| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pathology and Program in Molecular Biology University of Colorado Health Sciences Center Denver, Colorado 80262
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
For the sex steroid receptors and glucocorticoid receptor (GR),
dimerization appears to be a critical factor controlling receptor
binding to palindromic HREs (8, 9). Steroid receptors have been shown
to be able to form dimers in solution, and as full-length proteins they
bind preferentially to HREs as preformed dimers. In general, steroid
receptors interact with target palindromic HREs as a single type of
homodimer. Exceptions to this are hPR, which binds to target DNA as
three dimeric species composed of AA, AB, and BB subunits (10, 11, 12, 13) and
the more recent demonstration that GR can form a heterodimer complex on
DNA with mineralocorticoid receptor (MR) (14, 15). In addition, the
newly discovered estrogen receptor-ß (ERß) can heterodimerize with
ER
(16). In contrast, receptors for nonsteroid hormones, such as
vitamin D, thyroid hormone, and retinoids, preferentially bind DNA as
heterodimers with retinoic acid X receptor (17).
Two dimerization domains have been described within steroid receptors. A DNA-dependent dimerization interface located within the DBD is important for restricting steroid receptor recognition of palindromic HREs separated by three intervening nucleotides and for stabilization of receptor dimers on DNA (18). A second dimerization function resides in a region(s) outside the DBD that is involved in solution dimerization in the absence of DNA (18, 19, 20). The LBD of ER contains a strong hormone-dependent dimerization function that is required for solution dimerization of ER (8, 21, 22). Other studies with GR, androgen receptor, and PR have suggested that the LBD alone is not sufficient for stable solution dimerization and that amino-terminal sequences are also involved (23, 24, 25, 26, 27). Recently, we have analyzed the sequence regions involved in solution dimerization and have found that both the hinge and amino-terminal sequences contribute to PR dimerization (28).
Several synthetic ligands have been developed that compete effectively for binding of progestins to PR and are capable of inhibiting receptor activation. The mechanism by which antiprogestins inhibit activation remains incompletely understood. Original studies based on in vitro electrophoretic gel mobility shift assays (EMSA) classified ZK98299 (Onapristone) as a type I compound that failed to promote PR binding to DNA. Other antiprogestins such as RU486 (Mifepristone) were classified as type II compounds that efficiently enhance PR binding to DNA (13) and thus impair receptor transactivation at step(s) downstream of DNA binding. However, assays based on detection of PR-DNA binding in whole cells (29, 30) and our more recent results with altered EMSA conditions (30) indicate that ZK98299 does enhance PR binding to DNA. It has been well documented that antiprogestins induce conformational changes within the LBD of PR that are distinct from that induced by agonists (11, 12, 31, 32, 33, 34). Similar results have been shown with ER and estrogen antagonists (35). One consequence of the different conformation is to prevent the interaction of specific steroid receptor coactivators with the hormone agonist-dependent transcription activation function 2 (AF-2) (5, 6, 7). ZK98299 and RU486 have been shown to have different effects on PR conformation, suggesting that they may represent different classes of progestin antagonists based on inducing distinct conformational changes within PR (30, 36, 37).
The biological activity of antiprogestins as assessed by transfection experiments indicates that these compounds inhibit transcriptional activation of PR at substoichiometric concentrations with hormone. This suggests that antiprogestins do not act by simple competition with hormone for binding to PR (38). This unusually high potency of antiprogestins could, in part, be explained by the PR-antiprogestin complex functioning to repress the PR-agonist complex in trans. Since PR requires dimerization for function, this could be achieved by heterodimerization between a PR bound to agonist with a PR bound to antagonist.
Numerous studies have recently taken advantage of the two-hybrid
protein-protein interaction assay to examine dimerization of eukaryotic
transcription factors (39). Originally developed in yeast (40) and
later adapted to mammalian cells (41), the assay involves coexpression
of two proteins of interest fused to an autonomous DBD (from the GAL4
or LEXA transcription factors) and a strong transcriptional activation
domain (from GAL4 or VP16). Stable protein-protein interactions can
functionally reconstitute the separate DBD and activation domains of
the transcription factor, resulting in activation of a reporter gene
bearing DNA-binding sites for the DBD fusion protein. Previously, a
yeast two-hybrid assay was used to show ligand-inducible dimerization
of ER in vivo (42) and, more recently, a mammalian based
two-hybrid assay was used to demonstrate estrogen-dependent
heterodimerization between ERß and ER in whole cells (16). In the
present study, we used a mammalian two-hybrid assay to investigate
whether the three detected PR dimer forms complexed to DNA
in vitro can be formed in whole cells and whether or not
this is ligand dependent. In addition, we have used this system to
examine the effect of progesterone antagonists on receptor dimerization
and whether PR bound to agonist can heterodimerize in the cell with PR
bound to antagonist.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). The chimeric
constructs were transiently transfected into PR- negative COS-1 cells
along with a reporter plasmid, pG5CAT, that contains five GAL4
consensus binding sites and the minimal E1b promoter driving the
chloramphenicol acetyl transferase (CAT) gene. Interaction between two
PR molecules was assessed by the ability of cells transfected with
GAL4/PR and VP16/PR constructs to activate the pG5CAT reporter gene as
compared with cells transfected with either PR chimera alone.
|
). In addition,
each of the PR fusion proteins was expressed at approximately the same
level as native PR in T47D cells (Fig. 1B). It should be noted that
even though the VP16 domain is of smaller mol wt than the GAL4 DBD, the
VP16 fusion proteins exhibit slower mobility than the GAL4 fusion
proteins. A similar phenomenon was reported for androgen receptor
two-hybrid proteins suggesting that the highly acidic nature of VP16
causes anomalous electrophoretic mobility (43). The immunoreactive
doublets observed in the PR-A and PR-B GAL4 lysates may be a result of
proteolytic cleavage of the fusion proteins; however, these lower mol
wt species do not seem to interfere with our ability to detect
interactions with the VP16 fusion proteins (see below). The expressed
PR fusion proteins were also characterized for steroid binding activity
by a whole cell-binding assay. The results summarized in Table 1 show that all PR fusion proteins
exhibit similar specific binding to the synthetic progestin
[3H]R5020 with equal capacity, again approaching the
level of native PR in T47D cells. Thus, fusion of PR to either the GAL4
DBD or VP16 did not adversely affect the expression or the steroid
binding activity of PR.
|
When COS-1 cells were cotransfected with either, or both, parental
vectors that express GAL4 DBD or VP16 without PR sequences, activation
of the GAL4- controlled pG5CAT reporter gene was negligible (Fig. 2). Also, little or no activation of
pG5CAT was observed when cells were cotransfected with a single
VP16/PR-B chimeric vector cotransfected with a GAL4 DBD empty vector
control (Fig. 2). The low CAT activity observed under these conditions
was considered as the basal activity of the pG5CAT promoter in COS-1
cells. However, the GAL4/PR-B chimeric construct as a single hybrid
(cotransfected with VP16 empty control vector) did activate pG5CAT 2-
to 4-fold above the basal level in response to R5020 (Fig. 2).
Activation of the pG5CAT reporter by the single GAL4/PR-B chimeric
receptor is not entirely unexpected. Full-length PR-B harbors multiple
transactivation domains (49, 50) and has the potential, through direct
utilization of these domains, to mediate transcription when
tethered to the pG5CAT through its GAL4 DBD fusion sequences. In
contrast, PR-A, which is generally a weaker transcriptional activator
than PR-B (51, 52), mediates no induction of the pG5CAT reporter as a
single hybrid in response to hormone (not shown and Fig. 3). The data in Fig. 2 illustrate the
importance of factoring CAT expression mediated by the single GAL4/PR
hybrid to reliably detect induction of PR dimerization as reflected by
interaction of the two PR fusion proteins. In later two-hybrid
experiments, the GAL4/PR-B chimeric construct as a single hybrid
mediated a higher activation of the pG5CAT reporter in response to
R5020 (see below) than shown in Fig. 2, further emphasizing the need to
factor in the contribution of this single hybrid when calculating CAT
expression that represents PR dimerization.
|
|
). As shown in Fig. 3, hormone (R5020)
treatment of cells transfected with the GAL4/PR-A single hybrid did not
stimulate CAT activity above the ethanol control, whereas R5020
stimulated a significant (6-fold) increase of CAT expression in cells
transfected with the GAL4/PR-B single hybrid. When cells were
cotransfected with both GAL4/PR and VP16/PR fusion proteins, no
significant stimulation of CAT activity above the single hybrid
controls was obtained in the absence of ligand. Hormone (R5020)
addition to cells cotransfected with the different two hybrid
combinations in each case stimulated a significant increase of CAT
activity as compared with ethanol controls and, more importantly,
hormone also stimulated a significant increase of CAT activity over the
single hybrid controls treated with R5020 (Fig. 3). The two-hybrid
combinations included GAL4/PR-A plus VP16/PR-A (A-A interaction),
GAL4/PR-A plus VP16/PR-B (A-B interaction), and GAL4/PR-B plus
VP16/PR-B (B-B interaction). To determine the fold hormone stimulation
due to two-hybrid protein interaction (i.e. dimerization),
CAT activity contributed by the single GAL4/PR hybrid in the presence
of R5020 was factored out by calculating the relative CAT activity
obtained from the two-hybrid plus hormone divided by the relative CAT
activity obtained by the single hybrid plus hormone (see
Materials and Methods). By this analysis, hormone stimulated
a 3.44-fold increase in CAT activity mediated by PR-A/PR-A interaction,
a 7.16-fold increase mediated by PR-A/PR-B interaction, and 3.2-fold
increase mediated by PR-B-/PR-B interactions. Hormone-dependent
interaction between the A and B form of PR was also observed when PR-B
was fused to GAL4 DBD and PR-A was fused to VP16. However, in this
orientation the fold hormone (3-fold) stimulation was less than that
obtained with the fusion proteins in the opposite orientation in Fig. 3
(data not shown). This result, taken together with the fact that PR-B
has a higher intrinsic activity than PR-A, makes it difficult to
interpret whether the higher 7-fold stimulation mediated by GAL4/PR-A
and VP16/PR-B in Fig. 3 actually reflects a stronger dimerization
between PR-A and PR-B as compared with A-A and B-B interactions.
Nonetheless, the results in Fig. 3 strongly support the conclusion that
formation of the three possible forms of PR dimers (AA, AB, and BB)
occurs in the cell in a hormone-dependent manner.
Antagonist-Stimulated PR Dimerization in Whole Cells
Using the same PR chimeric constructs, we next asked how different
progesterone antagonists would effect PR-PR interactions in the intact
cell. The antagonists included RU486 and closely related ZK112993 (type
II), which behave as strong antagonists but display partial agonist
activity under certain conditions, and the more complete antagonist
ZK98299 (type I). None of these compounds stimulated pG5CAT activity
above basal levels in cells transfected with both GAL4/PR-B and
VP16/PR-B chimeras (data not shown). This was surprising since these
same antagonists have been shown to promote efficient binding of PR to
target DNA in whole cells (29, 30). We therefore considered the
possibility that antagonists may reduce the transcriptional activity of
the VP16 fusion proteins in such a manner that interaction between
VP16/PR and GAL4/PR chimeras is undetected, due to the failure of the
VP16 fusion protein recruited to the promoter to mediate a strong
transcriptional response. To test this idea, we analyzed the ability of
the VP16/PR-B chimera to function as a positive transactivator of a
PRE-controlled reporter gene (DHRE-E1b-CAT) in response to agonists and
antagonists. As shown in Fig. 4, R5020
stimulated a substantial 125-fold induction of DHRE-E1b-CAT reporter
activity in cells transfected with VP16/PR-B. VP16/PR-B transactivation
of the DHRE-E1b-CAT gene was also stimulated by all antagonists
examined, including ZK98299. However, the fold induction was only
4–13% (range is from three independent experiments) of that mediated
by hormone agonist (Fig. 4). The reduced ability of antagonists to
stimulate CAT activity does not appear to be due to promoting a less
efficient binding of the VP16/PR fusion protein to DNA than hormone
agonist. We have performed a control experiment by cotransfecting
VP16/PR-B with a carboxyl-terminal truncation UP-1 mutant of PR that
binds and responds to RU486 as an agonist (Ref. 33 and see below).
Cotransfected VP16/PR-B inhibited the transcriptional activity of
VP16/PR-B(UP-1) in a dose-dependent manner in cells treated with RU486
(not shown). This inhibitory effect suggests that the VP16/PR-B
chimeric receptor dimerizes and efficiently binds to DNA in the cell in
response to RU486. In further support of antagonists reducing the
transcriptional activity of the VP16/PR-B fusion protein as opposed to
reducing its binding efficiency for DNA, we recently showed by promoter
interference assays that all progestin antagonists used in this study
stimulate binding of PR to PREs in vivo (whole mammalian
cells) with the same efficiency as the agonist R5020 (30). How
antagonists reduce the transcriptional activity of VP16/PR-B fusion
protein is not known. A possible mechanism is suggested by recent
studies showing that PR and ER bound to antagonists strongly associates
with corepressors that can functionally silence partial agonist effects
of antagonists (53, 54, 55).
|
). In contrast, when COS-1
cells were transfected with VP16/PR-B(UP-1), all antagonists tested
stimulated a substantial induction of DHRE-E1b-CAT. The fold induction
stimulated by RU486 and ZK112933 was equal to that of R5020 induction
mediated by the VP16/PR-B construct (Fig. 4). VP16/PR-B(UP-1)
transactivation induced by ZK98299 was approximately 50% lower than
other antagonists but was still substantially greater than the level of
transactivation mediated by wild-type PR-B fused to VP16 in response to
ZK98299 (Fig. 4).
We next analyzed the effect of antagonists on interaction between the
two PR-UP-1 chimeras. As expected R5020, which does not bind the UP-1
PR mutant, did not stimulate CAT activity above the ethanol controls in
cells transfected with either the single hybrid GAL4/PR-B(UP-1) control
or with both GAL4 and VP16 fusions of PR-B(UP-1) (Fig. 5). In contrast, the antagonists RU486
and ZK98299 each stimulated a significant increase (3.2-fold and
2.8-fold, respectively) of CAT activity in cells cotransfected with
both fusion proteins over the single hybrid GAL4/PR-B(UP-1) control
treated with the same antagonist (Fig. 5). Although ZK112993 also
stimulated CAT activity (2.6-fold) in cells transfected with both
hybrid PR-B(UP-1) proteins over the single-hybrid control (Fig. 5), the
increase was statistically significant only at a P value of
0.08 suggesting a trend toward stimulating PR-PR interaction.
|
Heterodimerization between PR Bound to Agonist and PR Bound to
Antagonist in Whole Cells
An interesting and unexplained property of these progestin
antagonists is their ability to inhibit activation of PR at
concentrations well below that of the hormone, indicating that
antagonists do not act by simple competition with hormone for PR
binding sites (11, 38). Despite the fact that RU486, ZK112993, and
ZK98734 all have the same binding affinity for PR as the agonist R5020
(Ref. 29 and unpublished observations), these compounds inhibit
activation of PR at substoichiometric concentrations with R5020. This
is illustrated in Fig. 6 with ZK112993
and ZK98299 in a breast cancer cell line containing a stably integrated
mouse mammary tumor virus (MMTV)-CAT, a progestin-inducible reporter
gene (11). Cells were treated with a saturating amount of R5020 (5
nM) or were treated with 5 nM R5020 and
increasing amounts of ZK112993 (Fig. 6A) or ZK98299 (Fig. 6B). Reporter
gene activity stimulated by 5 nM R5020 was inhibited by
50% at a 10-fold lower concentration of ZK112993 (average
IC50 = 0.4 nM, n = 5), while maximal
inhibition was obtained at approximately 3 nM ZK112993
(Fig. 6A). The IC50 for the antagonist ZK98299 was between
3–10 nM (average 7 nM, n = 5), and
maximal inhibition was obtained between 100 and 500 nM of
ZK98299 (Fig. 6B). When the 10-fold lower affinity for PR as compared
with R5020 is corrected for, ZK98299 also antagonizes at
substoichiometric concentrations with R5020. Thus, both antagonists
have an observed IC50 that is approximately 10-fold lower
than the expected.
|
Cys) of hPR, which abolishes binding
of RU486 but retains normal progestin agonist binding affinity and
functional response to progestins (56). In the context of
full-length PR-B, we inserted the 722 (GlyCys) mutant into the GAL4
or VP16 fusion vectors and used these fusion proteins partnered with
PR-B(UP-1) fusion proteins to create a mammalian cell two-hybrid system
that expresses receptor hybrids with the desired agonist/antagonist
ligand specificities (see Fig. 7A). To
confirm that these mutant PRs as fusion proteins possess the ligand
specificities expected, a steroid binding assay in whole cells was
performed with [3H]R5020 and [3H]RU486 in
COS-1 cells transfected with each receptor GAL4 and VP16 chimeric
construct. The VP16/PR-B(722m)-fusion protein bound
[3H]R5020 specifically and exhibited no specific binding
to [3H]RU486. Conversely, the GAL4/PR-B(UP-1) fusion
protein bound [3H]RU486 in whole cells but not
[3H]R5020. As expected, native PR in T47D cells as a
control bound R5020 and RU486 equally (Fig. 7A). It should be noted
that whole-cell binding assays used a 50 nM concentration
of R5020 and RU486, which is approximately equal to the
concentrations used in two-hybrid assays (see Fig. 7B). Even at
these high concentrations, the mutant PRs show complete agonist
(PR722m) and antagonist (PR(UP-1)) specificity. The whole-cell steroid
binding assays also show (Fig. 7A) that PR(UP-1) and PR722m fusion
proteins are expressed at approximately equal levels in transfected
COS-1 cells, which was also confirmed by immunoblot analysis of the
cell lysates (not shown).
|
). In COS-1 cells cotransfected with both GAL4/PR-B(UP-1) and
VP16/PR-B(722m) hybrid constructs, no further increase in pG5CAT
activity over the single hybrid control (GAL4/PR-B(UP-1) + VP16) was
observed by addition of RU486 alone. Addition of R5020 alone also did
not stimulate a significant increase in CAT activity over the single
hybrid also treated with R5020. A significant 4.2-fold stimulation of
CAT activity over the single hybrid control was observed only when
cells were cotreated with R5020 and RU486 (Fig. 7B). Thus, both agonist
and antagonist were required for interaction between GAL4/PR-B(UP-1)
and VP16/PR-B (722m). These results provide strong evidence that PR
bound to an antagonist is capable of heterodimerization in the cell
with PR bound to antagonist. These results also underscore that
dimerization in whole cells requires ligand binding to both PR
partners.
Mixed Agonist/Antagonist PR Heterodimers in Vitro
Exhibit Reduced Ability to Bind to DNA
The two-hybrid assay results indicate that PR bound to an
antagonist can heterodimerize with PR bound to an agonist but does not
determine whether these mixed ligand heterodimers are functional. As
one means to examine this question, we have performed EMSAs of
separately expressed A and B receptors bound to either agonist or
antagonist. Because of the molecular size difference of the A (90 kDa)
and B receptors (120 kDa), mixing of the two PR isoforms results in the
formation of an intermediate mobility PR-A/PR-B heterodimer complex
that we and others have shown previously to be easily discernible by
EMSA from the faster mobility AA and slower mobility BB dimers (11, 12, 13, 38, 44, 45, 46). The A and B receptors were expressed in the baculovirus
insect cell system as previously described, and ligands were added to
Sf9 insect cell cultures for the last 6–8 h of expression to allow
binding to PR in the cell before extraction of receptor from nuclei
with 0.4 M NaCl. Free ligand was removed by treatment of
Sf9 cell extracts with dextran-coated charcoal as previously
described, and assays were performed at 0–4 C under conditions that
keep ligand dissociation to a minimum (45). When a constant amount of
PR-B bound to R5020 was mixed with increasing amounts of PR-A bound to
R5020 in a DNA binding reaction with a 32P-labeled PRE
oligonucleotide probe, a single PR-B-PRE complex was detected at the
lowest amount of PR-A, while two additional mobility complexes were
detected at higher amounts of PR-A (Fig. 8A, lanes 1–6). As shown previously by
supershifts with appropriate PR antibodies (11, 12, 44, 45), the
intermediate mobility complex is composed of PR-A-PR-B heterodimers,
while the fastest mobility complex contains PR-A homodimers. In
contrast, little or no intermediate mobility complexes were detected
when PR-B bound to R5020 was mixed with increasing amounts of PR-A
bound to the antagonist ZK112993 (Fig. 8A, lanes 7–12) or with
increasing amounts of PR-A bound to ZK98734 (Fig. 8A, lanes 13–18). We
and others previously observed similar results when one PR isoform
bound to RU486 was mixed with another bound to R5020. The amount of
intermediate mobility heterodimer complex was minimal compared with the
intermediate mobility heterocomplexes obtained when PR-A and PR-B were
bound to the same ligand (either R5020 or RU486) (12, 45). It should be
noted that the amount of PR-B (R5020) DNA complex decreased as
increasing amounts of PR-A antagonist were added to the EMSA assays,
albeit the decrease is more predominant with PR-A-ZK98734 than with
PR-A-ZK112993 (Fig. 8A). This decrease of the PR-B complex at the
expense of the PR-A complex, without the appearance of a third
intermediate mobility complex [which does occur when mixing PR isoform
bound to the same ligand (R5020)], is consistent with the formation of
mixed ligand heterodimers in solution that exhibit reduced DNA binding
affinity.
|
, PR-A was efficiently coimmunoprecipitated with the
PR-B-R5020 complex whether it was bound to R5020 or any of the
antagonists (lanes 1–5). Specificity of PR-A interaction with PR-B is
shown by the lack of PR-A coimmunoprecipitation with the B-30 MAb in
the absence of added PR-B (Fig. 8B, lanes 6–10). These results suggest
that PR bound to hormone agonist is capable of forming a solution
heterodimer with PR bound to antagonist in the absence of DNA. These
in vitro results, taken together with the results of the
two-hybrid assay with the ligand specificity mutant PRs, strongly
support the conclusion that PR bound to antagonist is capable of
forming a heterodimer with PR bound to agonist and that these
heterodimers have impaired ability to bind to target DNA sequences. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The two-hybrid system has been used previously to examine dimerization
of steroid receptors. Using a yeast two-hybrid assay, Wang et
al. (42) showed that dimerization of human ER in
vivo is estrogen dependent. A mammalian-based two-hybrid system
was used to show ligand-dependent interactions between the - and
ß-subtypes of mouse ER (16). Similar to our results with PR in a
mammalian-based system, a estrogen-dependent rise in reporter gene
activity was observed with a single GAL4/mER hybrid (and VP16 empty
vector). This activity was attributed to the ligand-inducible
transactivation function of ER itself, when directed to the promoter
of the GAL4 reporter construct by binding via the GAL4 DBD (16).
Reporter gene activity induced by the single GAL4/ER hybrid was
approximately 5-fold, which is similar to the induction level observed
to be mediated by the GAL4/PR-B single hybrid treated with hormone
(6-fold) (Fig. 3). Interestingly, we did not observe an R5020 induction
of reporter gene activity (pG5CAT) with the A isoform as a single GAL4
fusion protein. Both the A isoform of PR and ER contain two
transactivation functions, whereas the B isoform of PR contains an
additional third activation domain (49, 50). It would, therefore,
appear that the two activation functions (AF-1 and AF-2) of ER are
stronger than those of PR, at least in the context of the two-hybrid
assay. In our initial two-hybrid experiments, we attempted to reduce
the assay background activity of the single GAL4/PR-B hybrid protein by
taking advantage of known mutations in AF-2 of PR. Replacement of
conserved glutamate residues at positions 907 and 911 with alanine
results in a mutant receptor that is able to bind progesterone but has
greatly reduced AF-2 hormone-dependent activity (51). By using
PR-B(907, 911m) to construct the GAL4 DBD fusion, we observed a
reduction in hormone-dependent activity of the single hybrid protein.
However, the fold hormone induction of pG5CAT dependent on coexpression
of the second VP16/PR-B hybrid protein was no greater than that
obtained with wild-type PR-B as the GAL4 DBD fusion construct (not
shown). In the study of ER dimerization by the yeast two-hybrid assay,
no assay background activity was detected with the single GAL4/hER
hybrid (42). This clearly contrasts with our data with PR and with the
reported assay background detected with the GAL4 DBD/mER as a single
hybrid protein in mammalian-based two-hybrid assays (16). This apparent
discrepancy between yeast- and mammalian-based two-hybrid assays is
likely due to receptors in yeast possessing lower intrinsic
transcriptional activity due to the absence of steroid receptor
coactivators that are normally present in mammalian cells (6, 7, 57).
Our initial two-hybrid experiments using wild-type PR-B to analyze the
effects of antagonists on PR dimerization suggested that these
compounds do not promote dimerization in vivo. Recent
results showing that PR and ER bound to antagonists inappropriately
forms a strong association with nuclear receptor corepressors (NCOR or
SMRT) provides a likely explanation for the lack of antagonist-induced
two-hybrid dependent activation of the pG5CAT reporter (53, 54, 55). In the
context of the two-hybrid assay, we would argue that recruitment of
corepressors suppresses the transcriptional activity of the VP16/PR
fusion protein even though interactions may occur efficiently between
GAL4/PR and VP16/PR hybrids in the presence of antagonists. In support
of this idea we observed that the ability of the VP16/PR fusion protein
to transactivate a PRE-controlled reporter gene in response to
antagonists was reduced as compared with agonist (Fig. 4). To resolve
this issue we turned to the PR-UP-1 mutants in the two-hybrid assay.
The VP16/PR-B(UP-1) fusion protein, in response to antagonist, mediated
transactivation of the PRE-controlled reporter gene to a similar extent
as agonist-induced transactivation mediated by the VP16 fusion protein
constructed with wild-type PR-B (VP16/PR-B). This suggested to us that
the PR-UP-1 mutant would be useful to detect antagonist-dependent PR
dimerization in whole cells by two-hybrid assay. This indeed was the
case. The progestin antagonists RU486 and ZK98299 both stimulated a
significant interaction between GAL4/PR-B(UP-1) and VP16/PR-B(UP-1) as
detected by the induction of CAT reporter gene activity over that
obtained by the single GAL4/PR-B(UP-1) hybrid in response to
antagonists. Although statistical analysis did not show a significant
stimulation, the antagonist ZK112993 also exhibited a trend of
stimulating interaction between the GAL4/PR-B(UP-1) and VP16/PR-B(UP-1)
fusion proteins (Fig. 5). These two-hybrid results support the
conclusion that antagonists induce PR dimerization in whole cells and
do not appear to act by impairing receptor dimerization.
Antiestrogens were reported to induce dimerization of human ER when
measured in a yeast two-hybrid assay (42). Both tamoxifen and ICI
182,780 were observed to induce ER-ER interactions. However, the level
of reporter ß-galactosidase (ß-gal) activity induced by tamoxifen
and ICI 182,780 was only 15% and 20%, respectively, of the activity
induced by estradiol-17ß (42). Additionally, it was shown that
dimerization is perturbed when yeast cells were cotreated with estrogen
and antiestrogens. Our data with PR antagonists would suggest that the
lower induction of ß-gal observed by antiestrogens is due to
suppression of the heterologous transcriptional activation domain of
the GAL4 AAD/ER construct as opposed to a reduction of ER-ER
interaction per se. It is interesting that
antagonist-induced ER dimerization was detected using wild-type ER as
the fusion protein constructs, whereas we observed no dimerization
using wild-type PR to construct hybrid proteins. Antiprogestin-induced
PR-PR interactions were detected only with PR-UP-1 mutants. This
apparent difference could be explained as a difference between yeast-
and mammalian-based two-hybrid assays since it appears that yeast does
not express an NCOR- or SMRT-like molecule (53, 57).
It is now well established that binding of PR to antagonists induces a
conformational change in the LBD of the receptor that is different from
that induced by agonist (31, 32, 33, 34, 35, 36, 37). These conformational differences were
initially suggested to prevent the formation of heterodimers between
receptors bound to agonist and receptors bound to antagonist. This was
based on EMSA experiments in which mixing of one hPR isoform liganded
with RU486 and another liganded with R5020 failed to produce an
intermediate heterodimer complex bound to a PRE (12). However, the
ability of RU486- and R5020-liganded receptors to form heterodimers in
solution in the absence of DNA was not examined (12). In a subsequent
study, we examined whether mixed agonist/antagonist heterodimers could
form in solution by a coimmunoprecipitation assay. We observed that
PR-A bound to RU486 can heterodimerize as efficiently with PR-B bound
to R5020 as the two PR isoforms bound to the same ligand: either R5020
or RU486 (45). Similar to results of Meyer et al. (12), we
also observed a dramatically reduced ability of mixed R5020/RU486
heterodimers to bind to PREs by EMSA (45). By use of ligand specificity
PR mutants, we show here that mixed R5020/RU486 PR heterodimers can
form in whole cells as detected by the mammalian two-hybrid assay (Fig. 7B). This suggests that antagonists do not alter the dimer interface
sufficiently to prevent dimerization with PR bound to agonist. In
addition, the fact that both R5020 and RU486 were needed to form these
heterodimers provides new information that both PR partners must be
bound to ligand in order for dimerization to occur; ligand binding to
one partner is not sufficient. In further support of the compatibility
of dimer interfaces is the crystal structure of the ER LBD bound to
estradiol and the antiestrogen raloxifene (58). The dimerization
interfaces appear to be similar for the LBD bound to estradiol and
raloxifene; the major conformational difference induced by agonist and
antagonist is the position of helix 12, which contains the conserved
sequences in AF-2 (58). The recently published crystal structure of PR
LBD bound to progesterone, and modeling with the antagonist RU486, also
suggest that dimerization interfaces are similar for the LBD bound to
progesterone and RU486 (59). Similar to ER, it was predicted that the
major difference between the PR LBD bound to progesterone and RU486 is
the displacement of helix 12 and the C-terminal tail.
The fact that PRs bound to RU486 and R5020 are capable of forming
heterodimers in vitro (in solution) and in whole cells may
aid in the interpretation of the mechanistic basis for the biological
potency of RU486 and other structurally related progestin antagonists.
All antagonists that we have tested, including ZK98299 after correction
for its 10-fold lower affinity than R5020 for PR, inhibit activation of
PR at substoichiometric concentrations with R5020 (Refs. 11, 38 and
Fig. 6). This suggests that these compounds do not work solely by a
simple mechanism of competing with agonist for binding to PR. One
possible mechanism to explain these biological results is that
receptors bound to antagonists can inactivate receptors bound to
agonist in trans by heterodimerization. The implication is
that all receptors do not need to be occupied by antagonist for
complete antagonism to occur. This is diagrammed schematically in Fig. 9, where the agonist/antagonist
heterodimer forms in solution and then is unable to bind efficiently to
a PRE.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-methyl-3H]17,
21-dimethyl-19-norpregna-4,9-diene-3, 20-one; 87 Ci/mmol) and unlabeled
R5020 were obtained from Dupont/NEN Research Products (Boston, MA).
[3H]RU486 (Mifapristone: [17-
hydroxy-11[4-dimethyl-aminophenyl]17-propenyl-estra-4,5-
diene-3-one) and unlabeled RU486 were gifts from R. Daraedt
(Roussel UCLAF, Romanville, France). ZK98299 (Onapristone), ZK112993,
ZK98734, and [3H]ZK98299 were provided by David Henderson
(Schering, Berlin, Germany). 1294/H9 is an unpublished MAb that
recognizes the A and B forms of hPR. MAbs B-30 and AB-52 to PR have
been described previously (60).
Construction of Two-Hybrid Fusion Plasmids for PR
The Mammalian MATCHMAKER Two-Hybrid Assay Kit was obtained from
CLONTECH (Palo Alto, CA). Included within the kit is the pM cloning
vector for expressing GAL4 DBD (aa 1–147) fusion proteins and pVP16
(aa 411–455) for expressing the VP16 transcriptional activation
domain. Also included is pG5CAT, a reporter vector that contains the
CAT gene downstream of five consensus GAL4 binding sites and the
minimal promoter of the adenovirus E1b gene. Human PR-B was cloned into
the pM and pVP16 vectors in the following manner. A 2.93-kb
AflII-Asp718 fragment of YEphPR-B (33) was filled
in with Klenow and inserted into the unique BamHI
restriction site of pGAD424 (CLONTECH) that had been blunt ended with
Klenow. This plasmid, pGAD424.PR-B, was used as the source of hPR-B
cDNA for fusion with the GAL4 DBD and the VP16 protein of the herpes
simplex virus. A 2.9-kb EcoRI-PstI fragment from
pGAD424.hPR-B was ligated into the EcoRI-PstI
restriction sites in the multiple cloning cassette of both the pM and
pVP16 vectors to yield, respectively, pM.PR-B (referred to subsequently
as GAL4/PR-B) and pVP16.PR-B (referred to subsequently as VP16/PR-B).
This places the GAL4 DBD and the activation domain of VP16 immediately
amino-terminal to hPR cDNA. Cloning junctures were sequenced (Sequenase
2.0, US Biochemical, Cleveland, OH) and PR-B was determined to be
correctly oriented and in-frame with GAL4 DBD or VP16. For PR-A
constructs, a 2.4 kb BspHI-Asp718 fragment of
YEphPR-B (33) was filled-in with Klenow and inserted into the unique
SmaI restriction site of pGBT9 (CLONTECH). This plasmid was
used as the source of hPR-A cDNA for fusion with the GAL4 DBD and the
VP16. A 2.4 kb EcoRI-PstI fragment from
pGBT9.hPR-A was cloned into the EcoRI-PstI of pM
and pVP16 to yield respectively pM.PR-A (referred to subsequently as
GAL4/PR-A) and pVP16.PR-A (referred to subsequently as VP16/PR-A).
Cloning junctures were sequenced (Sequenase 2.0, US Biochemical) and
hPR-A was determined to be correctly oriented and in-frame with the
GAL4 DBD or VP16.
A C-terminal truncation mutant of hPR-B (pSVhPR-B 891) was provided by Donald McDonnell (Duke Medical Center, Durham, NC). To construct fusions with hPR-B 891, a 2.8-kb BamHI fragment from pSVhPR-B 891 was inserted into GAL4/PR-B and VP16/PR-B vectors that had been previously digested with BamHI to drop out the 2.8-kb PR-B fragment. This hPR cDNA contains a reading-frame mutation that generates a stop codon at aa 892, resulting in a carboxyl-terminal truncated receptor by 42 aa (33). The resulting plasmids, GAL4/PR-B(UP-1) and VP16/PR-B(UP-1), were sequenced to confirm the presence of the nucleotide deletion at base 2849.
A two-hybrid vector expressing PR-B containing a point mutation at aa
722 was constructed by digesting VP16/PR-B with the restriction enzymes
StuI and EcoN I and replacing the resulting
1.3-kb fragment with a 1.3-kb StuI-EcoN I
fragment from pBK·CMVhPR-BRU (provided by Dawn Wen, Ligand
Pharmaceutical, San Diego, CA). pBK·CMVhPR-BRU contains
full-length hPR-B cDNA with a G to T substitution at nucleotide
position 2339. This mutation results in a GlyCys change at aa 722 in
the human PR-B amino acid sequence. This generates a receptor that has
lost detectable RU486 binding, but retains wild-type binding affinity
for progestin agonists and is fully activated by progestin agonists
(56). The resulting plasmid, VP16/PR-B(722m), was sequenced to confirm
the presence of the G to T substitution.
For PR transactivation experiments, the progesterone responsive reporter, DHRE-E1b-CAT, was used. This contains two synthetic PREs linked to the TATA box of the adenovirus E1b (61) and CAT (30).
Transient Transfections
Monkey kidney COS-1 cells were plated in DMEM (GIBCO-BRL) with
10% FBS at a density of 1.75 x 105 to 2.0 x
105 cells per well in six-well dishes (Falcon, Franklin
Lakes, NJ) at 37 C. After 24 h the cells were transfected
by an nonrecombinant adenovirus-mediated DNA transfer technique as
previously described (62). This method involves the use of a
replication-defective adenovirus that is coupled with
poly-L-lysine to bind the plasmid DNA noncovalently. This
process facilitates cellular uptake of the plasmid DNA by
receptor-mediated endocytosis. For mammalian two-hybrid experiments, 50
ng/well of plasmids encoding either GAL4 DBD or VP16 fusion proteins
were used and 500 ng/well of the pG5CAT reporter plasmid. In early
two-hybrid experiments, we used charcoal-treated serum and found that
it made no difference compared with whole FBS, so we used whole FBS for
our studies. For experiments to measure PR as a transactivator, the
DHRE-E1b-CAT was added at 500 ng/well, and the PR expression plasmid
was added at 10 ng/well. To treat cells with various PR-ligands,
24 h after transfection, the medium was replaced with DMEM + 10%
FBS containing vehicle (0.01% ethanol) or ligand at the concentration
indicated in the figure legends, followed by incubation at 37 C. After
an additional 24 h, the cell monolayers were rinsed with CAT wash
buffer (40 mM Tris-Cl, pH 7.4, 150 mM NaCl, and
1 mM EDTA) and lysed in the well by the addition of 300
µl of cell lysis buffer (20 mM potassium phosphate, pH
7.4, 5 mM MgCl2, and 0.5% Triton X-100). To
control for variation in transfection efficiency, cells were
cotransfected with a CMV-ß-gal reporter plasmid, and ß-gal was
measured in a luminometer (Monolight 2001) with the Galacto-Light Plus
kit (Tropix, Bedford, MA) according to manufacturer’s
instructions.
Whole-Cell Steroid-Binding Assay
COS-1 cells plated in six-well dishes at a density of 1.75
x 105 cells per well were grown overnight at 37 C in DMEM
+ 10% FBS. Cells were then transfected with a single concentration (25
ng) of the indicated expression plasmids by the adenovirus-mediated DNA
transfer technique as described above. PR-rich T47D human breast cancer
cells (60) were also plated in six-well dishes at a density of
1.75 x 105 cells per well in MEM + 5% FBS. After
44 h the cells were incubated for an additional 4 h at 37 C
with 1–2 nM of [3H]R5020,
[3H]RU486, or [3H]ZK98299 ± unlabeled
100 nM homologous ligand. Radioactive steroid was extracted
from cells with ethanol and quantitated by liquid scintillation
counting. Parallel transfected cells were lysed and assayed for protein
concentration. Receptor expression levels were normalized to protein
and calculated as picomoles of steroid binding per mg of protein.
Nonspecific binding was determined by parallel incubations of
mock-transfected COS-1 cells. Data represents specific binding taken as
total binding minus nonspecific binding.
Immunoblot Analysis
COS-1 cells were plated in six-well dishes at a density of
2 x 105 cells per well in DMEM + 10% FBS for 24
h. Wells were transiently transfected by the adenovirus method (62)
with 25 ng of the appropriate plasmid (indicated in the figure
legends). After 44 h, the medium was removed and replaced with
medium containing either 100 nM of R5020 or 100
nM of RU486. After an additional 4 h at 37 C, the cell
monolayers were washed with CAT wash buffer, and cells from triplicate
wells were harvested and combined in 0.4 ml of 0.5% Triton X-100 lysis
buffer. The lysates were removed from the wells and centrifuged at
12,000 rpm for 10 min at 4 C. The supernatants were assayed for protein
concentration by Bradford assay (63). Total protein in 1% SDS sample
buffer (250 µg) was boiled and electrophoresed on 7.5%
polyacrylamide SDS gels. Separated proteins were transferred to
nitrocellulose and incubated with the receptor-specific (1 µg/ml)
MAb 1294/H9. Immune complexes were decorated with rabbit
antimouse immunoglobins (N. L. Cappel Laboratories, Cochranville,
PA) and [35S] protein A (Amersham) and detected by
autoradiography.
T47D human breast cancer cells were plated in six-well dishes at a density of 2 x 105 cells per well in MEM (GIBCO-BRL) with 5% FBS. At 44 h after plating, the cells were treated with 100 nM R5020. Cells from triplicate wells were harvested in a total of 0.4 ml of 0.5% Triton X-100 lysis buffer and prepared as the COS-1 lysates.
CAT Assays
CAT enzyme activity was assayed by a radiometric/organic phase
extraction method as previously described (64). Enzyme activity was
calculated as counts per min of [3H]acetylcoenzyme A
converted per µg of protein in the cell lysate. CAT assays were done
in duplicate for each lysate. Cell treatment groups were also performed
in duplicate. Protein concentration was measured by the method of
Bradford (63), and equal amounts of protein (10–30 µg) were added to
the assay. Specific CAT activity was determined by subtraction of assay
background obtained from lysates of nontransfected cells. To compensate
for variation in transfection efficiency, the cells were also
transfected with a CMV-ß-gal reporter plasmid. CAT results were
calculated as the ratio of CAT activity per unit of ß-gal
activity.
Baculovirus Expression of Human PR-A and PR-B
Human PR-A and PR-B were expressed as nonfusion full-length
proteins from baculovirus vectors in Sf9 insect cells as previously
described (44). To bind ligands to baculovirus-expressed receptors,
they were added at 200 nM to Sf9 cell cultures during the
last 6–8 h of expression. To prepare whole-cell extracts, Sf9 cells
were lysed in TEDG buffer [10 mM Tris-base, pH 7.4, 1
mM EDTA, 1 mM dithiothreitol, and 10%
glycerol], containing 0.5 M NaCl and a mixture of protease
inhibitors (60). Cell lysates were centrifuged at 100,000 xg for 30 min to yield a soluble supernatant and then dialyzed
against lysis buffer containing no NaCl.
Coimmunoprecipitation Assay
Protein A-Sepharose was prebound noncovalently with
receptor-specific MAbs and used as an immunoabsorbent as described
previously (28, 45). Resins were prebound to rabbit anti-mouse IgG
(Cappel) and used as a bridging antibody. Receptor-specific MAbs were
then bound to the immobilized rabbit antimouse IgG. Sf9 cell extracts
containing PR-A or PR-B were mixed at 10 µg/100 µl of resin in
siliconized microcentrifuge tubes and incubated on ice for 30 min.
Equal amounts of receptors were added to each assay as determined by
ligand-binding assay and immunoblot analysis. MAb-coated protein
A-Sepharose beads (100 µl) were added to each tube and incubated at 4
C for 1 h on an end-over-end rotator. Resins were then washed four
times by centrifugation in TEG (TEDG minus dithiothreitol) containing
100 mM NaCl, transferred to a new microcentrifuge tube, and
washed twice more. Immobilized proteins were eluted with 2% SDS
loading buffer and then analyzed by immunoblot, using
[35S]protein A and autoradiography as the detection
method.
EMSAs
For EMSA, we used a 28-bp oligonucleotide containing a
PRE/glucocorticoid response element derived from the MMTV-long terminal
repeat (11, 65). Sf9 insect cell extracts containing PR-A or PR-B were
incubated for 1 h at 4 C with 32P-labeled DNA (0.3 ng)
in a total reaction volume of 25 µl. Also included was 1 µg of poly
(dA-dT) as nonspecific competitor DNA. The DNA binding buffer contained
10 mM Tris-base, pH 7.4, 50 mM NaCl, 5
mM dithiothreitol, 2 mM MgCl2, 10%
glycerol, and 50 ng/ml of a carrier protein. Samples (25 µl) were
electrophoresed on 5% polyacrylamide gels prepared at 40:1 (wt/wt)
acrylamide-bis-acrylamide ratio using 20 mM Tris-acetate,
0.5 mM EDTA as the electrode buffer. To maintain constant
temperature during electrophoresis, 4 C water was recirculated through
the gel apparatus. Gels were dried and subjected to autoradiography
(65).
T47D Breast Cancer Cells Stably Transfected with MMTV-CAT to
Measure Biological Activity of Progestin Antagonists
T47D human breast cancer cells were cultured as described
(11) in MEM supplemented with 5% FBS. A cloned derivative (B-11) of
T47D, stably transfected with a construct containing the CAT gene
linked to the MMTV promoter/enhancer, has been previously described
(11). Cells plated in six-well dishes at a density of 0.5 x
106 cells per well were grown for 3 days at 37 C and then
incubated for another 24 h with medium stripped of steroid
hormones by treatment with dextran-coated charcoal (11). This was
followed by an additional 24 h incubation with hormone and other
compounds in the same medium. Harvested cells were washed twice in 40
mM Tris (pH 7.6), 150 mM NaCl, and 1
mM EDTA and lysed in buffer containing 20 mM
potassium phosphate (pH 7.8), 5 mM MgCl2, and
0.5% (vol/vol) Triton X-100. Cell lysates were analyzed for CAT enzyme
activity as described above.
Data Analysis
For two-hybrid experiments presented in Figs. 3, 5, and 7, CAT
activity mediated by the GAL4/PR construct as a single hybrid in the
absence of ligand was normalized to 1.0, and the CAT activity mediated
by other constructs and construct pairs was calculated relative to the
single-hybrid control. These data are expressed as relative CAT
activity. The two-hybrid CAT data represent mean values
(±SEM) from four to seven independent experiments. To
calculate the fold CAT induction mediated by PR dimerization in
response to ligand, we used the following calculation:
 |
|
ACKNOWLEDGMENTS |
|---|
This work was supported in part by Public Health Services Grant R01DK-49030 (to D.P.E.) and National Research Service Award fellowship F32DK-09662 (to S.A.L.).
|
FOOTNOTES |
|---|
Received for publication December 22, 1997. Revision received August 3, 1998. Accepted for publication September 8, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-substitued analogs of
the antiprogestin RU486 induce a unique conformation in the human
progesterone receptor resulting in mixed agonist activity. Proc Natl
Acad Sci USA 93:8739–8744This article has been cited by other articles:
 |
|
 |
|
  N. Z. Lu, S. E. Wardell, K. L. Burnstein, D. Defranco, P. J. Fuller, V. Giguere, R. B. Hochberg, L. McKay, J.-M. Renoir, N. L. Weigel, et al. International Union of Pharmacology. LXV. The Pharmacology and Classification of the Nuclear Receptor Superfamily: Glucocorticoid, Mineralocorticoid, Progesterone, and Androgen Receptors Pharmacol. Rev., December 1, 2006; 58(4): 782 - 797. [Full Text] [PDF]
|
|
|
|
 |
|
 A. K. Iyer, Y.-H. Zhang, and E. R. B. McCabe Dosage-Sensitive Sex Reversal Adrenal Hypoplasia Congenita Critical Region on the X Chromosome, Gene 1 (DAX1) (NR0B1) and Small Heterodimer Partner (SHP) (NR0B2) Form Homodimers Individually, as Well as DAX1-SHP Heterodimers Mol. Endocrinol., October 1, 2006; 20(10): 2326 - 2342. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Shao, E. Egecioglu, B. Weijdegard, K. Ljungstrom, C. Ling, J. Fernandez-Rodriguez, and H. Billig Developmental and hormonal regulation of progesterone receptor A-form expression in female mouse lung in vivo: interaction with glucocorticoid receptors. J. Endocrinol., September 1, 2006; 190(3): 857 - 870. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. K. Mani, A. M. Reyna, J. Z. Chen, B. Mulac-Jericevic, and O. M. Conneely Differential Response of Progesterone Receptor Isoforms in Hormone-Dependent and -Independent Facilitation of Female Sexual Receptivity Mol. Endocrinol., June 1, 2006; 20(6): 1322 - 1332. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Ma, P. Katiyar, L. P. Jones, S. Fan, Y. Zhang, P. A. Furth, and E. M. Rosen The Breast Cancer Susceptibility Gene BRCA1 Regulates Progesterone Receptor Signaling in Mammary Epithelial Cells Mol. Endocrinol., January 1, 2006; 20(1): 14 - 34. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. C. M. Simmen, R. R. Eason, J. R. McQuown, A. L. Linz, T.-J. Kang, L. Chatman Jr., S. R. Till, Y. Fujii-Kuriyama, F. A. Simmen, and S. P. Oh Subfertility, Uterine Hypoplasia, and Partial Progesterone Resistance in Mice Lacking the Kruppel-like Factor 9/Basic Transcription Element-binding Protein-1 (Bteb1) Gene J. Biol. Chem., July 9, 2004; 279(28): 29286 - 29294. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Botos, W. Xian, D. F. Smith, and C. L. Smith Progesterone Receptor Deficient in Chromatin Binding Has an Altered Cellular State J. Biol. Chem., April 9, 2004; 279(15): 15231 - 15239. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X.-L. Zhang, D. Zhang, F. J. Michel, J. L. Blum, F. A. Simmen, and R. C. M. Simmen Selective Interactions of Kruppel-like Factor 9/Basic Transcription Element-binding Protein with Progesterone Receptor Isoforms A and B Determine Transcriptional Activity of Progesterone-responsive Genes in Endometrial Epithelial Cells J. Biol. Chem., June 6, 2003; 278(24): 21474 - 21482. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Shao, E. Markstrom, P. A. Friberg, M. Johansson, and H. Billig Expression of Progesterone Receptor (PR) A and B Isoforms in Mouse Granulosa Cells: Stage-Dependent PR-Mediated Regulation of Apoptosis and Cell Proliferation Biol Reprod, March 1, 2003; 68(3): 914 - 921. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Leonhardt and D. P. Edwards Mechanism of Action of Progesterone Antagonists Experimental Biology and Medicine, December 1, 2002; 227(11): 969 - 980. [Abstract] [Full Text]
|
|
|
|
 |
|
 D. Zhang, X.-L. Zhang, F. J. Michel, J. L. Blum, F. A. Simmen, and R. C. M. Simmen Direct Interaction of the Kruppel-like Family (KLF) Member, BTEB1, and PR Mediates Progesterone-Responsive Gene Expression in Endometrial Epithelial Cells Endocrinology, January 1, 2002; 143(1): 62 - 73. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Giannoukos, D. Szapary, C. L. Smith, J. E. W. Meeker, and S. S. Simons Jr. New Antiprogestins with Partial Agonist Activity: Potential Selective Progesterone Receptor Modulators (SPRMs) and Probes for Receptor- and Coregulator-Induced Changes in Progesterone Receptor Induction Properties Mol. Endocrinol., February 1, 2001; 15(2): 255 - 270. [Abstract] [Full Text]
|
|
|
|
 |
|
 P. H. Giangrande, E. A. Kimbrel, D. P. Edwards, and D. P. McDonnell The Opposing Transcriptional Activities of the Two Isoforms of the Human Progesterone Receptor Are Due to Differential Cofactor Binding Mol. Cell. Biol., May 1, 2000; 20(9): 3102 - 3115. [Abstract] [Full Text]
|
|
|
|
 |
|
 D. P. Edwards, S. A. Leonhardt, and E. Gass-Handel Novel Mechanisms of Progesterone Antagonists and Progesterone Receptor Reproductive Sciences, January 1, 2000; 7(1_suppl): S22 - S24. [Abstract] [PDF]
|
|
|
|
|
|
C. Depoix, M.-H. Delmotte, P. Formstecher, and P. Lefebvre Control of Retinoic Acid Receptor Heterodimerization by Ligand-induced Structural Transitions. A NOVEL MECHANISM OF ACTION FOR RETINOID ANTAGONISTS J. Biol. Chem., March 16, 2001; 276(12): 9452 - 9459. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
