![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 61, Issue 6, 1385-1392, June 2002
Department of Psychiatry, Harvard Medical School, Boston, Massachusetts; and Laboratory of Molecular and Developmental Neurobiology, McLean Hospital, Belmont, Massachusetts
 |
Abstract |
|---|
 Top
 Abstract
 Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Two newly identified, overlapping (1 bp) glucocorticoid response
elements (GREs) at 
759 and 773 bp in the promoter of the rat
phenylethanolamine N-methyltransferase (PNMT; EC
2.1.1.28) gene are primarily responsible for its glucocorticoid
sensitivity, rather than the originally identified 533-bp GRE. A
dose-dependent increase in PNMT promoter activity was
observed in RS1 cells transfected with a wild-type PNMT
promoter-luciferase reporter gene construct and treated with
dexamethasone (maximum activation at 0.1 µM). The type II
glucocorticoid receptor antagonist RU38486 (10 µM) fully inhibited
dexamethasone (1 µM) activation of the PNMT promoter, consistent with classical glucocorticoid receptors mediating
corticosteroid-stimulated transcriptional activity. Relative
IC50 values from gel mobility shift competition assays
showed that the 759-bp GRE has a 2-fold greater affinity for the
glucocorticoid receptor than the 773-bp GRE. Site-directed mutation
of the 533-, 759-, and 773-bp GREs alone or in tandem
demonstrated that the 759-bp GRE was also functionally more
important, but both the 759- and 773-bp GREs are required for
maximum glucocorticoid responses. Moreover, the 533-bp GRE, rather
than increasing glucocorticoid sensitivity of the promoter, may limit
corticosteroid responsiveness mediated via the 759- and 773-bp
GREs. Finally, the glucocorticoid receptor bound to the 759- and
773-bp GREs interacts cooperatively with Egr-1 and/or AP-2 to
stimulate PNMT promoter activity in RS1 cells treated
with dexamethasone. In contrast, glucocorticoid receptors bound to the
533-bp GRE only seem to participate in synergistic activation of the
PNMT promoter through interaction with activator protein 2.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Glucocorticoids
are critical regulators of phenylethanolamine
N-methyltransferase (PNMT; EC 2.1.1.28), the final enzyme in
epinephrine biosynthesis, exerting both transcriptional and post-transcriptional influences. In vivo studies in rats have shown
that depletion of corticosteroids by hypophysectomy decreases PNMT mRNA
and enzyme expression (Evinger et al., 1992
; Wong et al., 1992a, 1995,
1996; Evinger, 1998). These changes can be reversed by administration
of adrenocorticotropin, which stimulates endogenous glucocorticoid
production, or direct corticosteroid replacement by administration of
the synthetic glucocorticoid dexamethasone. Changes in PNMT enzyme are
a consequence of alterations in both gene transcription and proteolytic
degradation (Berenbeim et al., 1979; Wong et al., 1985). In terms of
the latter, corticosteroids sustain methionine adenosyltransferase and
S-adenosylhomocysteine hydrolase, the metabolic enzymes
responsible for maintaining the cosubstrate and methyl donor,
S-adenosylmethionine. Sufficient AdoMet is thereby provided
for PNMT enzymatic activity; in addition, however, the binding of
AdoMet to PNMT protects it against proteolysis.
When intact rats are administered either dexamethasone or the
glucocorticoid agonist RU28362, PNMT mRNA levels rise markedly (Wong et
al., 1992b) because of increased gene transcription. Although it
remains unclear whether glucocorticoids are essential for PNMT
transcriptional activity, glucocorticoid receptor-deficient mice do not
express adrenal medullary PNMT although chromaffin cells are otherwise
ostensibly normal (Schmid et al., 1995; Finotto et al., 1999).
Glucocorticoid-induced transcriptional changes are mediated through
glucocorticoid response elements (GREs) in the proximal 5' flanking
sequences of the PNMT gene promoter. At least one putative
GRE has been identified for every species-specific PNMT gene, including human (Baetge et al., 1988; Kaneda et al., 1988), cow
(Baetge et al., 1986; Batter et al., 1988), rat (Ross et al., 1990),
and mouse (Morita et al., 1992). In the case of the rat PNMT
gene, a GRE was identified at 533 bp when the gene was first cloned
(Ross et al., 1990). Although this GRE seemed to be functional, its
responsiveness to glucocorticoid activation seems both variable and
weak. At best, glucocorticoid treatment (1 µM dexamethasone) elicits
no greater than a 2-fold induction of the PNMT promoter as
demonstrated through transient transfection assays with PNMT promoter-luciferase reporter gene constructs (Ebert et al., 1998) or
changes in PNMT mRNA measured by ribonuclease protection assays (Morita
et al., 1996). However, glucocorticoid-activated glucocorticoid receptors (GR), bound to the 533-bp GRE, seem to interact
cooperatively with other transcriptional activators bound to their
cognate recognition sites [e.g., the immediate early gene
transcription factor Egr-1 (Ebert et al., 1994) and the developmental
transcription factor AP-2 (Ebert et al., 1998)] to synergistically
stimulate the PNMT promoter.
This study is the first to definitively identify the primary GREs
mediating the glucocorticoid responsiveness of the rat PNMT gene. The sites at 759 and 773 bp have been characterized
extensively and their functionality has been established.
Glucocorticoid receptors bound to the GREs are further shown to
participate in cooperative interactions with two other PNMT
transcriptional activators, Egr-1 and AP-2, which is probably important
for their biological activity. Finally, it is demonstrated that
glucocorticoid receptor activation of these GREs and/or their
synergistic interactions with Egr-1 and AP-2 also stimulates the
endogenous PNMT gene in a manner consistent with their
stimulation of the PNMT promoter, whereas the original GRE
(533 bp) only shows apparent synergism with AP-2.
| |
Materials and Methods |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Plasmids and Oligonucleotides.
The wild-type construct
pGL3RP893 was generated by subcloning the proximal 863 bp of proximal
rat PNMT promoter sequences into the plasmid pGL3Basic
(Pharmacia-Upjohn, Kalamazoo, MI). Verification of the promoter
fragment by DNA sequence analysis (Her et al., 1999) showed that the
insert was 893 bp in length, rather than 863 bp as identified
originally (Ross et al., 1990). The difference arises in GC-rich
regions where G and C residue determination may be difficult and is
consistent with a recent report on this proximal extent of
PNMT promoter (Evinger, 1998). Hence, the full-length
construct was redesignated pGL3RP893.
) into the corresponding
restriction sites of the pGL3Basic vector upstream of the luciferase
reporter gene. The remaining mutant constructs, pGL3RP790mut759,
pGL3RP790mut773, and pGL3RP790mut759/773, were produced from pGL3RP790
using the polymerase chain reaction as described below.
For gel mobility shift assays, protein-DNA complexes were formed using
the wild-type 40-bp oligonucleotide GRE773/759. Competitor oligonucleotides included unlabeled GRE773/759, GRE773, GRE759, and
palGRE, with nucleotide sequences as follows:
5'GTACCAAGAATGTGTTCTGCACTCTCTGTTCTTACACGAG3' (790 
751,
GRE773/759); 5'GTACCAAGAATGTGTTCTGCA3' (790 770, GRE773);
5'TTCTGCACTCTCTGTTCTTAC3' (776 756, GRE759); and 5'AGAGGATCTGTACAGGATGTTCTAGAT [palindromic (Scheidereit and Beato, 1984) or palGRE].
Egr-1 (Sukhatme et al., 1988) and AP-2 (Mitchell et al., 1987)
expression and control constructs were kindly provided by Dr. Vikas
Sukhatme (Harvard Medical School, Boston, MA) and Dr. Trevor Williams
(Yale University, New Haven, CT) respectively.
Transient Transfection Assays.
Transient transfection assays
were executed as described previously (Her et al., 1999; Tai et al.,
2001) in the rat pheochromocytoma-derived RS1 cells (Ebert et al.,
1994). Briefly, cells were maintained in Dulbecco's modified Eagle's
medium supplemented with 5% bovine calf serum, 5% equine serum
(Hyclone, Logan, UT), 20 units/ml of hygromycin B (Calbiochem, La
Jolla, CA), and 50 µg/ml gentamicin sulfate (U.S. Biochemicals Corp.,
Cleveland, OH) at 37°C in an atmosphere of 5%
CO2/95% air. All sera were charcoal-treated to remove endogenous glucocorticoids.
) including 0.5 to 1.0 µg of
wild-type or mutant PNMT promoter-luciferase reporter gene construct, 0 to 1.5 µg of expression or null construct, pCMVEgr-1 or
pCMVETTL (Gupta et al., 1991) and/or pSPRSV-AP-2 or pSPRSV-NN (Williams
and Tjian, 1991), and 0.3 µg of pRSV-LacZ normalization control
construct. Total transfected DNA was adjusted to 3.0 µg by making up
the difference with the pGL3Basic plasmid vector. After transfection,
cells were exchanged to culture medium and maintained an additional
24 h. To examine the effects of dexamethasone and the
antiglucocorticoid RU38486 (Roussel-UCLAF, Romaineville, France),
transfected cells were exposed to dexamethasone (0-10 µM; Sigma
Chemical, St. Louis, MO) for 6 h or pretreated with the antagonist
for 1 h (0-10 µM), followed by 1 µM dexamethasone treatment
for 6 h. The cells were then collected, lysed, and assayed for
luciferase and 
-galactosidase as described below.
Luciferase and -Galactosidase Assays.
After removal of
the culture medium, the cells were washed twice with phosphate-buffered
saline and then lysed in 100 µl of lysis buffer (Promega, Inc.,
Madison, WI). Cellular debris was removed by centrifugation at
800g and luciferase activity measured in 20 µl of cell
lysate appropriately diluted to yield luciferase activity within the
linear range as defined with purified luciferase using the Luciferase
Assay System described previously (Ebert et al., 1994). Protein was
determined by the method of Bradford (1976) and luciferase activity was
adjusted for protein concentration. To correct for variation in
transfection efficiency, -galactosidase activity was also determined
(Ebert et al., 1994) and luciferase activity expressed relative to
-galactosidase generated from the pRSV-LacZ control construct. As
appropriate, the ratio of luciferase/-galactosidase for the
wild-type PNMT promoter-luciferase reporter gene construct
was set to unity, and the ratio of luciferase/-galactosidase for
other constructs expressed relative to it. Alternatively, the ratio of
luciferase/-galactosidase expressed by PNMT
promoter-luciferase constructs, wild-type, truncated or mutant
constructs, in the absence of dexamethasone, was set to unity, and
values in the presence of dexamethasone expressed relative to these
respective untreated controls. At least six replicates were included in
each sample group and experiments repeated two to three times.
Site-Directed Mutagenesis.
Site-directed mutation of the
759- and 773-bp GREs was performed by polymerase chain reaction as
described previously using the following 40-bp mutagenic primers,
spanning upstream sequences from 790 to 751 bp in the
PNMT promoter (Her et al., 1999): 5'GTACCAAGAATGTcaTCTGCACTCTCTGTTCTTACACGAG3' (mut773);
5'GTACCAAGAATGTGTTCTGCACTCTCTcaTCTTACACGAG3' (mut759); and
5'GTACCAAGAATGTcaTCTGCACTCTCTcaTCTTACACGAG3' (mut773/759).
Gel Mobility Shift Competition Assays.
As described
previously (Her et al., 1999), protein-DNA complexes were generated
using 1 ng of the wild-type oligonucleotide probe GRE773/759,
end-labeled with [
-32P]ATP and
T4 polynucleotide kinase (3 nM, specific
activity = 2.5 × 108 dpm/µg), and a
truncated GR protein (Dr. Keith Yamamoto, University of California, San
Francisco) (Freedman et al., 1988) in 20 µl of binding buffer
consisting of 25 mM HEPES buffer, pH 7.9, 50 mM KCl, and 0.05 mM EDTA.
Complexes were competed by including unlabeled oligonucleotides ranging
in concentration from 1 to 1000 ng. The amount of residual complex was
then quantified from its autoradiographic signal by scanning
densitometry using Image software, version 1.52 (National Institutes of
Health), on a Power MacIntosh 7500 computer and a Hewlett Packard 6100C
scanner. IC50 values were determined by linear
regression analysis of graphs of optical density units versus
ln[competitor oligonucleotide concentration] and calculation of the
x-intercept.
Statistical Analysis.
Data are presented as the mean ± S.E.M. with n = 6 for each experimental group.
Statistical significance between groups was determined by one-way
analysis of variance followed by post hoc comparison using Student's
t test. A p value 
0.05 was considered statistically significant.
| |
Results |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Glucocorticoid Responsiveness of the PNMT
Promoter.
To examine the glucocorticoid responsiveness of the
PNMT promoter, the proximal 893 bp of 5' promoter/regulatory
sequences were subcloned into the plasmid pGL3Basic (Pharmacia-Upjohn,
Kalamazoo, MI) upstream of the firefly luciferase reporter gene. When
this wild-type construct (pGL3RP893) was transiently transfected into the PC-12-derived RS1 cells (Ebert et al., 1994) and the cells treated
with 1 µM dexamethasone, luciferase activity was induced ~6.0-fold,
markedly higher than reported previously (Fig.
1).
|
745 and 798 bp were responsible for the marked glucocorticoid
induction of the promoter (Fig. 1). Two of three plasmid constructs
(pGL3RP557 and pGL3RP665) harboring only the proximal 533-bp GRE
(previously designated 513 bp) (Ross et al., 1990) did not show any
apparent corticosteroid activation, whereas a third (pGL3RP745) showed a 2-fold stimulation of the PNMT promoter. In addition,
sequences beyond 798 bp seemed to limit the glucocorticoid
responsiveness of the PNMT promoter. Highest dexamethasone
stimulation of PNMT promoter-driven luciferase activity
occurred with the construct containing 798 bp of promoter sequence
(pGL3RP798). Longer constructs expressed lesser amounts of luciferase;
the full-length construct pGL3RP893 showed only a 6.0-fold induction by
comparison to the 12.0-fold induction seen with the construct pGL3RP798.
Matching of the DNA sequences in the glucocorticoid sensitive region to
the consensus glucocorticoid response element, 5'TAGAACANNNTCTTCT3' (Scheidereit and Beato, 1984), using the Transfac database (version 3.2) identified two new GREs. Based on the highly conserved 3' hexanucleotide sequence of the consensus GRE, TGTTCT, the 3' termini of
these GREs were fixed at 759 and 773 bp, with a 1-bp overlap of
their 5' and 3' ends, respectively.
Effects of Dexamethasone and RU38486 on PNMT
Promoter Activation.
Because longer constructs seemed to contain
sequences inhibiting full glucocorticoid responsiveness of the
PNMT promoter, the construct pGL3RP790, which shows
comparable dexamethasone sensitivity to pGL3RP798, was used in
subsequent studies (data not shown). A dose-response curve was first
executed for dexamethasone (Fig. 2A).
When RS1 cells transfected with pGL3RP790 were treated with 0 to 10 µM dexamethasone for 6 h, a dose-dependent rise in PNMT promoter-driven luciferase reporter gene expression was
observed. Maximum stimulation of the promoter occurred at 0.100 µM.
No significant change in activation was observed when the dexamethasone
concentration was increased to 1 µM dexamethasone. However,
luciferase activity decreased to levels equivalent to 0.01 µM
dexamethasone when corticosteroid levels were increased to 10 µM.
|
Glucocorticoid Receptor Binding to the Upstream GREs and Activation
of the PNMT Promoter.
The previous results are
consistent with type II GRs mediating PNMT promoter
transcriptional activation. To demonstrate the specificity of the GR
and its relative binding affinity for the newly identified GRE target
sequences, gel mobility shift competition assays were executed (Fig.
3). Protein-DNA binding complex was formed between a 40-bp 32P-labeled wild-type
oligonucleotide spanning both the 759- and 773-bp GREs and 5 bp of
5' and 3' flanking sequence and a truncated GR protein (Dr. Keith
Yamamoto, University California, San Francisco). The complex was
competed with increasing amounts of unlabeled oligonucleotide (1-1000
ng), including the unlabeled 40-bp oligonucleotide and 21-bp
oligonucleotides encoding the 759-bp, 773-bp, and palindromic GRE
sequences (Fig. 3A). All oligonucleotides interacted with the GR
peptide as demonstrated by their displacement of the radiolabeled DNA
(Fig. 3B). The abundance of GR-GRE binding complex was quantified by
scanning densitometry and relative IC50 values determined by regression analysis of signal intensity versus
ln[competitor concentration] (Fig. 3, B and C). The palindromic GRE
and the oligonucleotide harboring both upstream GREs had the highest
affinity for the GR (relative IC50 values: 11.0 and 8.7, respectively). The affinities of the 759- and 773-bp GREs
for the GR were ~3.5 to 7.0-fold lower based on relative
IC50 values; the 759-bp GRE had ~2-fold
higher affinity (relative IC50, 30.6) for the GR
than the 773-bp GRE (relative IC50, 60.8). The
slopes of the regression curves for each competition assay did not
significantly change, confirming that the differences in x-intercepts
(IC50) were caused solely by affinity of the DNA
sequences for the GR and not differences in receptor abundance as well.
|
759- and 773-bp GREs,
site-directed mutations were introduced into the various GREs as
described under Materials and Methods to produce the single mutant constructs pGL3RP790mut533, pGL3RP790mut759, and
pGL3RP790mut773; the double mutant construct pGL3R790mut773/759; and
the triple mutant construct pGL3RP790mut533/759/773. The wild-type or
mutant constructs were then transiently transfected into RS1 cells and PNMT promoter-driven luciferase activity was determined in
the absence or presence of dexamethasone (1 µM) (Fig.
4A). When the 533-bp GRE was mutated,
rather than a decrease, a 
1.5-fold increase in
dexamethasone-stimulated PNMT promoter activation occurred at dexamethasone concentrations between 0.01 and 1 µM. By contrast, mutation of either the 759 or 773-bp GRE markedly attenuated dexamethasone-stimulated PNMT promoter activity. However,
when the 773-bp GRE was mutated, a dose-dependent increase in
PNMT promoter-driven luciferase activity was still apparent,
although maximum induction was only 3.1-fold (1 µM dexamethasone). As
observed with all GREs intact, 10 µM dexamethasone stimulated the
PNMT promoter less than 1 µM dexamethasone. When the
759-bp GRE was mutated, PNMT promoter-driven luciferase
expression was not significantly different from untreated, control
values.
|
533-bp GRE (pGL3RP790mut533) showed
PNMT promoter-driven luciferase expression similar to that
of the wild-type construct (pGL3RP790). At concentrations of 0.1 and 1 µM, luciferase activity was, in fact, slightly higher than the
wild-type control cells. The construct harboring a mutation in the
773-bp GRE (pGL3RP790mut773) showed a linear decrease in
dexamethasone-stimulated, PNMT promoter-driven luciferase
expression as would be expected based on its responses to dexamethasone
described above. Similarly, the mutant construct with altered 759-bp
GRE showed no significant dexamethasone and RU38486 responsiveness.
Finally, when the 759- and 773-bp or the 533-, 759-, and
773-bp GREs were mutated, neither dexamethasone nor the combination
of dexamethasone and RU38486 elicited any significant changes in
PNMT promoter expression whatsoever.
Cooperative Interactions between the GR and Other
PNMT Transcriptional Activators.
Many
transcriptional regulators can independently and cooperatively
stimulate gene expression. Although the 533-bp GRE in the rat
PNMT promoter is apparently a weak independent
glucocorticoid activation site, bound GR seemed to cooperatively
activate the PNMT promoter through interactions with Egr-1
and/or AP-2 bound to their respective consensus elements at 165,
674, and 587 bp (Wong et al., 1998). As described earlier (Fig. 1),
two deletion constructs, pGL3RP557 and pGL3RP665, which harbor the
533-bp GRE but not the 759 or 773-bp GREs, showed no significant
glucocorticoid activation. However, dexamethasone did induce luciferase
reporter gene expression 2.0-fold from a slightly longer construct,
pGL3RP745, a construct containing functional AP-2 binding sites (Ebert
et al., 1998).
533-bp GRE and the possible
role of the 759- and 773-bp GREs in cooperative stimulation of the
PNMT promoter, transient transfection assays were executed with the wild-type and mutant PNMT promoter-luciferase
reporter gene constructs in the absence or presence of 1 µM
dexamethasone and Egr-1 and/or AP-2 expression constructs. First,
cooperativity between the GR and Egr-1 was investigated (Fig.
5A). Although dexamethasone stimulated
only a 3.0-fold rise in PNMT promoter-driven luciferase
activity in this case, mutation of the 533-bp GRE increased
dexamethasone stimulation of the promoter 3.8-fold beyond that observed
with the wild-type construct. In contrast, mutation of the 759-bp
GRE, 773-bp GRE, both the 759- and 773-bp GREs, or the 533-,
759-, and 773-bp GREs almost completely abolished the
glucocorticoid responsiveness of the PNMT promoter. Egr-1 alone stimulated a 2.6-fold increase in PNMT promoter-driven
luciferase expression in the wild-type PNMT
promoter-luciferase construct and in combination with dexamethasone,
increased promoter activity slightly more than the additive inductions
by the GR and Egr-1, consistent with GR and Egr-1 acting cooperatively
to induce PNMT promoter-driven transcriptional activity.
However, if the 533-bp GRE was mutated, Egr-1 or Egr-1 combined with
dexamethasone stimulated the PNMT promoter 2- to 4-fold more
than when the site was intact (4.5 and 9.8-fold, respectively). When
the 759-bp GRE, the 773-bp GRE, the 759- and 773-bp GREs, or
the 533-, 759-, and 773-bp GREs were mutated, stimulation of the
PNMT promoter by Egr-1 and Egr-1 in combination with
dexamethasone declined precipitously. However, Egr-1-mediated,
PNMT promoter induction was still significantly greater than
their respective controls in the case of the 759- and
533-/759-/773-bp mutant constructs (5.0- and 2.0-fold, respectively). It was previously demonstrated that AP-2 induction of
the PNMT promoter required coactivation of the GR (Ebert et al., 1998). In the present study, AP-2 seems to independently stimulate
PNMT promoter-driven luciferase activity in the case of the
wild-type construct (3.7-fold), but synergistic activation by AP-2 and
dexamethasone still occurred (20.7-fold). When the 533- or 773-bp
GREs were mutated, similar responses were observed, although induction
was markedly attenuated. However, if the 759-bp, 759- and 773-bp,
or 533-, 759-, and 773-bp GREs were mutated, both the independent
AP-2 and GR and AP-2 synergistic stimulation of the PNMT
promoter was eliminated.
|
), greatest synergistic
activation of the PNMT promoter was elicited in the presence of all three transcriptional activators, the GR, AP-2, and Egr-1 (~28.0-fold). Moreover, mutation of the 533-bp GRE increased PNMT promoter activity 4.0-fold beyond that observed with
the wild-type construct (91.5-fold). In contrast, mutation of either the 759 or 773-bp GREs markedly attenuated synergistic stimulation of the promoter (25% of wild-type) although a significant 6.0- to
7.9-fold activation remained. Finally, when both the 759- and
773-bp GREs as well as all three GREs were mutated, cooperative interactions were completely lost.
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Glucocorticoids are important regulators of PNMT gene
expression, influencing adrenergic differentiation (Bohn et al., 1981; Teitelman et al., 1982; Bohn, 1983; Michelson and Anderson, 1992; Wong
et al., 1992a; Schmid et al., 1995; Ebert et al., 1997) and the
induction of PNMT in response to acute and chronic stress (Sabban et
al., 1995; Sabban et al., 1998; Serova et al., 1998). Consistent with
this role, a functional GRE was identified at 513 bp in the upstream
sequences of the rat PNMT gene when it was first cloned
(Ross et al., 1990). At best, however, only weak glucocorticoid
responses are elicited through this GRE based on changes in
PNMT promoter-reporter gene expression and PNMT mRNA expression in vitro (Wong et al., 1996; Ebert et al., 1997). Two additional glucocorticoid response elements have now been identified distal to the original GRE. These GREs, located at 759 and 773 bp
upstream of the site of transcription initiation, overlap by 1 bp
according to the consensus sequence defined by Scheidereit et al.
(1983) and together contribute to a maximum 12.0-fold induction of the
PNMT promoter in response to glucocorticoids.
The position of these new GREs has been designated by their 3' termini
based on the highly conserved 3'-hexanucleotide sequence TGTTCT in the
15-bp palindromic GRE, 5'AGAACANNNTCTTCT3' identified above. These
designations also correct for misalignment of the proximal
PNMT promoter sequences arising from ~30 bp of G and C
residues in GC rich regions where sequencing is difficult. Realignment of the promoter repositions the 513-bp GRE at 533 bp as well.
The relative contribution of each new GRE to the corticosteroid
responsiveness of the PNMT promoter was investigated by
site-directed mutagenesis, gel mobility shift competition assays, and
examination of the effects of dexamethasone and the
antiglucocorticoid RU38486. Both GREs must be intact to elicit full
glucocorticoid sensitivity of the PNMT promoter, and the
response seems greater than additive, indicating cooperativity between
activated GRs bound to the GREs or bound GRs and other transcription
factors. In addition, the 759-bp GRE has ~2-fold higher affinity
for the GR, which was reflected by a greater attenuation of
glucocorticoid-stimulated PNMT promoter activity when this
GRE was mutated. Thus, this GRE is probably functionally more
important. Finally, RU38486 effectively blocked glucocorticoid
activation from either response element in a dose-dependent manner,
consistent with the relative affinities of the GR for each GRE. The
latter results also confirm that glucocorticoid activation of the
PNMT gene promoter occurs through type II GRs because
RU38486 is a classic GR antagonist.
In keeping with previous reports, the 533-bp GRE did not seem to
markedly affect PNMT promoter activation through
glucocorticoid exposure alone. Two PNMT promoter constructs
(pGL3RP665 and pGL3RP557) containing this GRE, but not the 773- and
759-bp upstream GREs, failed to show an increase in PNMT
promoter activity when treated with corticosteroids. However, a 2-fold
induction of the luciferase reporter gene was observed with a slightly
longer construct pGL3RP745. The latter also harbors two functional AP-2
binding elements at 674 and 587 bp (Ebert et al., 1998). In
contrast to our earlier results, the present findings suggest that AP-2
alone can elicit a significant, but limited induction of the
PNMT promoter. Most notable again is the marked cooperative
induction of the promoter by GR bound to the 533-bp GRE and AP-2.
AP-2 and dexamethasone synergism is demonstrated by the nearly 2.0-fold
reduction (relative to wild-type) in PNMT promoter-driven
luciferase activity when the 533-bp GRE is mutated. Curiously, when
the 759- and 773-bp GREs are mutated and the 533-bp GRE left
intact, synergism disappears. It may be that binding of GRs to these
GREs alters PNMT promoter conformation in a fashion that
favors the interaction between GRs bound to the 533-bp GRE and AP-2.
Alternatively, GR bound to the 533, 759, and 773-bp GREs may
interact with one another and/or AP-2. Whereas the 533-bp GRE
participates in AP-2 and GR cooperative activation of the
PNMT promoter, synergism between bound GR and the immediate
early gene transcription factor Egr-1 does not occur, because there is
no significant difference in their combined effects on PNMT
promoter activity when this site is mutated. The latter results further
suggested that previously reported cooperative induction by Egr-1 and
dexamethasone might involve the newly identified tandem GREs. When the
533-bp GRE is mutated, leaving only these two GREs intact, Egr-1 and
dexamethasone elicit a 5.0-fold higher stimulation of the
PNMT promoter than observed with the wild-type construct
where all three GRE sites are intact. Moreover, when these GREs are
mutated independently, together or along with the 533-bp GRE,
activation by Egr-1 and dexamethasone is effectively eliminated. Thus,
the 759- and 773-bp GREs do seem to be the essential GREs for
cooperative activation of the PNMT promoter by Egr-1 and the
GR. In addition, the 759- and 773-bp GREs also participate in
synergistic activation of the PNMT promoter with AP-2,
although their contribution to GR and AP-2 activation of the promoter
is less than that orchestrated through the 533-bp GRE. When the
533-bp GRE is mutated, AP-2 still elicited a residual ~12.0-fold
stimulation of PNMT promoter-driven luciferase expression.
Finally, these independent synergistic effects are reflected in the
combined effects of AP-2, Egr-1, and the GR on the wild-type and mutant constructs.
Clearly, these cooperative interactions are complex and very dependent
on promoter length, acetylated histones, and DNA folding and
interaction as well as the availability of coactivator complexes containing factors such as SCRC1, GRIP1, CBP, P300, and PCAF (Wang et
al., 1999). Current studies are now investigating the effects of
selective silencing of the GREs, Egr-1 and AP-2 binding elements in
PNMT promoter constructs and endogenous PNMT gene
using viral vector driven antisense strategies.
In summary, the present study provides the first definitive
identification and characterization of the functional GREs in the
proximal sequences in the rat PNMT promoter. Through the
newly identified 759- and 773-bp GREs, marked and selective
glucocorticoid activation occurs, indicating that they are the primary
targets through which glucocorticoid sensitivity is conferred. In
addition, both the 533-bp GRE and these newly identified GREs seem to
participate in cooperative or facilitatory activation of the
PNMT promoter, the former with AP-2 and the latter with
AP-2, Egr-1, and/or both.
| |
Footnotes |
|---|
Received January 23, 2002; Accepted March 15, 2002
This work was supported by grant DK51025 from the National Institute of Diabetes and Digestive and Kidney Diseases, by the Spunk Fund, Inc., by the Sobel and Keller Research Support Fund, and by McLean Hospital.
Address correspondence to: Dona Lee Wong, Ph.D., Department of Psychiatry, Harvard Medical School, Laboratory of Molecular and Developmental Neurobiology, McLean Hospital, 115 Mill Street, MRC #116, Belmont, MA 02478. E-mail: dona_wong{at}hms.harvard.edu
| |
Abbreviations |
|---|
PNMT, phenylethanolamine
N-methyltransferase;
RU28362, 17-
-alkanyl-11,17-dihydroxy-androsterone;
GRE, glucocorticoid
response element;
GR, glucocorticoid receptor;
AP-2, activator protein
2;
RU38486, mifepristone.
| |
References |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
and HNF4 and the necessity of proper alignment of their cognate binding sites.
Mol Endocrinol
13:
604-618This article has been cited by other articles:
|
|
 |
|
  R. Kvetnansky, E. L. Sabban, and M. Palkovits Catecholaminergic Systems in Stress: Structural and Molecular Genetic Approaches Physiol Rev, April 1, 2009; 89(2): 535 - 606. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Root, J. Abrassart, D. A. Myers, T. Monau, and C. A. Ducsay Expression and Distribution of Glucocorticoid Receptors in the Ovine Fetal Adrenal Cortex: Effect of Long-term Hypoxia Reproductive Sciences, May 1, 2008; 15(5): 517 - 528. [Abstract] [PDF]
|
|
|
|
 |
|
 C. A. Ducsay, K. Hyatt, M. Mlynarczyk, B. K. Root, K. M. Kaushal, and D. A. Myers Long-term hypoxia modulates expression of key genes regulating adrenomedullary function in the late gestation ovine fetus Am J Physiol Regulatory Integrative Comp Physiol, November 1, 2007; 293(5): R1997 - R2005. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Newton and N. S. Holden Separating Transrepression and Transactivation: A Distressing Divorce for the Glucocorticoid Receptor? Mol. Pharmacol., October 1, 2007; 72(4): 799 - 809. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. C. Tai, D. C. Wong-Faull, R. Claycomb, and D. L. Wong Nerve Growth Factor Regulates Adrenergic Expression Mol. Pharmacol., November 1, 2006; 70(5): 1792 - 1801. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Adams, O. C. Meijer, J. Wang, A. Bhargava, and D. Pearce Homodimerization of the Glucocorticoid Receptor Is Not Essential for Response Element Binding: Activation of the Phenylethanolamine N-Methyltransferase Gene by Dimerization-Defective Mutants Mol. Endocrinol., December 1, 2003; 17(12): 2583 - 2592. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
