an additional 4 h.
mRNA analysis
Cells were treated with insulin and glucagon alone and together,
similar to the previous experiments. Total RNA was collected and
mRNA analysis conducted using a probe for CaM as described previously (4).
Immunocytochemical staining
Cells were grown in RS-coated glass chamber slides (Nalge Nunc
International, Naperville, IL). After achieving 60% confluency, cells were
treated overnight (12–16 h) with serum-free media containing combinations of insulin, glucagon, and signaling pathway inhibitors as de-
FIG. 1. The effect of dose and time of insulin (Ins) exposure on synthesis of Sp1 transcription factor in H-411E cells in tissue culture. A,
Western immunoblots of dose response in duplicate with actin control,
in response to increasing insulin concentration (one of three replicate
studies). Exposure is overnight (12–16 h). B, Western immunoblot of
Sp1, time study (one of three replicate studies). C, Time study from
multiple experiments using 10,000 ␮U/ml insulin, data are normalized. Number on the bar indicates replicates. Values shown as mean ⫾
SEM. *, Statistical significance from control at P ⬍ 0.01.
1514
Endocrinology, April 2002, 143(4):1512–1520
Keembiyehetty et al. • Glucagon Regulates Sp1
FIG. 2. The effect of glucagon (Glu), 1.5 ⫻ 10⫺5 M alone and
of glucagon ⫹ insulin (Ins), 10,000 ␮U/ml, together on Sp1
transcription factor extracted from H-411E cells in tissue
culture as outlined in the text. Exposure is overnight (12–16
h). A, Western blots of Glu (1.5 ⫻ 10⫺5 M) and Ins (10,000
␮U/ml) ⫹ Glu at increasing concentration of (1.5 ⫻ 10⫺8 to
1.5 ⫻ 10⫺5 M). B and C, Densitometrically analyzed values
of Western blots from multiple experiments, normalized to
control. Number on the bar indicates the number of replicates. Values shown as mean ⫾ SEM. *, Statistical significance from control at given P value.
analyzed by Western blotting. As shown in Fig. 1A, Sp1
levels were enhanced by insulin in a dose-dependent manner. Figure 1B illustrates a representative immunoblot showing that 10,000 ␮U/ml insulin enhances steady-state levels of
Sp1 in a time-dependent manner. Figure 1C demonstrates the
cumulative quantitative analysis of Sp1 in response to insulin. The levels of Sp1 increase after insulin treatment, and
peak at 4 h (P ⬍ 0.01).
Like insulin, glucagon stimulates Sp1 in a time- and dosedependent manner with time peak at 4 h, and dose peak at
1.5 ⫻ 10⫺5 m glucagon (data not shown). Data shown in Fig.
2A demonstrate that glucagon alone also stimulates Sp1, but
not as strongly as insulin. Furthermore, when glucagon and
insulin are present together, glucagon inhibits the enhancement of Sp1 possibly at lower concentrations of glucagon, but
clearly at 1.5 ⫻ 10⫺5 m. The effect of glucagon alone on Sp1
was quantified from immunoblots and is graphically demonstrated in Fig. 2B, where glucagon alone stimulates Sp1 by
approximately 100%, P ⬍ 0.002. In contrast, in Fig. 2C when
insulin (10,000 ␮U/ml) and glucagon (1.5 ⫻ 10⫺5 m) are
incubated together in the cells, glucagon inhibits the effect of
insulin on Sp1 by approximately 50%, P ⬍ 0.0002.
Effects of signal transduction inhibitors on Sp1
Wortmannin is an inhibitor of PI3K, thought to be the first
signal in many of insulin’s actions. Figure 3A (lower panel)
illustrates the ability of wortmannin to inhibit insulin’s stimulation of Sp1. In contrast, data in Fig. 3A (upper panel) demonstrate that glucagon stimulation of Sp1 is not inhibited by
wortmannin. Figure 3B illustrates data from multiple experiments demonstrating that insulin stimulates Sp1 synthesis
(P ⬍ 0.02 compared with basal) and wortmannin inhibits
(P ⬍ 0.02) this effect. In Fig. 3, A and B, glucagon alone also
stimulates Sp1 (P ⬍ 0.001), but here wortmannin fails to
FIG. 3. The effect of wortmannin (Wort or W), 0.01 ␮M, 0.1 ␮M, 1.0 ␮M
on insulin (Ins or I), 10,000 ␮U/ml and/or glucagon (Glu or G), 1.5 ⫻
10⫺5 M on stimulation of Sp1. Cells were exposed to wortmannin for
1 h and hormone ⫹ wortmannin for an additional 4 h. A, Western
immunoblot from a single representative study for glucagon; Western
immunoblot of single representative study for insulin (duplicate actin
controls). B, Densitometrically analyzed values of Western blots from
multiple experiments are normalized and plotted as percent control.
Number on the bar indicates the number of replicates. Values shown as
mean ⫾ SEM. *, Statistical significance from control at given P value.
Keembiyehetty et al. • Glucagon Regulates Sp1
Endocrinology, April 2002, 143(4):1512–1520 1515
FIG. 4. The effect of TNF-␣ on insulin (10,000 ␮U/ml)
and/or glucagon (1.5 ⫻ 10⫺5 M) stimulation of Sp1. Cells
were exposed to TNF-␣ for 1 h and hormone ⫹ TNF-␣ for
an additional 4 h. TNF-␣ is presented as T when in combination with hormone. A, Western immunoblot from a
single representative experiment (duplicate treatments).
B, Densitometrically analyzed values of Western blots from
multiple experiments normalized to the control. Number on
the bar indicates the number of replicates. Values shown as
mean ⫾ SEM. *, Statistical significance from control at given
P value. **, Statistical significance between Ins and
Ins ⫹ T.
inhibit the enhanced accumulation of Sp1 in response to
glucagon (glucagon vs. glucagon ⫹ wortmannin, not
significant).
TNF-␣ is known to antagonize the action of insulin in vitro
and in vivo (13, 13a). Our earlier findings that the enhanced
expression of the CaM gene in response to insulin could be
reversed by TNF-␣ are consistent with the proposed antagonism between TNF-␣ and insulin (4, 5). Therefore, it was of
interest to determine whether TNF-␣ altered the enhanced
steady-state levels of Sp1 elicited by insulin and/or glucagon. A representative immunoblot to assess Sp1 levels in
extracts of H-411E cells exposed to either insulin or glucagon
in the absence or presence of TNF-␣ is shown in Fig. 4A.
TNF-␣ treatment significantly inhibited the effect of insulin
on Sp1 (Ins vs. con, P ⬍ 0.04; Ins vs. Ins ⫹ TNF-␣, P ⬍ 0.04).
In contrast, TNF-␣ had a negligible effect on Sp1 levels in
glucagon-treated cells (Fig. 4A). Quantification of Sp1 levels
from multiple experiments (Fig. 4B) corroborated the visual
data shown in Fig. 4A.
In light of the apparent distinction between the actions of
insulin and glucagon revealed by TNF-␣, we sought to elucidate further mechanistic basis for these results. As depicted
in Fig. 3, the effect of insulin but not of glucagon on Sp1 could
be abrogated by inhibitors of PI3K. Initially, we extended
these data with inhibitors of cAMP, a second messenger
needed for glucagon signaling. Our initial attempts to inhibit
cAMP levels were equivocal at best (data not shown). Therefore, we tested a powerful inhibitor, cAMP-dependent PKI.
Treatment of H-411E cells with PKI reduced the effect of glucagon but not that of insulin on Sp1 levels in a dose-dependent
manner (data not shown). Quantification of these data indicate
that 10 ␮g/ml PKI inhibited glucagon-stimulated Sp1 levels by
45% (P ⬍ 0.03) (Fig. 5B).
Immunocytochemical studies
Consistent with our published observations (8), immunoreactive Sp1 could be readily seen in the nuclei of the
FIG. 5. The effect of cAMP-dependent PKI, 0.01, 0.1, 1.0 mg/ml, on
glucagon (GLU), 1.5 ⫻ 10⫺5 M, stimulated Sp1 synthesis. Cells were
exposed to PKI for 1 h and hormone ⫹ PKI for an additional 4 h. A,
Western immunoblot from a single representative experiment. B,
Densitometrically analyzed values of Western immunoblots normalized, from multiple experiments. Number on the bar indicates replicates. Values shown as mean ⫾ SEM. *, Statistical significance from
GLU at given P value.
untreated control H-411E cells and treatment with insulin
dramatically increased the nuclear staining intensity for
Sp1 (Fig. 6, a and b). Though, glucagon-stimulated cells
(Fig. 6c) demonstrated increased nuclear staining intensity
1516
Endocrinology, April 2002, 143(4):1512–1520
Keembiyehetty et al. • Glucagon Regulates Sp1
FIG. 6. Immunocytochemical staining with anti-Sp1 antibodies of cultured H-411E cells. Cells have been treated with insulin (10,000 ␮U/ml)
or glucagon (1.5 ⫻ 10⫺5 M) overnight (12–16 h). a, Unstimulated cells (control). There is weak nuclear staining in all cells. Magnification, ⫻100.
b, Insulin-treated cells. There is intense nuclear staining in all cells. The staining is strictly confined to the nucleus with only an occasional
cell (⬍1%) demonstrating some cytoplasmic reactivity. Magnification, ⫻100. c, Glucagon-treated cells. There is increased nuclear staining in
all cells, although of lesser intensity than in insulin treated cells. However, more than 10% of the cells have both nuclear as well as diffuse
cytoplasmic staining (arrows). Magnification, ⫻100. d, Detail of glucagon-treated cells demonstrating both nuclear and cytoplasmic staining.
Magnification, ⫻400.
for Sp1 antigens, as compared with control cells (Fig. 6a),
the increase was not as pronounced as in insulin-stimulated cells (Fig. 6b). On a subjective scale, control cells were
2⫹, glucagon stimulated cells 3⫹, and insulin stimulated
cells 4⫹. It is worth pointing out at this stage, that the
visible perception of staining intensity is logarithmic,
therefore a 10-fold change is required to perceive a clear
difference (11, 12). We also observed a noticeable presence
of Sp1 antigens in the cytoplasm of a small percentage
(10%) of glucagon stimulated cells (Fig. 6, c and d). The
cytoplasmic staining for Sp1 in glucagon-treated cell cultures was highly reproducible and was rarely seen in
H-411E cells stimulated with insulin.
Data shown in Fig. 7 demonstrate immunocytochemical
staining of Sp1 antigens in H-411E cells after exposure to
insulin or glucagon, in the presence of three known in-
hibitors of signal transduction, wortmannin, PKI, and
TNF-␣. Insulin (Fig. 7, a1) and glucagon (Fig. 7, b1) both
produced significant increases in Sp1 staining in cells (see
Fig. 6) compared with control (Fig. 7, c1). Wortmannin
inhibited insulin-induced Sp1 staining significantly
(Fig. 7, a2) but was less effective in blocking glucagoninduced Sp1 staining (Fig. 7, b2). In contrast, PKI, a cAMP
inhibitor failed to inhibit insulin-induced Sp1 staining
(Fig. 7, a3) but was very effective at inhibiting glucagoninduced Sp1 staining (Fig. 7, b3). TNF-␣ not only markedly
inhibited the stimulatory effects of insulin (Fig. 7, a4) and
to a slightly lesser extent glucagon (Fig. 7, b4) but brought
the Sp1 staining levels down, below those of control cells
(Fig. 7, c1). When both insulin and glucagon are studied
together (Fig. 7, c2), glucagon inhibits insulin-stimulated
Sp1 staining.
Fig. 7. Immunohistochemical staining of cultured H-411E cells with anti-Sp1 antibodies and counter-stained with hematoxylin; the presence
of Sp1 antigens results in brown staining, its absence in blue staining (see Materials and Methods). The cultured cells have been exposed to
different combinations of hormones and inhibitors. Figures labeled (a) depict cells exposed to insulin (10,000 ␮U/ml) overnight (12–16 h); figures
labeled (b) depict cells exposed to glucagon (1.5 ⫻ 10⫺5 M) overnight (12–16 h); and figures labeled (c) depict controls. Upper panels represent
low magnification (⫻80); lower panels depict higher magnification (⫻600). a1, Cells exposed to insulin alone show the most intense nuclear
staining of all the conditions tested. a2, Cells exposed to insulin and wortmannin (1 ␮M); there is a clear decrease in the staining intensity as
compared with insulin alone (a1). a3, Cells exposed to insulin and PKA inhibitor (10 ␮g/ml); there is no clear effect on Sp1staining with all nuclei
Keembiyehetty et al. • Glucagon Regulates Sp1
Endocrinology, April 2002, 143(4):1512–1520 1517
remaining strongly positive. a4, Cells exposed to insulin and TNF-␣ (5 ng/ml); there is almost complete disappearance of any demonstrable Sp1
antigen. Therefore, in the absence of brown staining for Sp1, there is emergence of hematoxylin staining (blue). b1, Cells exposed to glucagon
alone; although the staining intensity is less than in cells treated with insulin (a1), there is intense nuclear staining and occasional cytoplasmic
staining. b2, Cells exposed to glucagon and wortmannin; there is no clear effect in nuclear or cytoplasmic staining intensity as compared with
glucagon alone (b1). b3, Cells exposed to glucagon and PKA inhibitor; compared with glucagon alone (b1), there is a significant decrease in
staining intensity. b4, Cells exposed to glucagon and TNF-␣; there is almost complete disappearance of demonstrable Sp1 antigen. The inhibition
of the glucagon effect by TNF-␣ is present but weaker than the insulin effect (a4). c1, Cultured cells exposed to neither insulin nor glucagon
have a less intense staining than those stimulated by either insulin or glucagon. c2, Cultured cells exposed to insulin and glucagon concomitantly
show lower intensity of Sp1 staining compared with insulin alone (a1), demonstrating the antagonistic effects of these hormones.
1518
Endocrinology, April 2002, 143(4):1512–1520
Keembiyehetty et al. • Glucagon Regulates Sp1
Effects of insulin and glucagon on calmodulin
mRNA synthesis
One of the effects of insulin is to significantly enhance rates
of calmodulin gene transcription that are mediated by Sp1 (4,
5). Consistent with previous observations, data shown in
Table 1 demonstrate that insulin stimulates CaM gene expression (⫹70% for insulin). However, treatment with glucagon did not significantly effect the steady-state levels of
CaM mRNA (Table 1). In contrast to this, glucagon was
effective in neutralizing the insulin-mediated enhanced transcription of the CaM gene (Table 1).
Parallel aliquots of 500 ␮g of the total cellular protein from
each treatment were immunoprecipitated with Sp1 antibody,
immunoblotted, and probed with anti O-GlcNAc antibody to
detect glycosylated Sp1 (8B). Insulin and glucagon both stimulated the steady-state levels of Sp1 protein. However, after
insulin exposure, the glycosylated Sp1 was markedly increased, whereas glucagon-treated cells exhibited almost no
difference in glycosylated Sp1 from untreated controls. The
overall effect of insulin and glucagon on the levels of Sp1
were probed in the blots after stripping the GlcNAc antibodies and were similar to what was observed in the whole
cell extracts (data not shown).
Effects of insulin and glucagon on glycosylation of Sp1
Both insulin and glucagon stimulated the steady-state levels of total Sp1 in H-411E cells (Fig. 8A). Because additional
levels of regulation may be exerted by preferentially glycosylating Sp1, we tested if the levels of O-GlcNAc-modified
Sp1 were altered by either insulin and/or glucagon. Data
shown in Fig. 8 illustrate Western blots of cell extracts after
exposure to insulin or glucagon for 4 h. An aliquot of extracts
was immunoblotted and probed with antibodies directed
against Sp1 (8A). To assess the equivalence of protein loading, membranes were probed with antibodies to ␤-actin.
TABLE 1. Effects of insulin and glucagon on steady-state levels
of CaM mRNA
Treatments
Exp 1
Exp 2
Exp 3
Control
Insulin
Glucagon
Insulin ⫹ Glucagon
100
160
93
107
100
175
105
119
100
163
72
Densitometric quantitation of autoradiographs of Northern blots
for CaM mRNA in H-411 E cells exposed to insulin and glucagon. The
data from three independent experiments are shown. Data are normalized to control (100%). Please see Materials and Methods in text
for further details.
FIG. 8. The effect of insulin (Ins, 10,000 ␮U/ml) and glucagon (Glu,
1.5 ⫻ 10⫺5 M) on total and glycosylated Sp1 transcription factor
extracted from H-411E cells. Time of incubation with the hormones
was 4 h. Equal amounts of cell extracts were immunoblotted and
probed with antibodies to Sp1 and then stripped and reprobed for
␤-actin to assess equal protein loading (A). The effect of Ins or Glu on
glycosylated Sp1 was examined from cell extracts (500 ␮g of total
cellular protein/sample) after immunoprecipitation with polyclonal
Sp1 antibodies and probing with anti-O-linked GlcNAc antibody (B).
The membrane was stripped and reprobed with anti-Sp1 antibody to
assess total Sp 1 (data not shown). Representative Western blots from
one of four replicate experiments are shown.
Discussion
Sp1, the best characterized member of the Sp transcription
factor family (14), is critical for both basal and insulin-stimulated CaM gene transcription (6, 7). Present studies demonstrate a dose and time dependency for insulin stimulation
of Sp1. Because of a close but usually antagonistic relationship between insulin and glucagon, we investigated the effect of the latter hormone on Sp1. To our surprise, similar to
insulin, glucagon alone stimulated accumulation of Sp1.
However, when added together, glucagon totally inhibited
insulin’s stimulatory effect on Sp1. To some extent this may
reflect, what we see clinically; insulin lowers blood glucose,
glucagon raises it and insulin deficiency or glucagon excess,
lead to diabetes mellitus, i.e. insulin and glucagon act antagonistically toward each other in physiological settings (9,
10). This anabolic/catabolic switch may have common effector pathways. Tang et al. (15) propose a mechanism of
repressive activity of Sp1 in the C/EBP␣ promoter that is
activated during adipocyte differentiation. Ahlgren et al. (16)
show evidence of positive impact of Sp1 transcription factor
on activity of cAMP-dependent bovine CYPIIA gene. Several
genes have Sp1 binding sites within the broadly defined
cis-acting module encompassing cAMP responsive elements
of their promoters. Thus, modulation of metabolic activities
and gene expression by both insulin and glucagon may be
mediated by common transcription factors that bind to
cAMP-responsive and GC-rich elements.
To examine the mechanism of antagonism between these
two hormones, we chose specific inhibitors known to affect
pathways important to either insulin or glucagon action.
Wortmannin, an inhibitor of PI3K and an important second
messenger for many of insulin’s actions (17, 18), markedly
reduced insulin-stimulated Sp1 synthesis but had no effect
on glucagon stimulation of Sp1. This is consistent with PI3K
activity being a key factor in the signaling pathway for insulin (3). Several other insulin-stimulated activities are
blocked by wortmannin, i.e. glucose transport (17, 19) and
down-regulation of gene expression of CYP2E1 (20). However, modulation of these effects by interaction between PI3K
signaling and/or Sp1 has not yet been critically examined.
We demonstrate an essential role of PI3K in induction of
transcription via Sp1. Another important antagonist of insulin is the cytokine TNF-␣ (13). TNF-␣ is thought to work
by phosphorylating serine on the insulin receptor substrate
1 (IRS-1) protein, thereby preventing insulin from stimulating its receptor to phosphorylate tyrosine on IRS-1 and ini-
Keembiyehetty et al. • Glucagon Regulates Sp1
tiating its own signal (13, 13a, 21). TNF-␣ clearly abolished
insulin-stimulated Sp1 synthesis in H-411E cells. TNF-␣ also
inhibits Sp1 synthesis (8) and CaM gene transcription (21).
However, TNF-␣ did not inhibit Sp1 synthesis stimulated by
glucagon on the Western blot. This was predictable because
glucagon is thought to initiate most or all of its effects
through cAMP, so inhibition of IRS-1 that is not tied to the
cAMP pathway should not affect glucagon. The opposing
effects of insulin and glucagon may reflect opposing actions
of these hormones on apoptosis, with insulin being antiapoptotic (lowering cAMP) and glucagon and TNF-␣ (raising cAMP) promoting apoptosis. However, we have not
directly investigated the potential action of these hormones
on H-411E cell survival and its relationship to Sp1 biogenesis
and accumulation.
Could we define the pathway that mediated glucagon’s
stimulation of Sp1 and ultimately transcription of multiple
genes? We initially tried some inhibitors of cAMP synthesis
but results were equivocal at best. However, PKI, a powerful
inhibitor of cAMP, unequivocally inhibited glucagon-stimulation of Sp1 in a dose-dependent manner. It was reported
earlier that PKI could abolish glucagon’s stimulatory effect
on transcription of CYP2E1 gene (20).
We also performed immunocytochemical staining of the
H-411E cells using anti-Sp1 antibody, after treatment with
insulin, glucagon, or both together. Previous data from our
laboratory have shown Sp1 staining in the nucleus, clearly
intensified by insulin, and inhibited by TNF-␣ (8). As judged
by immunocytochemical staining, both insulin and glucagon-stimulated Sp1 accumulation in the H-411E cell nuclei.
However, in contrast to insulin 10 –15% of the glucagonexposed cells have demonstrable Sp1 antigens in the cytoplasm as well as the nucleus. Because the Sp1 protein is
synthesized in the endoplasmic reticulum, it would seem
that in a subset of cells glucagon prevents or delays its migration to the nucleus. Perhaps this could reflect the extent
of either phosphorylation or glycosylation of Sp1 (22–26, 28,
29), influencing subcellular localization. Alternatively, it
could also be due to differences in the rate of accumulation
of Sp1 in subcellular compartments at different stages of the
cell cycle. TNF-␣ clearly inhibited Sp1 staining induced by
insulin (Fig. 7), but this diminution in staining for Sp1 was
also seen to a lesser extent with glucagon and TNF-␣ (Fig. 7).
The former effect (insulin) was anticipated; the latter (glucagon) was not, as it does not match with our Western blot
(Fig. 4) or our presumptive understanding of mechanisms,
involved. TNF-␣ inhibition of glucagon stimulated Sp1 seen
here could be secondary to its previously reported ability to
inhibit cAMP (27). Additional immunocytochemical studies
seen in Fig. 7 are in concordance with the Western blot data
with a few exceptions. The differences in the time of exposure
of cells to TNF-␣ for immunostaining (12–16 h) and Western
blotting (4 h), may partially be responsible for the apparent
discrepancy between the two data. Because O-glycosylation
of Sp1, which occurs with insulin but not glucagon, likely
alters its susceptibility to degradation, the initial increase in
glucagon-stimulated Sp1 at 4 h seen by Western blotting may
be significantly weakened due to degradation during the 12to 16-h incubation for immunostaining. This would be even
more likely, if TNF-␣ itself inhibits glycosylation or other-
Endocrinology, April 2002, 143(4):1512–1520 1519
wise promotes Sp1 degradation directly. To examine further
these and other differences between insulin and glucagon
stimulation of Sp1 and possible compartmentalization, transport or migration differences between cytoplasmic structures, nucleus, etc., we studied the glycosylation state of the
Sp1. Data presented in Fig. 8 provide strong preliminary
support for a role for O-glycosylation of Sp1 in translating the
effects of insulin and glucagon. In cells treated with insulin
alone, Sp1 is most probably hyperglycosylated and translocates to the nucleus and facilitates transcription more efficiently (25, 28, 29). In cells treated with glucagon alone, Sp1
is probably in a hypoglycosylated form, which can mean: 1)
the hypoglycosylation marks the Sp1 for proteosomic degradation; 2) the hypoglycosylated Sp1 may not yet be properly folded into its final quaternary structure (25); and 3) the
hypoglycosylated Sp1 is not functioning and gives an unclear
defective translocation signal. Because insulin stimulates
both glycosylation and nuclear staining of Sp1 and enhances
transcription of a Sp1-driven CaM promoter, it is more potent
than glucagon, which does not glycosylate the Sp1, and,
hence, leaves it less active in the cytoplasm unable to migrate
to the nucleus.
Sp1 is able to regulate the synthesis of many new proteins.
To extend the above observations to the next level, we performed Northern blots of H-411E cells exposed to insulin or
glucagon. Consistent with our previously published data (4,
5), our data in Table 1 again demonstrate that insulin stimulates CaM gene expression, but, as can be seen, glucagon
does not. With CaM gene transcription clearly linked to
insulin’s stimulation of Sp1, why did another hormone, i.e.
glucagon which also stimulates Sp1, not affect CaM gene
transcription? As outlined above, these differences could be
due to differences in signal transduction, subcellular compartmentalization, phosphorylation, or more likely glycosylation. Future investigations will be needed to explain the
cause of these differences, i.e. transcription vs. translation.
However, it is not far-fetched to suppose that the insulin/
glucagon ratio, which is a critical factor in homeostasis and
maintenance of blood glucose and multiple metabolic parameters (ketosis, lipolysis, etc.), also regulates selective gene
transcription, orchestrating the net response to reflect the
insulin/glucagon ratio. In this way, along with selective
compartmentalization, glycosylation and/or phosphorylation and use of multiple signal transduction pathways, the
appropriate cellular response to physiologic or pathologic
challenges can be orchestrated.
Acknowledgments
We thank Ms. Marjorie Palazzolo for Northern blots, Ms. Phyllis Love
in the VAMC for immunocytochemical staining of the cells, Mr. Paul
Markowitz for general technical support, and Ms. Ann Oswalt for secretarial support.
Received October 12, 2001. Accepted December 20, 2001.
Address all correspondence and requests for reprints to: Solomon
S. Solomon, M.D., Veterans Affairs Medical Center, Research Service
(151), 1030 Jefferson Avenue, Memphis, Tennessee 38104. E-mail:
[email protected].
This work was supported by Merit Review grants from the Department of Veterans Affairs (to S.S.S. and R.R.) and the NIH (to S.S.S. and
R.R.). R.P.C. is supported by a Medical Student’s Short Term Research
1520
Endocrinology, April 2002, 143(4):1512–1520
Training Grant (T35-DK-07405-16, NIH). R.R. is a Senior Research Career
Scientist of the Department of Veterans’ Affairs.
References
1. Solomon S, Raghow R, Molecular mechanisms of insulin modulated signaling
and regulation of gene transcription. Recent research developments in endocrinology, Transworld Research Network, Trivandrum, India, in press
1a.Solomon SS 1975 Effects of insulin and lipolytic hormones on cyclic AMP
phosphodiesterase activity in normal and diabetic rat adipose tissue. Endocrinology 96:1366 –1372
2. Solomon SS, Steiner MS, Little WL, Rao RH, Sanders LL, Palazzolo MR 1987
Inhibitor of calmodulin and cAMP phosphodiesterase activity in BB rats.
Diabetes 36:210 –215
3. Solomon SS, Palazzolo MR, Green S, Raghow R 1990 Expression of calmodulin genes is down regulated in diabetic BB rats. Biochem Biophys Res Commun 168:1007–1012
4. Solomon SS, Palazzolo MR, Smoake JA, Raghow R 1995 Insulin stimulated
calmodulin gene expression in rat H-411E cells can be selectively blocked by
antisense oligonucleotides. Biochem Biophys Res Commun 210:921–930
5. Solomon SS, Palazzolo MR, Takahashi T, Raghow R 1997 Transcription
factor Sp1 is necessary for basal calmodulin gene transcription and for its
selective stimulation by insulin. Endocrinology 138:5052–5054
6. Solomon SS, Palazzolo MR, Takahashi T, Raghow R 1997 Insulin stimulates
rat calmodulin gene transcription through activation of Sp1. Proc Assoc Amer
Phys 109:470 – 477
7. Pan X, Solomon SS, Shah R, Palazzolo MR, Raghow R 2000 Members of the
Sp transcription factor family regulate rat calmodulin gene transcription. J Lab
Clin Med 136:157–163
8. Pan X, Solomon SS, Borromeo DM, Martinez-Hernandez A, Raghow R 2001
Insulin deprivation leads to deficiency of Sp1 transcription factor in H-411E
hepatoma cells and in streptozotocin-induced diabetic ketoacidosis in the rat.
Endocrinology 142:1635–1641
9. Cherrington AD, Chiasson JL, Liljenquist JE, Jennings AS, Keller U, Lacy
WW 1976 The role of insulin and glucagon in the regulation of basal glucose
production in the post-absorptive dog. J Clin Invest 58:1407–1418
10. Nielsen M, Wise S, Dinneen S, Schwenk WF, Basu A, Rizza 1997 Assessment
of hepatic sensitivity to glucagon in NIDDM: use as a tool to estimate the
contribution of the indirect pathway to nocturnal glycogen synthesis. Diabetes
46:2007–2016
11. Martinez-Hernandez A 1987 Methods for electron immunohistochemistry. In:
Cunningham LW, ed. Methods in enzymology. Structural and contractile
proteins: extracellular matrix, vol 145. New York: Academic Press; 103–133
12. Amenta P, Martinez-Hernandez A 1995 Immunological approaches to the
identification of extracellular matrix components in tissues and cell cultures.
In: Haralson M, Hassel J, eds. The extracellular matrix: a practical approach.
Oxford, UK: IRL Press; 303–325
Keembiyehetty et al. • Glucagon Regulates Sp1
13. Duckworth WC 1997 Tumor necrosis factor and insulin resistance: specificity
of sequence accounts of inhibition of insulin action. J Lab Clin Med 130:114 –115
13a.Solomon SS, Mishra SK, Palazzolo, MR, Postlethwaite AE, Seyer JM 1997
Identification of specific sites in the TNF-␣ molecule promoting insulin resistance in H-411E cells. J Lab Clin Med 130:139 –145
14. Fry CJ, Farnham PJ 1999 Context-dependent transcriptional regulation. J Biol
Chem 274:29583–29586
15. Tang Q, Jiang M, Lane MD 1999 Repressive effect of Sp1 on the C/EBP␣ gene
promoter: Role in adipocyte differentiation. Mol Cell Biol 19:4855– 4865
16. Ahlgren R, Suske G, Waterman MR, Lund J 1999 Role of Sp1 in cAMPdependent transcriptional regulation of the Bovine CYPIIA gene. J Biol Chem
274:19422–19428
17. Combettes-Souverain M, Issad T 1998 Molecular basis of insulin action. Diabetes Metab 24:477– 489
18. Davies SP, Reddy H, Caivano M, Cohen P 2000 Specificity and mechanism
of action of some commonly used protein kinase inhibitors. Biochem J 351:
95–105
19. Virkamaki A, Ueki K, Kahn CR 1999 Protein-protein interaction in insulin
signaling and molecular mechanisms of insulin resistance. J Clin Inves 103:
931–943
20. Woodcroft KJ, Novak RF 1999 The role of phosphatidylinositol 3-kinase, Src
kinase, and protein kinase A signaling pathway in insulin and glucagon
regulation of CYP2E1 expression. Biochem Biophys Res Commun 266:304 –307
21. Solomon SS, Mishra SK, Cwik C, Rajanna B, Postlethwaite AE 1997 Pioglitazone and Metformin reverse insulin resistance induced by tumor necrosis
factor-␣ in liver cells. Horm Metab Res 29:379 –382
22. Hunter T, Karin M 1992 The regulation of transcription by phosphorylation.
Cell 70:375–387
23. Meek D, Street A 1992 Nuclear protein phosphorylation and growth control.
Biochem J 287:1–15
24. Jackson S, MacDonald J, Lees-Miller S, Tjian R 1990 GC box binding induces
phosphorylation of Sp1 by a DNA-dependent protein kinase. Cell 63:155–165
25. Jackson S, Tjian R 1988 O-glycosylation of eukaryotic transcription factors:
implication for mechanisms of transcriptional regulation. Cell 55:125–133
26. Han I, Roos M, Kudlow J 1998 Interaction of transcription factor Sp1 with the
nuclear pore protein p62 requires the C-terminal domain of p62. J Cell Biochem
68:50 – 61
27. Christ B and Nath A 1996 Impairment by interleukin 1 beta and tumor necrosis
factor ␣ of the glucagon-induced increase in phosphoenolpyruvate carboxykinase gene expression and gluconeogenesis in cultured rat hepatocytes. Biochem J 320:161–166
28. Han I, Kudlow J 1997 Reduced O-glycosylation of Sp1 is associated with
increased proteosome susceptibility. Mol Cell Biol 17:2550 –2558
29. Du XL, Edelstein D, Rossetti L, Fantus IG, Goldberg H, Ziyadeh F, Wu J,
Brownlee M 2000 Hyperglycemia-induced mitochondrial superoxide overproduction activates the hexosamine pathway and induces plasminogen activator inhibitor-1 expression by increasing Sp1 glycosylation. Proc Natl Acad
Sci USA 97:12222–12226