| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
The Division of Endocrinology, Department of Medicine, Children’s Hospital, Harvard Medical School (Y.I., J.A.M), Boston, Massachusetts 02115; and the First Department of Internal Medicine, Nagoya University School of Medicine (Y.I., Y.O., H.S.), Nagoya 466, Japan
Address all correspondence and requests for reprints to: Joseph A. Majzoub, M.D., Division of Endocrinology, Children’s Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115. E-mail: majzoub{at}a1.tch.havard.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The vasopressin gene is approximately 2 kb in size and consists of 3 exons and 2 introns, encoding a signal peptide, vasopressin, neurophysin II, and a carboxy-terminal glycopeptide (7, 8). After transcription, splicing and translation, the preprohormone, preprovasopressin, is cleaved by signal peptidase, making provasopressin, which contains the latter three peptides. The prohormone is sorted from the endoplasmic reticulum to secretory granules through the Golgi apparatus, during which it undergoes posttranslational modifications such as processing, amidation, and glycosylation, to yield the mature peptide secretory products (9).
Regulation of vasopressin secretion is relatively well characterized (2). Also, using immunohistochemical and molecular biologic techniques, the differential regulation of vasopressin expression in vivo has recently been clarified, with vasopressin expression in magnocellular neurons of the SON and PVN being stimulated by osmotic and hypovolemic stimuli, and that in parvocellular neurons being negatively regulated by glucocorticoid (10, 11). However, the mechanisms of transcriptional regulation of the vasopressin gene within the cell are not well known. Hypothalamic vasopressin messenger RNA (mRNA) content is increased after water deprivation in hyperosmolar rats (12, 13, 14, 15, 16, 17). Osmotic stimulation enhances cAMP production in the SON in the rat (18, 19). cAMP is known to stimulate vasopressin gene expression (20, 21), and putative cAMP-responsive elements (CREs) have been identified in the bovine vasopressin gene promoter (21). Together, these data suggest a role for this pathway in the positive regulation of vasopressin gene transcription. In contrast, glucocorticoid has been shown to exert a negative effect on vasopressin gene expression (22). The negative effect of glucocorticoid on vasopressin gene expression may have clinical significance. Vasopressin synthesized in hypothalamic parvocellular neurons, which is involved in stress-induced ACTH secretion, is negatively regulated by glucocorticoid (4, 10). Also, glucocorticoid receptor (GR) are present in both parvocellular (23) as well as magnocellular vasopressinergic neurons, although in the latter only during hypoosmolality (24); this may explain the vasopressin hypersecretion frequently observed in patients with chronic glucocorticoid insufficiency (25).
To study the transcriptional regulation of the vasopressin gene in vitro, we isolated 3 kb of the 5' promoter region of the rat vasopressin gene, as well as a series of promoter deletion mutants, and fused these to a luciferase reporter gene. Because no homologous cell line derived from vasopressinergic neuron is available, these constructs were expressed in the human choriocarcinoma cell line JEG-3. Changes in promoter activity by various reagents were quantitated using the luciferase assay. We examined the roles of various intracellular signaling systems of potential physiological importance, including the protein kinase A (PKA) and GR pathways, in the regulation of vasopressin gene expression.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Plasmid constructions
An 8.2-kb EcoRI DNA fragment of the rat vasopressin
gene, including 3-kb 5' flanking region, 2-kb structural gene
containing three exons and two introns, and 3.2-kb 3' flanking region,
was isolated from a rat genomic library in lambda phage as previously
reported (26). To isolate the 3-kb 5' flanking region of the gene, 292
bp of downstream region (positions -266 to +26; transcription start
site is designated as +1) of the promoter sequence was first
synthesized by the PCR with Taq DNA polymerase
(Perkin-Elmer/Cetus Corp., Norwalk, CT) using sense and antisense
oligonucleotide primers containing XhoI and KpnI
restriction sites, respectively (shown by bold type) as
described below:
Sense: 5'-TTTCTCGAGCCTCCCTGATTGCACAGCAC-3'
Antisense: 5'-TTTGGTACCGGCACTGCGTGCAGCTCT-3'
The PCR product was subcloned into the XhoI/KpnI sites of pBluescript IISK+ plasmid (Stratagene, La Jolla, CA), and the sequence was confirmed by dideoxy chain termination method. Then the remaining 2.7-kb length upstream region of the 5' flanking sequence was isolated from the original gene by endonuclease digestion, and ligated at the BalI site (-211) of the downstream region, thus completing the 3-kb full-length 5' promoter (AVP3000; -3000 to +26). Six serial shorter deletion mutants of the 5' promoter genes were also made by endonuclease digestions at unique restriction sites (AVP803, AVP441, AVP266, AVP211, AVP102, and AVP36; -803, -441, -266, -211, -102, and -36 to +26, respectively). Also, various mutants, in which defined parts of the promoter sequences were deleted from AVP441, were made by site-directed mutagenesis (27): [for AVP441D250-201, AVP441D150–101, and AVP441D100–71; -250 to -201, -150 to -101, and -100 to -71 bp of the promoter sequences were deleted, respectively; for AVP441DCRE1 and AVP441DCRE2, CRE1 (-227 to -220) and CRE2 (-123 to -116) were deleted, respectively; for AVP441DCAAT, the CCAAT box (-247 to -243) was replaced with GTCGA].
To make the vasopressin promoter/luciferase fusion genes, each promoter sequences was subcloned into BamHI/KpnI sites of a pXP1 plasmid containing the luciferase reporter gene (28). Some of the vasopressin promoter constructs were also subcloned into a different luciferase vector pA3Luc (29). As a positive control, the long terminal repeat of the rous sarcoma virus (RSV) promoter was subcloned into the pXP1 plasmid (pXP1/RSV).
To make a competitor plasmid to bind intracellular activator protein 2 (AP2) transcription factor, two oligonucleotides containing tandem AP2 consensus sequences found between -95 and -75 of the rat vasopressin promoter were synthesized as shown below (AP2 consensus sequences are shown by bold type):
Sense: 5'-GATCCTCCCCAGTGGCTCCCCAGGAGA-3'
Antisense: 5'-AGCTTCTCCTGGGGAGCCACTGGGGAG-3'
These two oligonucleotides were annealed, and cloned between HindIII/BamHI sites of the pBluescript IISK+ plasmid. pBluescript IISK+ was used as a control.
Transient plasmid transfections
JEG-3 cells were transfected transiently with each sequentially
deleted or mutated vasopressin-luciferase fusion gene along with other
expression vectors (as indicated) using the standard calcium phosphate
precipitation method (30). Cells of approximately 50% confluency in
60-mm diameter culture dishes were incubated for 3–4 h with
CaPO4 precipitates containing 3 µg of DNA in total [0.6
µg of each vasopressin plasmid, 0.3 µg of GR expression vector
(either wild-type or mutant), 0.3 µg of a human GH (hGH) expression
vector (pRSV-hGH) as an internal control (when indicated) (31), other
plasmids for cotransfection experiments (as indicated), and pBS(+)
plasmid (Stratagene) as carrier plasmid to keep the total amount of
transfected DNA constant], followed by 2 min of 15% glycerol shock.
Cells were used for experiments 24–48 h after the start of the
transfection.
Evaluation of promoter activity
For the initial characterization, AVP3000 full-length
promoter-luciferase fusion gene was used. In the subsequent
experiments, AVP441, which contains all the putative CREs, was used to
study the positive regulation by cAMP-PKA pathway, whereas AVP803,
which includes putative glucocorticoid-responsive element (GRE) (-622
to -608), was used to study negative regulation by glucocorticoid.
Cells were treated with forskolin, an activator of adenylate cyclase, 8-bromo-cAMP (8Br-cAMP), a membrane permeable cAMP analog (Sigma Chemical Co., St. Louis, MO), the phorbol ester 12-O-tetradecanoylphorbol 13-acetate (TPA; Sigma), an activator of protein kinase C (PKC), or dexamethasone (Sigma), a potent synthetic glucocorticoid, with defined concentrations and time intervals. When the expression of different plasmid constructs was compared in the same experiment, pRSV-hGH (31) was cotransfected as an internal control, and hGH secreted into culture medium of each dish was measured by human GH RIA kit (Daiichi Isotope Co., Tokyo, Japan). Forskolin and dexamethasone were dissolved in ethanol, and TPA was dissolved in dimethyl sulfoxide, whereas 8Br-cAMP was dissolved in culture medium. Stock solutions for all reagents were made in 1000x concentrations except 8Br-cAMP (100x maximum), and 0.1% volume of the stock solution or solvent alone was added directly into the culture medium of each dish.
Luciferase assay
Luciferase assay was performed as previously described (32) with
some modification. After the defined time of incubation with various
reagents, cells were washed 2–3 times with phosphate buffer saline
without Ca++, Mg++, harvested with lysis buffer
containing 1% (vol/vol) Triton X100, 25 mM glycylglycine,
pH 7.8, 15 mM MgSO4, 4 mM EGTA, and
1 mM DTT, and the cell lysate was centrifuged. For the
luciferase reporter assay, 100 µl of each supernatant was added to
360 µl of luciferase assay buffer containing 25 mM
glycylglycine, pH 7.8, 15 mM MgSO4, 4
mM EGTA, 15 mM potassium phosphate, pH 7.8, 2
mM ATP, 1 mM DTT, and 0.5 mM
coenzyme A (Sigma). Reactions were mixed in a luminometer (LKB Model
1251, Pharmacia, Stockholm, Sweden; or Berthold Lumat LB9501, Bundoora,
Australia) by injection of 200 µl of the luciferin solution
containing 0.2 mM D-luciferin (Analytical
Luminescence Laboratory, San Diego, CA, or Sigma), 25 mM
glycylglycine, pH 7.8, 15 mM MgSO4, 4
mM EGTA and 2 mM DTT. Light output was measured
for 20 sec at 25 C. The net light output was calculated by subtracting
the background value from the value of each sample and displayed in
arbitrary units. When hGH was used as internal control, each value of
the net light output was divided by the hGH value of the corresponding
dish, and displayed in arbitrary units.
Data analyses
Most of the experiments were carried out more than two times.
Samples in each group of experiments were performed in triplicate or
quadruplicate. All data were expressed as mean ± SE.
When statistical analysis was performed, data were compared by one-way
ANOVA with Duncan’s multiple range test, and P values below
0.05 were considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We then studied the dose-response effect of forskolin or other
stimulation of the cAMP/PKA pathway in cells transfected with AVP441,
which contains all of the putative cAMP-responsive elements (see
below). Forskolin treatment for 5 h (0.1–100 mM)
again stimulated vasopressin gene expression in a dose-dependent
fashion, with the maximal effect at a concentration of 10
mM (Fig. 1A
, left
panel). A similar effect was observed when cells were treated for
5 h with 8Br-cAMP (0.1–5 mM), a membrane permeable
cAMP analog (Fig. 1A, right panel). We also carried out
cotransfection experiments using an expression vector of the PKA
catalytic subunit a (PKAcat), kindly provided by Dr. Richard Maurer
(33). As expected, the expression of PKAcat markedly enhanced the
vasopressin promoter transcription in a dose-dependent fashion (up to
approximately 20-fold) (Fig. 1B). Because PKA is known to be activated
by cAMP, these results indicate that the cAMP/PKA pathway is a major
positive regulator of vasopressin gene expression.
|
Regulation of vasopressin gene expression by glucocorticoid
To determine the role of glucocorticoid in the transcriptional
regulation of the vasopressin gene, we first studied the effect of
dexamethasone on vasopressin promoter transcription in JEG-3 cells
expressing AVP803, which contains a putative GRE (-622 to -608 bp).
Because JEG-3 cells express no intrinsic GR (34), the cells were
cotransfected with a GR expression vector (35), kindly provided by Dr.
Ronald Evans, along with the vasopressin-luciferase construct.
Dexamethasone by itself had no significant effect on luciferase
activity in cells transfected with either vasopressin-luciferase or
RSV-luciferase (pXP1/RSV) constructs (data not shown). Dexamethasone
(100 nM) coadministered with forskolin (10 mM)
significantly suppressed forskolin-stimulated vasopressin promoter
activity in a time-dependent fashion, with the maximal suppressive
effects (31% of forskolin-stimulated activity) observed 5 h after
the coaddition of forskolin and dexamethasone (data not shown). We
therefore subsequently used 5 h of coincubation with dexamethasone
and forskolin.
We examined the dose-dependent effect of dexamethasone on vasopressin
expression using JEG-3 cells with the GR expression vector.
Dexamethasone treatment for 5 h (1–1000 nM) at doses
as low as 1 nM suppressed forskolin-stimulated vasopressin
gene expression in a dose-dependent fashion, with maximal suppression
seen at 10 nM (Fig. 2A). To
confirm that the suppressive effect of dexamethasone is GR-dependent,
we carried out a titration experiment of GR expression vector
cotransfection. No suppression by dexamethasone (100 nM)
was observed without GR, whereas the suppressibility was enhanced as
the amount of GR cotransfected increased (Fig. 2B). These results
together indicate that glucocorticoid exerts a negative effect on
vasopressin gene expression, and the effect is GR dependent.
|
). A similar negative effect was
observed when the vasopressin pro-moter construct was cotransfected
with a mutant GR in which the N-terminal transactivation domain was
deleted (GRD77-262). In contrast, when the vasopressin promoter
construct was cotransfected with another mutant GR in which the domain
containing both zinc fingers was deleted (GRD428–490), dexamethasone
unexpectedly showed an enhancing rather than suppressive effect (Fig. 3).
|
). To determine whether these putative
cis-acting elements are responsible for the positive and
negative regulation of rat vasopressin gene promoter transcription, we
studied basal, forskolin (10 µM)-, or forskolin (10
µM) and dexamethasone (100 nM)-treated
promoter activity in a series of the 5' deletion mutant constructs
(AVP3000, 803, 441, 266, 211, 102, and 36) in JEG-3 cells (Fig. 5A). Only a minimal increase (less than
3-fold) was observed in the promoterless plasmid (data not shown).
Among the promoter constructs tested, the maximal activity was observed
with the promoters AVP803 or AVP441, which were higher than that in
full-length AVP3000, suggesting repressor activity in upstream
element(s). Shorter constructs (AVP266 to AVP36) showed a clear
decrease in absolute basal as well as forskolin-treated activity.
However, the relative stimulation by forskolin was consistently more
than 10-fold in all the deletion mutants including the shortest
construct (AVP36), which lacks both putative CREs and the putative
nurr1/nur77 binding site. Dexamethasone had a consistent suppressive
effect of 30–40% on forskolin-stimulated promoter activity, in
constructs AVP3000 to AVP102, which included those lacking a putative
GRE at nucleotides -622 to -608, and the potential nurr1/nur77 site
noted above. Of interest, negative regulation by dexamethasone was not
seen with AVP36, suggesting the possible presence of a negative GRE
between -102 and -36 bp of the vasopressin promoter. Neither
forskolin nor dexamethasone had a significant effect on RSV-luciferase
(pXP1/RSV) expression (data not shown).
|
|
).
The two deletion mutants AVP441D250–201 and AVP441D150–101, each
of which lacks a different one of the two putative CREs, when
compared with the intact promoter, AVP441, showed a substantial
decrease in the absolute level of basal and forskolin-stimulated
transcriptional activity, whereas AVP441D100–71, which lacks a tandem
putative recognition site for AP2, had normal activity (Fig. 5B). In
all constructs, however, a similar relative degree of forskolin
stimulation and dexamethasone suppression was observed.
To even more precisely identify the cis-acting element(s)
required for positive and negative regulation, we studied basal,
forskolin (10 µM)-, or forskolin (10 µM)
and dexamethasone (100 nM)-treated promoter activity in
another series of four mutant constructs; AVP441DCRE1, in which CRE1
(-228 to -221) was eliminated; AVP441DCRE2, in which CRE2 (-123
to -116) was eliminated; AVP441DCRE1/DCRE2, in which both CREs were
eliminated; and AVP441DCAAT, in which the putative cCAATc enhancer
sequence (-246 to -243) was replaced with the sequence, gTCGAc (Fig. 5C). Deletion of either CRE sequence alone had no effect, whereas
deletion of both CREs resulted in a substantial (>50%) decrease in
the absolute level of both basal and forskolin-stimulated
transcriptional activity (Fig. 5C). However, as in the constructs
containing larger mutations (Fig. 5, A and B), elimination of both CREs
had no effect on the relative magnitude of forskolin stimulation
(
7-fold) and dexamethasone suppression (50%), both effects being
similar to those seen in the wild-type construct, AVP441. Modification
of the CAAT box had no effect on either the absolute or relative
magnitude of vasopressin promoter response to forskolin or
dexamethasone. Taken together, the results in Fig. 5 suggest that the
two CRE-like sequences are important for both basal vasopressin gene
transcription and the absolute level of response to forskolin, although
the relative modulation of transcriptional activity by cAMP or
glucocorticoid is not dependent on the presence of these DNA
sequences.
Effects of cotransfection of KCREB, a dominant-negative inhibitor
of CREB, or an AP2 competitor plasmid, on vasopressin promoter
activity
To examine the role of cAMP-responsive element binding protein
(CREB) in mediating forskolin-stimulated vasopressin gene expression,
we cotransfected JEG-3 cells with AVP441-luciferase construct and with
an expression vector encoding either CREB or KCREB (40), the latter a
dominant-negative form of CREB (Fig. 6A).
When KCREB was coexpressed with AVP441 (but not with AVP102, which
lacks both putative CREs), both absolute basal and
forskolin-stimulated vasopressin promoter activities were
significantly lower (less than 50%) than seen with CREB, although the
relative level of forskolin stimulation (7-fold) was not affected by
KCREB expression. That KCREB had no effect on basal or
forskolin-stimulated transcription of construct AVP102 suggests that
forskolin-mediated transcriptional activation of this construct may be
mediated by a mechanism not involving CREB. We also cotransfected JEG-3
cells with AVP441-luciferase construct and with either AP2 competitor
or control plasmids. Cotransfection of the competitor plasmid
significantly decreased the both basal and forskolin-stimulated
vasopressin expression (Fig. 6B) compared with that of the control
plasmid. The relative level of forskolin stimulation, however, was not
affected.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) and had no effect on the magnitude of PKA
stimulation, which remained consistently approximately 7-fold above
basal promoter activity. We also found that the negative effect of
glucocorticoid is GR dependent and may be mediated by a mechanism not
involving GR binding to DNA. Moreover, when the zinc finger domain was
removed, GR activity remained ligand dependent but was switched from a
repressor to an activator of vasopressin transcription. Finally, we
found no evidence for PKC activation of vasopressin gene
transcription. Our results show that, as previously reported (20, 21, 41, 42), forskolin, an activator of adenylate cyclase, or 8Br-cAMP, a membrane-permeable cAMP analog, potently stimulate vasopressin gene expression. A marked dose-dependent stimulatory effect was also observed with cotransfection of the catalytic subunit of a PKA expression vector (33), further confirming the positive role of this pathway. In vivo studies in the rat have demonstrated accumulation of cAMP within the SON after hypertonic saline injection or salt loading (18, 19), suggesting that this regulatory system may operate in intact organisms.
We further examined the cis- and trans-acting
element(s) that may mediate the positive effect of the cAMP/PKA
pathway. Because the rat 5' vasopressin promoter sequence has two
putative CREs, a putative nurr1/nur77 element and four putative AP2
consensus sequences within 0.5 kb of the promoter (Fig. 4), any of
which might confer cAMP responsiveness (21, 36, 39), we selectively
mutated each of these DNA regions to ascertain their involvement in the
positive regulation by forskolin. Elimination of DNA sequences between
-3000 and -803 caused a consistent increase in basal and stimulated
promoter activity, suggesting repressor properties in upstream
element(s) in this region. Elimination of a consensus nurr1/nur77 site
had no effect on positive or negative regulation of the vasopressin
promoter. This element, present in two other neuroendocrine genes of
the hypothalamic-pitutiary-adrenal axis that encode POMC and CRH, has
been shown to bind the nurr1/nur77 subfamily of nuclear hormone
receptors and mediate both positive and negative regulation by the PKA
and GR pathways, respectively (39). Deletion analysis showed that the
elimination of either one of the two putative CREs within the
vasopressin promoter had no effect, whereas removal of both CREs
markedly diminished transcriptional activity, suggesting that both
basal and forskolin-stimulated transcription is mediated by multiple
CREs. This redundant mode of regulation has been described in the
transcriptional control of promoters of various neuropeptide genes
(43). Pardy et al. (21) recently identified a CRE between
-210 and -112 of the bovine vasopressin promoter [corresponding to
CRE2 (-123 to -116) of the rat vasopressin promoter] using a
relatively short construct that does not contain the CRE sequence
corresponding to rat CRE1.
5' deletion constructs with less than 266 bp of promoter (AVP211, AVP102, and AVP36), internal deletion constructs lacking CRE1 (AVP441D250–201), or CRE2 (AVP441D150–101), as well as the double point deletion construct lacking both CRE1 and CRE2 (AVP441DCRE1/DCRE2) all share an intriguing characteristic: whereas the absolute level of basal and forskolin-stimulated transcriptional activity is regulated in all of these constructs, the relative increase in vasopressin promoter transcriptional activity following cAMP stimulation remains intact. The decrease in absolute basal activity may reflect a decreased ability of these mutated vasopressin promoter constructs to bind a cAMP-stimulated transcription factor such as CREB, which can support constitutive gene transcription in a phosphorylation-independent fashion via interaction of its Q2 domain with dTAF110, a component of the basal transcription factor TFIID (44). The maintenance of the relative magnitude of transcriptional stimulation following PKA activation suggests that the degree of activation involves a process that is independent of the magnitude of promoter binding of the transcription factor, such as the phosphorylation of a transcription factor already bound to the promoter DNA sequence. Phosphorylation of CREB promotes the interaction of its KID domain with the KIX domain of CREB binding protein (CBP) to stimulate transcription (45). Thus, a vasopressin promoter construct containing CRE mutations might bind CREB at a level 25% of that of an intact promoter and exhibit a basal transcription rate 25% that of the intact promoter. CREB phosphorylation following PKA stimulation might activate promoter-bound CREB to the same degree in both intact and mutated promoter constructs, resulting in the same fold stimulation, although to different absolute levels, of transcriptional activity. This is likely a general phenomonen that was not generally appreciated previously because of the relative insensitivity of past methods used to monitor transcriptional activity, in which mutations (e.g. within a CRE) affecting PKA-stimulated transcription were reported to alter the degree of transcriptional activation by PKA (46).
To further define the nature of PKA-responsive factors involved in vasopressin promoter regulation, we examined the effect of KCREB, a dominant negative mutant of CREB (40), upon vasopressin promoter transcription. Coexpression of KCREB clearly attenuated the basal and absolute level of forskolin-stimulated expression of AVP441, a vasopressin promoter containing two putative CREs, but again did not affect the relative increase in transcriptional activity following forskolin addition. One explanation for these results is that KCREB blocks binding of CREB to CREs at -220 and -120 bp in AVP441, resulting in reduced basal and stimulated transcriptional activity. That the residual forskolin-stimulated activity following coexpression of AVP441 and KCREB may be due to binding of a PKA-stimulated factor (either CREB or another DNA binding protein such as AP2) to a non-CRE sequence is supported by the absence of effect of KCREB on PKA-stimulated transcriptional activation of AVP102, which lacks CRE motifs. The role of AP2 was tested as well, because this factor is also known to mediate cAMP-stimulated transcriptional activation (47). In an attempt to compete away any endogenous AP2-like binding factors, we cotransfected a plasmid harboring the same AP2-like sequences observed in -72 to -100 of the vasopressin promoter. This procedure also decreased vasopressin promoter activity, suggesting that AP2, or related factors that bind to these sequences, may play a positive role in vasopressin gene expression.
We found that glucocorticoid inhibited PKA-stimulated, but not basal
transcription, that this effect was GR-dependent, and that the zinc
finger domain, but not the N-terminal transactivation domain, was
crucial for this suppressive effect. A similar effect was observed
previously in the regulation of the human interleukin 6 promoter (48).
Although the zinc finger domain of GR is known to be necessary for DNA
binding to GREs that stimulate gene transcription, the repressive
effect of dexamethasone was observed in vasopressin promoter constructs
which have no obvious GRE, suggesting that this effect may be caused
not by DNA binding of GR, but perhaps by protein-protein sequestration
(squelching) of a positive transcription factor by GR. Although such
squelching could be an artifact caused by the transfection of excessive
amounts of GR, this possibility was minimized by titration of the
amount of GR transfected to the lowest amount (0.3 µg/transfection)
at which a dexamethasone-dependent effect was observed (Fig. 2B).
Furthermore, we have observed qualitatively similar
dexamethasone-dependent negative regulation of these same vasopressin
gene promoter constructs following their introduction into AtT-20
cells, which express endogenous GR (Iwasaki, Y., unpublished
observations). Recently, activation of transcription by GR has been
shown to require CBP (49), and it has been suggested that intracellular
levels of CBP may be limiting (50). Thus, an explanation for our
findings may be that glucocorticoid-activated GR within the nucleus
binds to CBP, thereby decreasing CREB-CBP interaction, resulting in a
decrease in PKA-stimulated transcription. By this mechanism, GR would
not inhibit basal transcription, because nonphosphorylated CREB
stimulates constitutive transcription via direct interaction with
TFIID, without the involvement of CBP (44). That GR-mediated inhibition
of vasopressin promoter transcription was abolished in a GR mutant
lacking the zinc finger domain raises the possibility that this motif
may mediate interaction between GR and CBP. The switch from negative to
positive regulation associated with this GR mutant suggests that GR may
possess both activities, either of which predominates when the other is
blocked. This model is consistent with recent studies that suggest that
GR can modulate transcriptional regulation without DNA binding, and
that the zinc finger domain of GR may play a role not only in DNA
binding but also in homodimerization and interaction with other
proteins (51, 52, 53). This model is also consistent with our finding that
glucocorticoid reduced PKA-activated transcription to a similar degree
in vasopressin promoter constructs AVP3000 to AVP102, irrespective of
the absolute level of activation by PKA (Fig. 5). However, we cannot
explain why dexamethasone did not suppress forskolin activation of
construct AVP36, although in this case DNA within 7 nucleotides of the
TATA box is deleted, which might normally bind components of the basal
transcription complex involved in glucocorticoid-mediated negative
regulation.
Our studies suggest that the positive and negative regulation of vasopressin gene transcription is complex. They should guide future experiments that will more directly examine the physical interaction of transcription factors with the vasopressin gene promoter, to more precisely characterize the nature of this regulation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received May 7, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  Y. Iwasaki, M. Nishiyama, T. Taguchi, M. Kambayashi, M. Asai, M. Yoshida, T. Nigawara, and K. Hashimoto Activation of AMP-activated protein kinase stimulates proopiomelanocortin gene transcription in AtT20 corticotroph cells Am J Physiol Endocrinol Metab, June 1, 2007; 292(6): E1899 - E1905. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Yoshida, Y. Iwasaki, M. Asai, S. Takayasu, T. Taguchi, K. Itoi, K. Hashimoto, and Y. Oiso Identification of a Functional AP1 Element in the Rat Vasopressin Gene Promoter Endocrinology, June 1, 2006; 147(6): 2850 - 2863. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Nigawara, Y. Iwasaki, M. Asai, M. Yoshida, M. Kambayashi, H. Sashinami, K. Hashimoto, and T. Suda Inhibition of 11{beta}-Hydroxysteroid Dehydrogenase Eliminates Impaired Glucocorticoid Suppression and Induces Apoptosis in Corticotroph Tumor Cells Endocrinology, February 1, 2006; 147(2): 769 - 772. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. C. Meijer, B. Topic, P. J. Steenbergen, G. Jocham, J. P. Huston, and M. S. Oitzl Correlations between Hypothalamus-Pituitary-Adrenal Axis Parameters Depend on Age and Learning Capacity Endocrinology, March 1, 2005; 146(3): 1372 - 1381. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E Yamamori, M Asai, M Yoshida, K Takano, K Itoi, Y Oiso, and Y Iwasaki Calcium/calmodulin kinase IV pathway is involved in the transcriptional regulation of the corticotropin-releasing hormone gene promoter in neuronal cells J. Mol. Endocrinol., December 1, 2004; 33(3): 639 - 649. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Asai, Y. Iwasaki, M. Yoshida, N. Mutsuga-Nakayama, H. Arima, M. Ito, K. Takano, and Y. Oiso Nuclear Factor of Activated T Cells (NFAT) Is Involved in the Depolarization-Induced Activation of Growth Hormone-Releasing Hormone Gene Transcription in Vitro Mol. Endocrinol., December 1, 2004; 18(12): 3011 - 3019. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Murgatroyd, A. Wigger, E. Frank, N. Singewald, M. Bunck, F. Holsboer, R. Landgraf, and D. Spengler Impaired Repression at a Vasopressin Promoter Polymorphism Underlies Overexpression of Vasopressin in a Rat Model of Trait Anxiety J. Neurosci., September 1, 2004; 24(35): 7762 - 7770. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Asaba, Y. Iwasaki, M. Yoshida, M. Asai, Y. Oiso, T. Murohara, and K. Hashimoto Attenuation by Reactive Oxygen Species of Glucocorticoid Suppression on Proopiomelanocortin Gene Expression in Pituitary Corticotroph Cells Endocrinology, January 1, 2004; 145(1): 39 - 42. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Yamamori, Y. Iwasaki, Y. Oki, M. Yoshida, M. Asai, M. Kambayashii, Y. Oiso, and N. Nakashima Possible Involvement of Ryanodine Receptor-Mediated Intracellular Calcium Release in the Effect of Corticotropin-Releasing Factor on Adrenocorticotropin Secretion Endocrinology, January 1, 2004; 145(1): 36 - 38. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Yoshida, Y. Iwasaki, M. Asai, T. Nigawara, and Y. Oiso Gene Therapy for Central Diabetes Insipidus: Effective Antidiuresis by Muscle-Targeted Gene Transfer Endocrinology, January 1, 2004; 145(1): 261 - 268. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kuwahara, H. Arima, R. Banno, I. Sato, N. Kondo, and Y. Oiso Regulation of Vasopressin Gene Expression by cAMP and Glucocorticoids in Parvocellular Neurons of the Paraventricular Nucleus in Rat Hypothalamic Organotypic Cultures J. Neurosci., November 12, 2003; 23(32): 10231 - 10237. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Mune, H. Murase, N. Yamakita, T. Fukuda, M. Murayama, A. Miura, T. Suwa, J. Hanafusa, H. Daido, H. Morita, et al. Eutopic Overexpression of Vasopressin V1a Receptor in Adrenocorticotropin-Independent Macronodular Adrenal Hyperplasia J. Clin. Endocrinol. Metab., December 1, 2002; 87(12): 5706 - 5713. [Abstract] [Full Text] [PDF]
|
77–262, a mutant GR in which N-terminal
transactivation domain was deleted; 
