| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
National Research Council (L.S., L.R., E.P.), Institute of Biomedical Technology, Rome; Department of Experimental Medicine (M.P.F.), University of L’Aquila, L’Aquila; Department of Experimental Medicine and Pathology (M.R.C., M.A.R., L.F., A.G.), University "La Sapienza", Rome; and Neuromed Institute (L.F., A.G.), Pozzilli, Italy
Address all correspondence and requests for reprints to: Dr. Elisa Petrangeli, Istituto di Tecnologie Biomediche, Consiglio Nazionale delle Ricerche, via G. B. Morgagni, 30/E, 00161 Roma, Italy. E-mail: petrang{at}itbm.rm.cnr.it
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(ER). This action was retained by
the 36-bp core promoter fragment and did not require the receptor DNA
binding domain, as demonstrated by analyzing the role of ER deletion
mutants on EGFR gene promoter-derived constructs. The 36-bp promoter
fragment does not contain an estrogen response element but an
imperfect thyroid hormone response element half-site that overlaps the
Sp1 binding site. ER does not bind this imperfect thyroid hormone
response element half-site but is able to enhance binding of Sp1 to its
site, in gel mobility shift assays, suggesting that the mechanism by
which the receptor stimulated the transcription involved
protein-protein interactions that replaced DNA binding. To explain this
action, we propose a model in which induction of the EGFR gene
expression by estrogens in HeLa cells is dependent upon the formation
of a transcriptionally active ER-Sp1 complex that binds to the
GC-rich (Sp1) region of the minimal promoter. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In human breast cancer, the ER expression is inversely correlated with the epidermal growth factor (EGF) receptor (EGFR) expression: ER-positive breast cancer cells generally express very low levels of EGFR, whereas ER-negative cells express high levels of EGFR (8). EGFR overexpression predicts a poor survival prognosis (9) and failure of endocrine therapy (10). It has also been shown that EGF can partially replace estrogen in the growth promotion on MCF-7 breast cancer cells implanted in ovariectomized nude mice (11). Furthermore, in the rat uterus and in human breast cancer cells, 17ß-estradiol (E2) transiently up-regulates EGFR messenger RNA (mRNA) and protein levels (12, 13), acting with a direct mechanism because new protein synthesis is not required (13).
The promoter of the EGFR gene is GC-rich, contains multiple transcriptional start sites, and lacks both CAAT and TATA boxes (14). Full activity of a 1081-bp region (-1100 to -19 bp) of the EGFR gene was reduced 5-fold in a proximal 134-bp 5' fragment (-153 to -19 bp) that contains promoter/enhancer elements responsible for basal activity and inductive responses to EGF, the tumor promoter TPA, dexamethasone, and (Bu)2cAMP, indicating positive control elements within the removed region (15). A 36-bp (-112 to -77 bp) promoter element, which functions both as a minimal promoter with internal start sites and as a strong enhancer element (16), was identified within the 134-bp region. Both the promoter and the enhancer activities map to a GC cassette that is a binding site for the transcription factor Sp1 (17, 18).
The EGFR promoter sequence does not display a canonical estrogen response element (ERE) (4) but three imperfect palindromes located upstream with respect to the 134-bp region (13). An imperfect thyroid hormone response element (TRE) half-site, located in the 36-bp promoter element, was also identified. T3R and RAR are able to bind this half-site (19, 20), which presents sequence homology to the upstream half of a TRE in the rat GH gene promoter (21).
In this study, we investigated the effects of E2
on the modulation of EGFR gene transcription in the human ER-negative
HeLa cells, transfected with the wild-type or the deletion mutant ER
expression vectors, and EGFR gene promoter-derived constructs linked to
a luciferase reporter gene. Data show that the EGFR gene is
estrogen-responsive and that this effect is exerted through both the
ligand-bound wild-type ER and its DNA binding domain (DBD) deletion
mutant, by means of a DNA-binding-independent action of ER on the
core region of the promoter, which involves protein-protein interaction
with the transcription factor Sp1.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The 1081-bp (-1100 to -19 bp, oriented to the translation start
site as position +1) 5' region of the EGFR gene, and its 134-bp
fragment (-153 to -19 bp), ligated into the HindIII site
of the luciferase expression vector pSVOAL
5' (15, 22), were provided
by Gordon Gill (Department of Medicine, University of California-San
Diego). The wild-type human ER (hER) expression vector (HEGO)
(23), and the receptor deletion mutant HE11 (24), HE15, and HE19 (25)
expression vectors, were provided by Pierre Chambon (CNRS, Strasbourg,
France). Recombinant human Sp1 and ER proteins were purchased from
Promega Corp. and PanVera (Madison, WI), respectively.
Human anti-ER H222 antibody was provided by Geoffrey L. Greene (Ben
May Institute, University of Chicago, Chicago, IL). Anti-ER C-314
monoclonal antibody and anti-Sp1 polyclonal antibody were purchased
from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA).
Ethanol was used as solvent for E2. All other
chemicals and biochemicals were of the highest quality available from
commercial sources. Oligonucleotides were synthesized by
Biogen (Rome, Italy), and their structures are
listed below. The consensus ERE and the imperfect TRE half-site, on the
X. laevis vitellogenin B1 and the 36-bp EGFR promoters,
respectively, are underlined; the Sp1 binding sites, on the
consensus Sp1 and the EGFR core promoter oligonucleotides, are shown in
bold script, and every mutated base of the Sp1
site on the EGFR minimal promoter (36-bp EGFR-mut) is indicated with an
asterisk: vitellogenin-ERE: 5'-GTCCAAAGTCAGGTCACAGTGACCTGATCAAAGTT-3' 36-bp
EGFR
promoter: 5'-AGCTTCGCGTCCGCCCGAGTCCCCGCCTCGCCGCCAACGCCA-3' 36-bp
EGFR-mut
promoter: 5'-AGCTTCGCGTCCGCCCGAGTCTTTGTCTCGCCGCCAACGCCA-3' *** 1 consensus
Sp1: 5'-AGCTTATTCGATCGGGGCGGGGCGAGCG-3'
Cloning
The wild-type and mutant 36-bp (-112 to -77 bp) regions of the
EGFR gene promoter were cloned into the pSVOAL5' vector at the
HindIII site to give the pSVOAL5'-36 and pSVOAL5'-36m
constructs, respectively. Ligation products were transformed into HB101
cells, and clones were verified by sequencing.
Transient transfection and luciferase activity assays
Cultured HeLa cells were maintained in phenol-red free DMEM
dextran-charcoal-stripped FBS (DCCFBS) 4 days before transfection.
Eight micrograms of EGFR gene promoter-derived constructs, 5 µg
wild-type or variant ER expression plasmids, and 6 µg
ß-galactosidase-lacZ plasmid (Amersham Pharmacia Biotech, Uppsala, Sweden) were cotransfected using the calcium
phosphate method. pSG5 was used as an empty vector (control) and to
maintain uniform levels of added DNA. Background was determined by
transfection of the promoterless pSVOAL5' luciferase vector.
Trypsinized cells, suspended in 5% DCCFBS medium, containing the
calcium phosphate DNA precipitate, were plated in a 10-cm tissue
culture dish. After 2 h, cultures were treated with ethanol
(0.01% total vol) or E2. Twenty hours after
transfection, the medium was changed to DMEM containing 2.5% DCCFBS,
and hormone treatment was continued for an additional 24 h (19).
Cells were then scraped from the plates, and protein extracts were
prepared (26). Forty microliters of cell extract were added to a
luciferase assay reaction containing 25 mM
glycylglycine (pH 7.8), 2 mM ATP, 10
mM MgSO4, in a vol of 350
µl. The reaction was initiated by the injection of 100 µl of 0.2
mM luciferin, and light readings were integrated,
over 10 sec, on a 1251 luminometer (LKB Wallac, Turku,
Finland) (22). ß-galactosidase activity was assayed: 0.8 ml of 60
mM
Na2HPO4, 40
mM
NaH2PO4, 10
mM KCl, 1 mM
MgSO4, 38 mM
ß-mercaptoethanol, and 0.2 ml of
O-nitrophenyl-ß-D-galactopyranoside (4 mg/ml in 0.1
M sodium phosphate, pH 7.5) were added to 0.2 ml
cell extracts, and OD was measured at 420 nm (27). Data were normalized
as light units of luciferase activity per unit of ß-galactosidase
activity, to correct for differences in transfection efficiency between
samples. Results are presented as the mean ±
SE for at least four separate experiments.
Electrophoretic mobility shift assays
Oligonucleotides were annealed and labeled at the 5' end using
T4-polynucleotide kinase and [
-32P] ATP.
With the fragment corresponding to consensus ERE of the vitellogenin
(vit-ERE), 10 µg nuclear extracts from HeLa cells (28) and 0.6 pmol
ER were preincubated in 1 x binding buffer (10% glycerol, 20
mM HEPES, pH 7.9, 2 mM
MgCl2, 1 mM dithiothreitol, 0.1 mg/ml
BSA) with 6 µg poly (dI-dC) for 15 min at 4 C. When the
oligonucleotide containing the EGFR promoter sequence between positions
-112 to -77 was used, 4–16 µg nuclear extracts, 0.06–30 pmol
ER, 1–9 µg poly (dA-dT), or 0.05–3.5 µg poly (dI-dC) were
preincubated for 15 min at 4 C. The mixture was then incubated for 30
min at room temperature, in the presence of 100 nM
E2, with 0.6 ng (20,000 cpm)
32P-vit-ERE or 32P-EGFR
core promoter oligonucleotides. In supershift experiments, up to 3 µg
anti-ER H222 antibody (directed against the C-terminal region of the
protein) were added during the preincubation. For competition
experiments, 100-fold molar excess cold oligonucleotides, containing
the vit-ERE or the EGFR core promoter sequence, were added to the
binding mixture, 20 min before addition of the
32P-labeled probes (29). For ER-enhanced Sp1
binding studies, 0.4–0.8 pmol ER were incubated for 15 min at 4 C
in 1 x binding buffer (6% glycerol, 1 mM
MgCl2, 0.5 mM EDTA, 0.5
mM dithiothreitol, 50 mM NaCl, 10
mM Tris-HCl, pH 8), 0.1 mg/ml BSA, and 100 nM
E2; 3 ng Sp1 protein were added to the mixture
and incubated for 10 min at 4 C; 0.6 ng (20,000 cpm)
32P-consensus Sp1 or
32P-EGFR core promoter oligonucleotides were
added to the reaction mixture, in the presence of 1 µg poly (dI-dC)
and incubated for an additional 30 min at room temperature (30). For
supershift experiments, 6 µg anti-Sp1, or anti-ER H222 and C-314
(raised against the amino terminal domain of the protein) antibodies,
were added during the preincubation. For competition experiments,
100-fold molar excess cold oligonucleotides, containing the consensus
Sp1 or the wild-type or mutant EGFR core promoter sequence, were added
to the binding mixture, 20 min before addition of the
32P-labeled probes. Samples were loaded onto a
4% polyacrylamide gel (acrylamide-bisacrylamide ratio, 37.5:1) and run
for 2 h at 160 V in 0.25 x TBE (0.09 M Tris,
0.09 M boric acid, and 2 mM EDTA, pH 8). Gels
were then dried, and protein-DNA binding was visualized by
autoradiography.
Statistical analysis
Unless otherwise stated, statistical significance was determined
by one-way ANOVA with Dunnett’s test. The level of statistical
significance was set at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
on reporter gene
activation in response to E2, the 1081-bp EGFR
promoter-luciferase construct and increasing amounts of ER
expression vector (HEGO) were cotransfected, with or without the
addition of 10 nM and 100 nM
E2, in HeLa cells (data not shown). In the
presence of 5 µg ER expression vector, the basal promoter activity
showed no change. After E2 treatment, the EGFR
promoter conferred a significant estrogen-responsiveness to the
downstream luciferase-reporter gene at the higher dose of hormone (1.1-
and 1.5-fold, expressed as ratio of activity in the presence of 10
nM and 100 nM E2,
respectively, divided by activity in the absence of hormone; Fig. 1A
). Estrogen treatment
was found to be ineffective on the EGFR promoter-luciferase construct
cotransfected with the empty pSG5 vector. Furthermore, no activity was
observed after cotransfection of the promoterless vector pSVOAL5'
and HEGO, with or without addition of E2 (data
not shown). Transfections were then performed with the 134-bp EGFR
promoter-luciferase construct, which retains 15% of basal activity
(data not shown). The maximal ligand-dependent reporter gene activation
was observed at the dose of 5 µg ER expression vector and was more
evident than with the full-length promoter (1.8- and 2.2-fold, for 10
nM and 100 nM E2,
respectively; Fig. 1B). Also the (-153 to -19 bp) EGFR
promoter-luciferase construct, cotransfected with the empty pSG5
vector, failed to direct the transcription, after hormone treatment. A
further (-112 to -77 bp) deletion promoter element, which retains 2%
of full-length promoter basal activity (data not shown), was
investigated. This construct strongly enhanced reporter gene
transcription induced by ligand-activated ER (3.3- and 4.7-fold, for
10 nM and 100 nM E2,
respectively; Fig. 1C). The 36-bp region, therefore, retains the
estrogen responsiveness of the EGFR gene.
|
and EGFR gene
minimal promoter could bind the same
sequence on the EGFR gene promoter. ER binding to the vit-ERE and to
the 36-bp EGFR gene promoters was determined in gel mobility shift
assays. On the 32P-vit-ERE oligonucleotide, the
ER protein caused a retarded band formation, in the presence of HeLa
cell nuclear extracts (Fig. 2A, lane 3).
The intensity of the bound ER-32P-vit-ERE band
was not significantly enhanced after incubation with estrogens (lane
4), it was decreased by competition with a 100-fold molar excess of
unlabeled vit-ERE oligonucleotide (lane 6), and it was not affected by
a 100-fold molar excess of unlabeled EGFR promoter oligonucleotide
(lane 7). Addition of the anti-ER H222 antibody to the binding
reaction produced a supershifted band (lane 5). When the
32P-36-bp-EGFR promoter oligonucleotide was used
(Fig. 2B), the addition of ER protein, in the presence of nuclear
extracts, did not lead to the formation of a specific retarded band, at
any of the doses tested, in the presence of either
E2 or H222 antibody (lanes 3–5). On the other
hand, the addition of 30 pmol protein led to an increase in the
intensity of retarded bands observed in the presence of nuclear
extracts alone. These data undoubtedly highlight the specificity of
ER-vit-ERE interaction and suggest the existence of an alternative
mechanism for estrogen-induced activation on the EGFR gene promoter,
because ER is unable to bind the imperfect TRE half-site.
|
-induced EGFR gene
transcription involved in the
transcriptional regulation, the 1081-bp, the 134-bp, and the 36-bp EGFR
promoter-luciferase constructs were cotransfected with the three hER
deletion mutant expression vectors HE19, HE11, HE15 (Fig. 3A), that lack,
respectively, the N-terminal region of the protein (1–178), the DNA
binding domain (185–251), and the ligand binding domain
(282–595). The N- and C-truncated proteins did not enhance reporter
gene transcription in the presence of their ligand, on any of the
promoter fragments, whereas the HE11 receptor induced a significant and
strong estrogen-dependent increase of transcription (on the 1081-bp
promoter fragment, 2.5- to 5.5-fold, expressed as ratio of activity in
the presence of 10 nM and 100 nM
E2, respectively, divided by activity in the
absence of hormone, Fig. 3B; on the 134-bp promoter segment, 2.8- to
5.7-fold, Fig. 3C; on the 36-bp promoter element, 3.4- to 4.6-fold,
Fig. 3D). Analysis of the ER functional activity on the expression
of the EGFR promoter-luciferase constructs confirms that the
protein acts through a DNA-binding-independent mechanism and that this
effect is retained by the smaller promoter fragment.
|
-enhancement of Sp1 binding to the EGFR minimal
promoter with other DNA-bound transcription factors, such
as Sp1 (30, 31, 32, 33, 34, 35, 36). The observation that the 36-bp fragment of the EGFR
gene promoter contains a GC-rich site at position -110 to -84 (18),
which overlaps the imperfect TRE half-site localized at -102 to -98
bp (17), prompted us to investigate whether Sp1 could be a target for
ER. The results in Fig. 4 show the binding of
Sp1 protein to both the 32P-consensus Sp1 and the
32P-EGFR core promoter sequences. Formation of a
retarded band was observed using both oligonucleotides. The intensity
of the shifted Sp1 band (Fig. 4A, lane 2) was enhanced after incubation
with 0.4 or 0.8 pmol ER protein (lanes 3 and 4) and was supershifted
after addition of anti-Sp1 antibody (lane 5) but was not affected after
addition of anti-ER C-314 or H222 antibodies (lanes 6 and 7). Also
the GC-rich site within the EGFR core promoter bound Sp1 protein (Fig. 4B, lane 3), even if with lower affinity than the consensus Sp1
sequence but not ER protein (lane 2). The intensity also of the
bound Sp1-32P-EGFR band was enhanced after
addition of 0.4–0.8 pmol ER protein (lanes 4 and 5), it was
decreased by competition with a 100-fold molar excess of unlabeled EGFR
core promoter or consensus Sp1 oligonucleotides (lanes 6 and 7), but it
was not affected by a 100-fold molar excess of unlabeled EGFR-mut
oligonucleotide (lane 8), which presents 4 base substitutions in the
GC-box (for the sequence, see Materials and Methods).
Addition of the anti-Sp1 antibody to the binding reaction produced a
supershifted band (lane 9), whereas the addition of both anti-ER
H222 and C-314 antibodies was ineffective (data not shown). The
intensity of the Sp1-32P-EGFR-retarded bands was
not affected by the presence or absence of 100 nM
E2 (data not shown). These results prove the
existence of a functional synergy between ER and Sp1 nuclear
transcription factors for binding to the Sp1 consensus sequence.
|
5'-36m
construct (containing the 36-bp EGFR-mut promoter fragment) was
cotransfected with pSG5 control vector, and the ER or HE11
expression vectors, with or without hormone treatment. Basal activity
of the mutated construct was reduced to 44% of the wild-type 36-bp
EGFR promoter-luciferase construct (0.9% of full-length promoter basal
activity, data not shown). Both receptors induced an estrogen-dependent
reporter gene transcription, directed by the mutant promoter fragment,
being significantly lower, with respect to the transcription induced
from the pSVOAL5'-36 construct (Table 1). Therefore, the disruption of the Sp1
binding site on the 36-bp EGFR promoter fragment causes inhibition of
the estrogen-induced reporter gene transcription, ranging from
39–41%, thus confirming that a considerable part of the ER-induced
EGFR promoter transcriptional activity involves the interaction between
Sp1 protein and its binding site.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Computer analysis showed that the EGFR promoter sequence does not
contain a canonical ERE, the consensus sequence of which is
GGTCAnnnTGACC (4). Nevertheless, three imperfect ERE palindromes, with
one mismatch in each half-site, were identified at positions -778 to
-767, -753 to -739, and -404 to -392, and ER ability to bind
the two more distal elements, even if with lower affinity than to a
consensus ERE, was shown (13). The results depicted in Fig. 1
, A and B,
highlight the 1081-bp EGFR promoter estrogen-responsiveness retained by
the 134-bp fragment, localized downstream with respect to the three
imperfect palindromes. These observations would seem to suggest that
estrogens stimulate EGFR transactivation through a mechanism not
involving ER binding to the putative EREs. Evidence that the 134-bp
promoter segment was still able to promote gene expression prompted us
to investigate the molecular mechanism of this action. In the (-153 to
-19 bp) EGFR promoter fragment, the sequence most resembling a
consensus ERE was the imperfect TRE half-site sequence AGTCC localized
at position -102 to -98. This type of pentameric sequence presents
two point mutations, with respect to a downstream canonical ERE
half-site, with replacement, respectively, of the thymine by adenine,
in the first position, and of the adenine by thymine, in the third
position. Even if ER can bind, as a monomer, to a TRE (37), the
results in Fig. 2B demonstrated that the receptor was unable to bind
this imperfect TRE half-site. This finding was not followed by loss of
estrogen-responsiveness, as demonstrated by transient transfections of
the pSVOAL5'-36 construct, containing the imperfect TRE half-site
(Fig. 1C), suggesting that ER action on the EGFR promoter could be
mediated through a DNA-binding-independent mechanism. This observation
is further confirmed by analysis of the role of ER deletion mutants
HE11, HE15, and HE19 (Fig. 3A) in promoting reporter gene
transcription. The results demonstrated that the EGFR gene retained
estrogen-responsiveness, even in the absence of the ER DBD, whereas
the inability of the N- and the C-terminal mutant receptors, HE19 and
HE15, to enhance reporter gene transcription when ligand-activated
(Fig. 3; B, C, and D) suggests a functional importance for the two
activation domains, AF1 and AF2. The finding that action was
independent of the receptor DBD supports the hypothesis that the
mechanism by which the receptor stimulates transcription involves
protein-protein, rather than protein-DNA, contacts. Several
estrogen-responsive genes in which ER does not act through consensus
EREs have been detected (30, 31, 32, 33, 34, 35, 36, 38, 39). ER/AP-1 interactions have
been reported in which the AP-1 complex serves as a tether, when bound
to its cognate DNA element, to target steroid receptors, such as the
ER, in the absence of a canonical response element (38, 40). Other
authors have shown the existence, in some target gene promoters, of
estrogen-responsive regions containing Sp1 and ERE-half sites with
variability in the ERE half-site sequences. It has been suggested that
gene transactivation may be caused by interaction between ER and Sp1
complexes, which are stabilized by interactions with an
Sp1(N)xERE half-site DNA-binding motif (30, 32, 41). The ER/Sp1 complex activity was not abolished by the DBD deletion
or by mutation of the ERE half-site, suggesting that estrogen
inducibility was mediated through protein-protein interaction and
required GC-rich Sp1 binding sites that were, therefore, potential
targets for ER-mediated transactivation (30, 31, 32, 33, 34, 35, 36).
Both ER and Sp1 interact physically with other nuclear proteins (38, 42, 43); furthermore, evidence is accumulating showing that nuclear
receptors can come directly into contact with some of the basal factors
of the preinitiation complex; in particular, the ER, through its AF2
domain, is able to bind TFIIB (44), and both the AF1 and the AF2
activation domains bind to components of the TFIID complex in
vitro (45, 46). These observations are in agreement with the
hypothesis that Sp1 plays a role in recruiting the transcription
apparatus on TATA-less promoters, through a tethering activity
anchoring the basal initiation complex to the promoter by binding to
the Sp1 site (47). Deoxyribonuclease I footprinting showed that Sp1 can
bind to four CCGCCC sequences (-457 to -440, -365 to -286, -214 to
-200, and -110 to -84) in the EGFR gene promoter and may, therefore,
play a role in its regulation (18). The 36-bp fragment, which retains
the estrogen-responsiveness, contains the most proximal Sp1 motif,
which partially overlaps with the imperfect TRE half-site. These data
suggest the existence of a complex, also on the EGFR gene, in which
ER could be the tether between the basal initiation complex and Sp1.
On the other hand, ER does not seem to activate the gene expression
through an indirect mechanism, because other investigators have shown
that estrogens can increase EGFR mRNA levels, even in the absence of
new protein synthesis (13).
The results depicted in Fig. 2B show a clear enhancement of the
retarded bands, achieved in the presence of nuclear extracts from HeLa
cells and 32P-36-bp EGFR promoter oligonucleotide
containing the mutated TRE half-site and the Sp1 site, after
coincubation with ER. This finding is confirmed by the molecular
mechanism, evaluated by means of the recombinant Sp1 protein and the
addition of antibodies, as shown in Fig. 4. The ER enhances the
Sp1-DNA on rate, in a concentration-dependent manner, demonstrating the
existence of a functional synergism between ER and Sp1 proteins on
both the consensus Sp1 and the 36-bp EGFR core promoter
oligonucleotides, as also reported in other studies (30, 31, 32, 33, 34, 35, 36, 42). The
synergistic action does not involve the formation of a ternary
ER-Sp1-DNA complex, as confirmed by the loss of supershifted bands
in the presence of both proteins and after addition of either
anti-ER H222 or C-314 antibodies. This finding could be attributed
to the weakness of the ternary complex, which cannot be shown in an
in vitro system, although it possesses functional activity
(30, 31, 32, 33, 34, 35, 36, 42). Of considerable importance, on the other hand, is the
presence of a whole Sp1 binding site, as demonstrated by gel mobility
shift assays, which showed that the (-112 to -77 bp) EGFR
oligonucleotide with mutated Sp1 sequence (EGFR-mut) was ineffective in
competitively binding the Sp1 protein (Fig. 4B, lane 8). The EGFR-mut
oligonucleotide contains mutations that change residues between -98
and -96 and residue -94. These are a part of the 9 G residues that
Sp1 contacts, within the region extending from -110 to -84 bp, in the
36-bp EGFR promoter fragment (17). As previously reported (16, 17, 18),
these 4 mutations suffice to impair the 36-bp EGFR promoter activity.
Cotransfections of either the wild-type or the DBD-deletion mutant
ER expression vectors with the pSVOAL5'-36m reporter gene
construct show a partial, but significant, repression of the
E2-induced transcriptional activation of the
36-bp EGFR promoter (Table 1), highlighting the involvement of the
mutated residues of the Sp1 binding site in the estrogenic
transcriptional regulation of the EGFR gene.
In conclusion, the results of this study show that ligand-activated
ER confers estrogen-responsiveness to the EGFR gene. This activity
was retained by the 36-bp promoter fragment, and, because DNA binding
was not required, it was independent of the receptor DBD but dependent
on the two activation domains, AF1 and AF2. In the attempt to offer an
explanation for the mechanism of this action, we propose a model in
which ER enhances EGFR gene transcription through protein-protein
interactions that replace DNA binding. The ER target is, at least in
part, the transcription factor Sp1, and the receptor might act as a
tether between the basal initiation complex and the transcription
factor, contributing to the stabilization of an Sp1-dependent
initiation complex on a TATA-less template, through enhanced binding of
Sp1 to its site.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received January 3, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ERß at AP1 sites. Science 277:1508–1510This article has been cited by other articles:
 |
|
 |
|
  S. Khan, F. Wu, S. Liu, Q. Wu, and S. Safe Role of specificity protein transcription factors in estrogen-induced gene expression in MCF-7 breast cancer cells J. Mol. Endocrinol., October 1, 2007; 39(4): 289 - 304. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Bionaz and J. J. Loor Identification of reference genes for quantitative real-time PCR in the bovine mammary gland during the lactation cycle Physiol Genomics, May 11, 2007; 29(3): 312 - 319. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Maor, D. Mayer, R. I Yarden, A. V Lee, R. Sarfstein, H. Werner, and M. Z Papa Estrogen receptor regulates insulin-like growth factor-I receptor gene expression in breast tumor cells: involvement of transcription factor Sp1 J. Endocrinol., December 1, 2006; 191(3): 605 - 612. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M L Panno, L Mauro, S Marsico, D Bellizzi, P Rizza, C Morelli, M Salerno, F Giordano, and S Ando' Evidence that the mouse insulin receptor substrate-1 belongs to the gene family on which the promoter is activated by estrogen receptor {alpha} through its interaction with Sp1 J. Mol. Endocrinol., February 1, 2006; 36(1): 91 - 105. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. O'Lone, M. C. Frith, E. K. Karlsson, and U. Hansen Genomic Targets of Nuclear Estrogen Receptors Mol. Endocrinol., August 1, 2004; 18(8): 1859 - 1875. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Khan, M. Abdelrahim, I. Samudio, and S. Safe Estrogen Receptor/Sp1 Complexes Are Required for Induction of cad Gene Expression by 17{beta}-Estradiol in Breast Cancer Cells Endocrinology, June 1, 2003; 144(6): 2325 - 2335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. T. Baron, E. M. Cora, J. M. Lafky, C. H. Boardman, M. C. Buenafe, A. Rademaker, D. Liu, D. A. Fishman, K. C. Podratz, and N. J. Maihle Soluble Epidermal Growth Factor Receptor (sEGFR/sErbB1) as a Potential Risk, Screening, and Diagnostic Serum Biomarker of Epithelial Ovarian Cancer Cancer Epidemiol. Biomarkers Prev., February 1, 2003; 12(2): 103 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Abdelrahim, I. Samudio, R. Smith III, R. Burghardt, and S. Safe Small Inhibitory RNA Duplexes for Sp1 mRNA Block Basal and Estrogen-induced Gene Expression and Cell Cycle Progression in MCF-7 Breast Cancer Cells J. Biol. Chem., August 2, 2002; 277(32): 28815 - 28822. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Shimada, Y. Suzuki, S.-J. Kim, P.-C. Wang, M. Matsumura, and S. Kojima Transactivation via RAR/RXR-Sp1 Interaction: Characterization of Binding Between Sp1 and GC Box Motif Mol. Endocrinol., October 1, 2001; 15(10): 1677 - 1692. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Samudio, C. Vyhlidal, F. Wang, M. Stoner, I. Chen, M. Kladde, R. Barhoumi, R. Burghardt, and S. Safe Transcriptional Activation of Deoxyribonucleic Acid Polymerase {{alpha}} Gene Expression in MCF-7 Cells by 17{{beta}}-Estradiol Endocrinology, March 1, 2001; 142(3): 1000 - 1008. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Castro-Rivera, I. Samudio, and S. Safe Estrogen Regulation of Cyclin D1 Gene Expression in ZR-75 Breast Cancer Cells Involves Multiple Enhancer Elements J. Biol. Chem., August 10, 2001; 276(33): 30853 - 30861. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||
