Human Progesterone Receptor A

Form Is a Cell- and Promoter-

Specific Repressor of Human
Progesterone Receptor B Function

Elisabetta Vegeto, Manouchehr M. Shahbaz, Dawn X. Wen,

Mark E. Goldman, Bert W. O’Malley, and Donald P. McDonnell

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

Ligand Pharmaceuticals Inc. (E.V., M.M.S., D.X.W., M.E.G., D.P.M.)
San Diego, California 92121
Department of Cell Biology (E.V., B.W.O.)
Baylor College of Medicine
Houston Texas 77030
Universita degli Studi di Milan0 (E.V.)
lnstituto Di Science Farmacologiche
20133 Milan, Italy

Two distinct isoforms of the human progesterone the existence of two distinct forms of the hPR. (Mo-
receptor (hPR-A and hPR-B) have been identified lecular Endocrinology 7: 1244-1255, 1993)
previously. They differ only in that hPR-B contains
an additional 164 amino acids at the amino terminus.
Among various species these two forms arise as a INTRODUCTION
result of either alternate initiation of translation from
the same mRNA or by transcription from alternate
promoters within the same gene. In order to under- The steroid hormone progesterone is a potent hormonal
stand the reason for their existence, we studied the effector implicated in the control of proliferation, differ-
entiation, and development of mammary and uterine
transcriptional capacity of these individual receptors
tissues (1). The endocrine effects of this hormone are
and observed that their activity was influenced
manifested only in cells containing a specific intranu-
strongly by cell and promoter context. More surpris-
clear receptor, the progesterone receptor (PR). The
ing was the observation that in promoter and cell
interaction between PR and its cognate ligand induces
contexts where hPR-A was inactive, it acted as a
a series of structural and functional changes in the
potent frans-dominant repressor of hPR-B-mediated
protein, leading ‘ultimately to an association of the re-
transcription. This event occurred at substoichiom-
ceptor with specific DNA sequences in the regulatory
etric concentrations of hPR-A and was hormone
regions of target genes. The cellular and promoter
dependent. Human PR-A was not a general repres- context of the bound receptor determines the pheno-
sor of ligand-mediated transcription, as it had no typic consequence of this interaction (2, 3).
effect on vitamin D receptor function. Interestingly, The PR is a member of a closely related subgroup of
hPR-A but not hPR-B was capable of a similar inhi- nuclear receptors that includes the androgen, mineral-
bition of glucocorticoid, androgen, and mineralocor- ocortiroid, and glucocorticoid receptors (AR, MR, and
ticoid receptor-mediated gene transcription. This tiic, respectively) (4). Within this subgroup, human PR
suggests a specific role for the hPR-A isoform in this (hPR) is unique in that it occurs in target tissues as two
regulatory process. The trans-dominant effects of distinct subtypes, PR-A and PR-B, of 94 and 114
hPR-A were induced also by the antiprogestins kilodaltons, respectively (5, 6). The PR-B isoform con-
ZK112993 and ZK98299 and a DNA binding defective tains an N-terminal fragment of 164 amino acids (B164),
hPR-A mutant, suggesting that the inhibitory function which is absent on the PR-A isoform. It is likely that
of hPR-A does not require DNA binding. The dual both forms can arise as a result of either alternate
role of hPR-A as an activator or repressor of tran- initiation of translation from the same mRNA or by
scription defines a potential mechanism by which transcription from alternate promoters within the same
cells can generate dissimilar responses to a single gene (7, 8). Interestingly, Kastner et al. (8) have identi-
hormone and provides a molecular explanation for fied two distinct promoters in the hPR gene. These
promoters, which regulate the synthesis of specific
0688-8809/93/1244-1255$03.00/O transcripts corresponding to hPR-A and hPR-B, are
Molecular Endocrinology
Copyright 0 1993 by The Endocrine Society regulated independently. It is likely, therefore, that the

1244
hPR-A Effects on hPR-B Function 1245

expression levels of PR-A and PR-B can differ with between PR-A and PR-B in the regulation of progester-
respect to each other in certain target tissues. Two one-responsive genes.
forms of PR, corresponding to hPR-A and hPR-B, have
been identified in most species examined, the exception
being the rabbit, where PR may exist as a single unique RESULTS
B subtype. The specific role for each of these two PR
subtypes is unclear. However, the existence of elabo- Differential Transcriptional Activity of the A and B
rate mechanisms regulating their production suggests lsoforms of hPR
that the relative levels of hPR-A and hPR-B in cells is
critical for appropriate cellular response to progester- In order to define the activities of the hPR-A and hPR-

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

one.
The biochemical properties of the PR isoforms have
been analyzed extensively in vitro, and, interestingly,
they displayed similar DNA and hormone binding affini-
ties (9, 10). However, when analyzed in reconstituted
progesterone response systems in heterologous cells
it became apparent that hPR-A and hPR-B have differ-
ent promoter specificities (8, 11). A similar result was
obtained when the transcriptional activities of chicken
PR-A and PR-B were assessed (12, 13). Two distinct
regions within hPR required for transcriptional activa-
tion (TAFs) have been identified, TAFl located in the
amino terminus and TAF2 within the carboxyl terminus.
Interestingly, both TAFs are contained within PR-A and
PR-B. The 8164 region, unique to hPR-B, does not
contain additional transcriptional activators but is re-
quired for maximal TAFl function in the context of the
full-length receptor (14). It is possible, therefore, that in
cell and promoter contexts where TAFl activity is re-
quired, PR-B will be a more efficient transcriptional
regulator than PR-A.
The precise mechanism by which the promoter-
bound receptor exerts its transcriptional effect is un-
clear at present. The reconstitution of steroid receptor-
dependent transcription in vitro has been informative in
this regard (15). On chromatin free templates it is clear
that at least one of the functions of the receptor is to
recruit and/or stabilize the transcription preinitiation
complex at the core promoter. It is probable, however,
that in the context of the intact cell, additional factors
and processes in unison with the activated receptor are
required for appropriate function. The existence of
adaptors or cofactors which influence the interaction
between bound receptor and the general transcription
apparatus has been implicated by several studies (12,
16). It is likely that these cofactors are differentially
expressed in cells and in the case of the PR may be
the determinants of the cellular and promoter context
preferences of the PR-A and PR-B. Recent genetic
studies in yeast have identified two distinct nuclear
proteins, SSN6 and SW12, which have profound effects
on steroid receptor transcriptional activity (17, 18). The
mechanism of action of the mammalian homologs of
these proteins has not yet been defined.
In an attempt to define the evolutionary significance
and the specific cellular functions of the individual PR
isoforms, we have performed a systematic analysis of
the promoter and cellular preferences of these proteins.
We extended our analysis to determine the interplay
B isoforms we used the expression vectors phPR-A
and phPR-B that encode exclusively either PR-A or PR-
B. These expression constructs were transiently trans-
fected into either CV-1 (monkey kidney fibroblasts),
HeLa (human epithelial), or HepG2 (human hepatocel-
lular carcinoma) cells together with the progesterone-
responsive mouse mammary tumor virus-luciferase re-
porter (MMTV-LUC). Western immunoblot analysis us-
ing two PR-specific monoclonal antibodies confirmed
that no hPR-A was synthesized from our PR-B con-
structs.
In CV-1 cells, the endogenous level of PR is low. As
a result, there was no significant hormone-dependent
activation of the MMTV promoter in the absence of
transfected receptor (Fig. 1 A). Transfection of increas-
ing amounts of the phPR-B expression vector permitted
progesterone-mediated activation of the MMTV pro-
moter, the degree of which was proportional to input
plasmid. In contrast, in the same cell background, no
progesterone-induced activation of the MMTV pro-
moter by PR-A was observed. These results indicated
that progesterone-induced activation of the MMTV pro-
moter was regulated differentially by hPR-A and hPR-
B, in agreement with observations made previously
using both chicken and human PR-A and PR-B (8, 11,
12,13).
The influence of cell type on PR subtype-specific
activation of MMTV was evaluated by performing ad-
ditional cotransfection experiments in HeLa (Fig. 1B)
and HepG2 cells (Fig. 1C). Interestingly, PR-A was
transcriptionally active on this promoter in HeLa cells;
its efficacy was 15% that of PR-B, reaching a maximal
lo-fold induction at 250 ng transfected DNA (Fig. IB).
A similar transcription assay performed in the HepG2
cell line revealed that in this cell line both PR-A and PR-
B functioned as efficient activators of the MMTV pro-
moter (Fig. 1 C).
To eliminate the possibility that the differential regu-
latory activity of the two PR subtypes was an artifact
of their expression level, we performed hormone bind-
ing analysis of extracts prepared from transfected CV-
1 cells. The expression levels of PR-A and PR-B were
determined to be 178 fmol/mg and 35 fmol/mg respec-
tively, indicating that PR-A was expressed approxi-
mately five times as well as PR-B. In repeated experi-
ments, in different cell lines we observed no more than
20% variation in receptor levels. lmmunoblot analysis
confirmed the expression of intact receptors of the
correct molecular weight and that no detectable PR-A
was produced by the PR-B expression plasmid (data
MOL ENDO. 1993 Vol7No.10
1246

A not shown). Thus, it was unlikely that receptor expres-

18 _ cv-1 sion level was a significant factor in these studies.
16 _ q Vehicle We conclude, therefore, that cell context influences
n PROG PR-A- and PR-B-mediated activation of the MMTV pro-
cn 14-
.?t moter, implicating the existence of additional transcrip-
s 12- tional cofactors that may influence their transcriptional
2 IO- efficiency.
-1
9 8-
‘;, Dominant Negative Effect of PR-A on PR-B
_m 6-
B 4-

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

We have shown that PR-A and PR-B have distinct cell-
2- specific activities when assayed individually on the
0-h *A MMTV promoter. However, since these two subtypes
-- are coexpressed in endometrial and breast cancer cell
-R hPR-A hPR-B lines (19), it was of interest to determine if PR-A func-
tioned as a modulator of PR-B function. To address
B this question we assayed PR-B activity in the presence
of increasing concentrations of PR-A. In CV-1 cells, PR-
90 , HeLa B but not PR-A allowed hormone-dependent regulation
of the MMTV promoter (Fig. 2A). Surprisingly, the pro-
gesterone-dependent activation obtained when equal
amounts of phPR-A and phPR-B expression vectors
were transfected into these cells was only 7% that of
PR-B alone (Fig. 2A). Decreasing the concentration of
PR-A restored the PR-B-dependent, progesteronein-
duced transcriptional activity such that at A:B DNA
ratios of 15 or I:25 the recovery of PR-B-mediated
transcription was 13% and 57%, respectively. Even at
the lowest concentration of phPR-A transfected there
was a strong reduction in PR-B activity. The progester-
-- one concentration required for half-maximal inhibition
-R hPR-A hPR-B was lo-’ M, a concentration that corresponds to the
affinity of the receptor for ligand (data not shown). The
C efficiency of this repression revealed a dominant nega-
tive role of PR-A in the progesterone-mediated regula-

10
60 He@2
Vehicle
50 n PROG
--
-R hPR-A hPR-B
Fig. 1. Transcriptional Activity of PR-A and B lsoforms on the
MMTV Promoter in Different Mammalian Cell Lines
Increasing amounts of the phPR-A or phPR-B plasmid DNA
(0.05, 0.1, 0.25, 0.5, 2.5, and 5 fig) together with 5 fig/ml
MMTV-LUC reporter DNA and 5 @g/ml pCHll0 (an SV40-p-
galactosidase expression vector) as an internal control, were
transiently transfected into CV-1 (A), HeLa (B), or HepGP (C)
ceil lines as described in Materials and Methods. Cells were
treated with or without lo-’ M progesterone as indicated for
24 h and assayed for fl-galactosidase and LUC activity. (LUC
activity was normalized to @-galactosidase activity.) The rela-
tive LUC activity is calculated by dividing the normalized LUC
tion of MMTV transcription in CV-1 cells.
We wanted next to evaluate the role of PR-A as a
repressor of PR-B-activated MMTV promoter transcrip-
tion in HeLa cells where PR-A was found to be minimally
active. As shown in Fig. 28, the transcriptional repres-
sion was 84%, 78%, and 56% at l:l, 1:5, 1:25 A:B
ratios, respectively, confirming that PR-A-mediated in-
hibition of PR-B activity correlated with the inability of
PR-A to trans-activate. As expected, coexpression of
PR-A and PR-B in HepG2 cells, where these receptor
isoforms were shown to be independently active, re-
sulted in an overall additive effect of the coexpressed
receptors on MMTV transcription (data not shown).
To eliminate the possibility that inhibition of PR-B
activity was due to an alteration of the levels of the
individual activators, we performed a Western immu-
noblot and in vitro hormone binding analysis on extracts
prepared from transfected CV-1 cells. The results indi-
at a given point by that obtained in the absence of transfected
receptor or ligand. Expression of PR-A or PR-B had no direct
effect on S/40-driven P-galactosidase activity. A representa-
tive experiment is detailed above. Data shown indicate the
mean + the average deviation from the mean of triplicate
estimations. -R, Transcriptional activity in the absence of any
transfected receptor.
hPR-A Effects on hPR-6 Function 1247

transcriptional activator, it acts as a hormone-depend-

ent trans-dominant repressor of hPR-B function in con-
texts where PR-A activity is minimal.

Promoter Specificity of the Dominant Negative

Effect of PR-A

The previous experiments demonstrated that cell con-

text determines whether PR-A acts as a transcriptional
activator or repressor when assayed on the MMTV

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

promoter. To determine if inhibition of transcription by
PR-B - - .25 .25 .25 .25 PR-A was a general phenomenon or was restricted to
PR-A - 25 - .Ol .05 .25 the MMTV promoter, PR-A and PR-B function were
&vPW evaluated on other PR-responsive promoters. For this
B purpose we chose the PRE2tk-LUC reporter plasmid,
which contains two copies of a progesterone response
70 lHeLa
n PROG element (PRE) linked to the thymidine kinase (tk) pro-
60 moter. In CV-1 cells, PR-B was transcriptionally active,
” whereas PR-A was transcriptionally inactive (Fig. 3A).
5 50
When phPR-B was transfected together with phPR-A,
we observed a repression of PR-B activity similar to
that observed on the MMTV promoter. This demon-
strated that PR-A-mediated repression of PR-B function
in CV-1 cells is not promoter specific. Similar results
10 were obtained when the same experiment was per-
1
n+ formed in HeLa cells (data not shown). In contrast,
PR-B - - .25 .25 .25 .25 when the identical promoter and receptor combinations
PR-A - .25 - .Oi .05 .25 were assayed in HepG2 cells, PR-A was found to be
(wPW an activator of transcription and had no effect on PR-B
Fig. 2. hPR-A is a Ceil-Specific Trans.-Dominant Repressor of transcriptional activity (Fig. 3B), similar to what was
hPR-6 Function observed using the MMTV promoter (Fig. 1C). To-
CV-1 (A) or HeLa (B) cells were transiently transfected with gether, these data suggested that PR-A-specific down-
either 0.25 pg phPRA or 0.25 pg phPR-B alone or 0.25 pg regulation of transcription could be extended to another
phPR-B in the presence of increasing concentrations of phPR- progesterone-responsive promoter and that this phe-
A together with 5 pg MMTV-LUC reporter and 5 pg pCHll0
nomenon maintained cell specificity.
as an internal control. Cells were treated with or without 1 O-’
Our data indicated that PR-A-regulated transcrip-
M progesterone as indicated for 24 h and assayed for @-
galactosidase and LUC activity. (LUC activity was normalized tional repression was a cell-specific event. It was im-
for P-galactosidase activity.) The relative LUC activity is cal- portant to assess whether PR-A was inactive on all
culated by dividing the normalized LUC value at a given point progesterone-responsive promoters in CV-1 and HeLa
by that obtained in the absence of transfected receptor or cells, suggesting that a factor required for PR-A function
ligand. A single experiment representative of three independ- was absent, or whether in these cellular contexts PR-
ent experiments is detailed above. Data shown indicate the A could function as an activator of some promoters. To
mean + average deviation from the mean of triplicate estima- address this question we extended our analysis to the
tions.
progesterone-responsive tyrosine amino transferase
(TAT) promoter. Significantly, in HeLa cells PR-A but
not PR-B functioned as a hormone-dependent regulator
cated that coexpression of PR-A and PR-B using SV40 of the TAT promoter (Fig. 3C). At this concentration of
promoter-based expression plasmids did not alter the expressed receptor, PR-A promoted a 1 O-fold induction
level of either receptor (data not shown). Reproducibly, of TAT promoter activity. Importantly, this result indi-
PR-A was expressed about three to five times as well cated that PR-A was not a general repressor of PR-B-
as PR-B. Therefore at equimolar concentrations of PR- mediated transcription in HeLa cells, but rather inhibited
A and PR-B we observe greater than 90% inhibition of transcription in a promoter-specific manner. In support
MMTV-LUC transcriptional activity in CV-1 cells. In of this hypothesis we have observed that in this context
addition, identical results were obtained when using PR-B is not an inhibitor of PR-A function, attesting to
receptor expression vectors that used a Rous sarcoma the specific function of PR-A as a transcriptional re-
virus promoter. This indicates that the effects of PR-A pressor.
are not peculiar to the particular expression system Further analysis of TAT promoter regulation revealed
used. that it responded to progesterone in HepG2 cells after
These experiments identified a unique regulatory transfection of either phPR-A or phPR-B (data not
function for PR-A. In addition to its accepted role as a shown). In CV-I cells the TAT promoter was unrespon-
MOL ENDO. 1993 Vol7No.10
1248

sive to either receptor subtype (data not shown). These

results highlighted further the existence of promoter-
and cell-specific events that converge on the receptor-
mediated signal transduction pathway to modulate its
outcome.

Dominant Negative Effects of PR-A on GR Function

Since the MMTV promoter has been shown to respond

to glucocorticoids (20, 21) we asked whether PR-A

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

could antagonize the GR-mediated response. To ex-
PR-B - - .25 .25 .25 .25 amine this, we established a GR-responsive MMTV
PR-A - .25 - .Oi .05 .25 transcription unit in transfected CV-1 cells (22). HeLa
(wDW
cells were not used for these experiments, as they
B contain a high level of endogenous GR. An expression
vector encoding GR (pRShGR) was transfected into
CV-1 cells together with the MMTV-LUC reporter. Half-
maximal activation of MMTV by dexamethasone-acti-
vated GR occurred at 250 ng transfected DNA (data
not shown). In subsequent experiments a constant
amount (250 ng) of GR expression vector was trans-
fected with equimolar or decreasing amounts of phPR-
A DNA together with the MMTV reporter (Fig. 4A).
Transcriptional activity was measured after four differ-
ent hormonal stimuli; vehicle alone, dexamethasone,
progesterone, or dexamethasone and progesterone.
Pi=-A -
This allowed the independent measurement of recep-
(wDN4
tor-specific responses at each different receptor ratio.
As shown in Fig. 4A, there was negligible hormone-
C dependent induction of transcription in cells transfected
14 - HeLa (TATA2950) with an empty control vector, representing most likely
12
the transcriptional activity of a low level of endogenous
GR. When GR was transfected alone, a 450-fold induc-
tion of MMTV promoter activity was detected in the
presence of dexamethasone or in the presence of dex-
amethasone and progesterone in combination. In this
context PR-A was minimally active on its own. Interest-
ingly, in the absence of progesterone, PR-A decreased
the dexamethasone induction of GR activity to about
53% of control levels. Hormone-independent transcrip-
tional interference was less apparent at lower concen-
- PR-A PR-0 trations of receptor, suggesting most likely that this
Fig. 3. Human PR-A Repressor Function is Both Cell Type activity related to an overexpression of PR-A. In the
and Promoter Context Dependent presence of progesterone and dexamethasone, GR
CV-1 (A) or HepG2 (B) cells were transfected with either activity decreased dramatically to 2% of induced control
0.25 pg/ml phPR-8, phPR-A alone, or phPR-B in the presence levels. Even at the lowest concentration of PR-A (0.01
of increasing concentrations of phPR-A, together with 5 pg/ml pg), we observed a significant hormone-dependent in-
PRE2tk-LUC reporter and 5 fig/ml pCHll0 as an internal hibition of GR function.
control. C, HeLa cells were transfected with either phPR-B or To determine the specificity of PR-A-mediated
phPR-A together with 5 pg/ml pTATA2950 reporter and 5 pg/ repression of GR function in this cell and promoter
ml pCH110 as an internal control. Cells were treated with or context, a similar assay was performed in the presence
without lo-’ M progesterone for 24 h and assayed for LUC
of PR-B. Vectors expressing PR-B and GR were co-
and /3-galactosidase activity. The relative LUC activity is cal-
culated by dividing the normalized LUC value at a given point transfected into CV-1 cells with the MMTV-LUC re-
by that obtained in the absence of transfected receptor or porter, and receptor activity was assayed under the
ligand. The data shown are representative of several experi- same conditions as in the previous experiment. Inter-
ments performed. The data shown represent the mean + estingly, as shown in Fig. 4, A and B, GR-mediated
average deviation from the mean of triplicate estimations. activation of MMTV promoter activity was about 10
times greater than PR-B and 50 times greater than PR-
A in this context. We reasoned, therefore, that if PR-A-
mediated repression was due solely to its ability to
hPR-A Effects on hPR-B Function 1249

240 A
CV-1 cells were transfected with MMTV-LUC and either

0)
.+E
200

=I 160
; GR alone or GR in combination with phPR-A or phPR-
B in the presence of 5 x 1Om8 M dexamethasone
increasing concentrations of progesterone.
and
As shown

#
0
3 120
L in Fig. 5, progesterone had no significant effect (12%

n$
increase) on GR activity when transfected alone. How-
.-F
53 80 ever, when GR and PR-A were transfected together
2 there was a 50% reduction in GR activity at lo-’ M
40

0 : /I
I.I progesterone. These experiments suggested that PR-
A’s activity as a repressor occurred at physiological

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

-II IllI
concentrations of hormone.
GR - .25 - .25 .25 .25
PR-A - - .25 .Ol .05 .25
(WNA) hPR-A is not a General Repressor of Nuclear
0 “ehlcle Receptor Transcriptional Activity
w OH
250 ,D q PROG In order to analyze further the specificity of PR-A action
we examined its ability to affect other receptor-depend-
ent transcription systems. For this purpose, phPR-A
was cotransfected with a vector expressing the vitamin
D receptor (VDR) and assayed on a VDRE,tk-LUC
reporter. On a similar promoter containing two PRE
sequences, PR-A functioned as a repressor of PR-B
activity in CV-1 cells (Fig. 3A). As shown in Fig. 6,
hormone-dependent trans-activation by VDR was not
substantially affected by coexpression of PR-A either
uuu uuu in the absence or presence of progesterone. The minor
GR - .25 - .25 .25 .25
PR-B - - .25 .Ol
reduction noted at the higher concentrations of PR-A
.05 .25
most likely related to nonspecific squelching of a tran-
(I.IW’JN
scription factor required for both receptors and is un-
Fig. 4. Human PR-A but not hPR-B is an Efficient Repressor likely to relate to the function of PR-Aper se. In addition,
of GR Function we observe no effects of PR-A expression on the
CV-1 cells were transfected with either pRShGR, phPR-A, transcriptional activity of the SV-40 promoter (assayed
phPR-B, or pRShGR in the presence of increasing concentra- as a fusion to /3-galactosidase; data not shown). These
tions of phPR-A (A) or pRShGR in the presence of increasing
concentrations of phPR-B (B). In addition, 5 pg/ml MMTV-LUC
and 5 pg/ml pCHll0 were included in all transfections. Cul-
tures were treated with or without lo-’ M dexamethasone
(DEX), progesterone (PROG), or DEX and PROG for 24 h and
assayed for LUC and fl-galactosidase activity. The data shown
are representative of several identical experiments. The data
shown are the mean values k the average deviation from the
mean of triplicate determinations.

displace GR and occupy the response elements of the -

MMTV promoter with a less efficient activator, then PR- 0-l
Blank 11 10 9 8 7 6
B should also decrease GR-activated transcription to a
level corresponding to PR-B maximal activity. However, -Log[M] Progesterone

as shown in Fig. 46, coexpression of PR-B had minimal

Fig. 5. The Repressor Function of hPR-A Occurs at Physiolog-
effects on GR function under the conditions examined.
ical Concentrations of Progesterone
Therefore, progesterone-mediated repression of MMTV CV-1 cells were transfected with either 0.25 pg/ml pRShGR
transcription appeared to be specific for the A subtype alone or 0.25 pg/ml pRShGR + 0.05 pg/ml phPR-A together
of PR, and as such it was unlikely that repression of with 5 Fg/ml MMTV reporter and 5 gg/ml pCH110 as an
GR was a consequence of occupancy of the GRE by a internal control. The cells were incubated for 24 h with 5 X
less efficient trans-activator. 1 O-’ M dexamethasone and increasing concentrations of pro-
gesterone as indicated and assayed for P-galactosidase and
LUC activity. The data are presented as percent activation
The Repressor Function of hPFi-A Occurs at
where the 100% value represents maximally activated GR in
Physiological Concentrations of Progesterone the presence of 5 x 1 Om8 M dexamethasone alone (>450-fold
normalized response). The data shown represent the mean
The next aspect we wanted to define was the profile of values + the coefficient of variation of quadruplicate estima-
inhibition with respect to progesterone concentration.
 MOL END0.1993 Vol7No.10
1250

A
160 1 +5 x 10~8M DEX T

20 n =H
q GRthPR.4
0
VDR - .25 .25 .25 .25 Blank 11 10 9 8 7 6

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

PR-A - - .25 .Ol .05 .25
(@NA) -Log[M] ZK 96299

B
Fig. 6. Human PR-A is not a General Repressor of Nuclear 140, +5x IO-8M DEX ,
Receptor Transcriptional Activity
CV-1 cells were transfected with either an expression vector
producing the human VDR (pRShVDR) or hPR-A alone or with
both vectors in combination. In addition, the cells were trans-
fected with 5 pg/ml VDREztk-LUC reporter plasmid and 5 pg/
ml pCHll0 as an internal control. Cells were treated with
either 1 O-’ M 1 ,25(OH)nD3, progesterone, or both hormones in
combination as indicated. The data shown represent the mean
value + the average deviation from the mean of triplicate
..\
determinations. I-
Blank 11 10 9 6 7 6

-Log[M] ZK 112993

data suggested a selective role for PR-A in the regula-

Fig. 7. Antiprogestins Efficiently Induce hPR-A Repressor
tion of steroid receptor transcriptional activity. Function
CV-1 cells were transfected with either pRShGR alone or
Repression of PR-B Function by PR-A Does Not pRShGR and phPR-A together with 5 pg/ml MMTV-LUC re-
Require DNA Binding porter and 5 pg/ml pCHll0 as an internal control. The cells
were treated with 5 x 1 O-’ M dexamethasone and increasing
In order to define more precisely the mechanism of PR- concentrations of ZK98299 (A) or ZK112993 (B) as indicated.
A-mediated repression, it was important to determine The cells were harvested after 24 h, and the LUC and p-
whether inhibition required that PR-A be bound to DNA. galactosidase activities were measured. The results are pre-
Therefore, we specifically examined the role of DNA sented as percent activation, where 100% represents the
binding in this process. maximal activity of GR in the presence of 5 x lo-* M dexa-
methasone (>450-fold normalized response). The data shown
In the first series of experiments we evaluated the
represent the mean f the coefficient of variation of quadrupli-
ability of PR-A to repress GR transcriptional activation
cate estimations.
of MMTV-LUC in the presence of two dissimilar antipro-
gestins. One compound, ZKI 12993, promotes the as-
sociation of PR with DNA, whereas a second, ZK98299,
interferes with DNA binding, possibly by preventing used a mutant of PR-A, PR-A587. In this mutant, two
dimerization of the receptor (23, 24; our unpublished point mutations were created in the DNA binding do-
data). In initial experiments the effects of ZK112993 main of PR-A, replacing the critical cysteine 587 residue
were examined (Fig. 78). At high concentrations, this with an alanine. The PR-A587 mutant was not capable
compound directly inhibited GR function, but cointrod- of binding DNA in vitro, as demonstrated by gel-shift
uction of PR-A, but not PR-B (data not shown), greatly experiments, and was expressed at the same level as
potentiated this inhibition (>I OOO-fold). This result sug- the wild type protein (data not shown). Results from
gested that PR-A could function as a transcriptional the in vivo competition assay using this receptor mutant
repressor in the presence of either hormone agonists are reported in Fig. 8. The transcriptional activity of PR-
or antagonists. It suggested further that this class of B was reduced by 90% when equimolar amounts of
compounds had the potential to function as potent phPR-A587 and phPR-B DNA were cotransfected into
antiglucocorticoids in a relevant though indirect manner CV-1 cells. Apparently, therefore, the binding of PR-A
via PR-A. This experiment was repeated in the presence to DNA is not an obligate step for the inhibitory activity
of ZK98299. The results of this analysis, shown in Fig. of PR-A. Using this type of mutant receptor it is possible
7A, indicated that this compound also mediated an that some of the inhibition observed may result from
inhibition of GR function on the MMTV promoter by PR- the formation of nonfunctional hPR-A/hPR-B heter-
A. These results suggested that DNA binding may not odimers.
be required for PR-A-mediated repression of transcrip- These data strongly indicated that DNA binding was
tion. not required for the PR-A to function as a transcriptional
In order to examine this question more directly, we repressor. It is important, however, that even though
hPR-A Effects on hPR-I3 Function 1251

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

0
PR-B - .25 - - .25 .25 .25 .25 25 .25
PR-As8, - - .25 - .Ol .05 .25 - - - 0
PR-A - - - .25 - - - .Ol .05 .25 NOH PROG DHT ZK993 ZK299

MPJA)

Fig. 8. Repressor Function of hPR-A Does Not Require DNA B

c
Binding
150 -
CV-1 cells were transfected individually with either phPR-B,
phPR-A587
increasing
or phPR-A
concentrations
tion, the cells were transfected
reporter and 5 fig/ml pCHll0
alone, or phPR-B in combination
of phPR-A587 or phPR-A.
with 5 wg/ml MMTV-LUC
as an internal control.
were grown in the presence or absence of 1 O-’ M progesterone
with
In addi-

Cells
.8=
5
125

ioo-
L
.2 75-
for 24 h. All values were normalized for transfection efficiency 2
by simultaneous estimation of LUC and P-galactosidase activ- 8 50-
ities. The relative LUC activity is calculated by dividing the
normalized LUC value at a given point by that obtained in the 25 -
absence of transfected receptor or ligand. Data shown repre-
sent the mean + the average deviation from the mean of O-
AR AR + hPR-A
triplicate estimations.
Fig. 9. Human PR-A is a Repressor of AR Transcriptional
Activity
DNA binding is not required, PR-A functions as a pro- CV-1 cells were transfected with an expression vector for
moter- and cell-specific transcriptional repressor. the hAR (pRShAR) alone or in combination with phPR-A to-
gether with 5 fig/ml MMTV-LUC reporter and 5 Kg/ml pCHll0
Modulation of AR and MR Activity as an internal control. A, The cells were treated with lo-’ M
progesterone (PROG), 5 x 1 O-’ M dihydrotestosterone (DHT),
The GR and PR are members of a subfamily of steroid lo-’ M ZK112993 (ZK993), or lo-’ M ZK98299 (ZK299) for
receptors that include also the AR and MR, all of which 24 h. The relative LUC activity is calculated by dividing the
can bind to and operate through similar DNA regulatory normalized LUC value at a given point by that obtained in the
sequences (4). In order to determine whether the activ- absence of transfected receptor or ligand. B, Cells were grown
ity of AR or MR could also be modulated by PR-A, we in the presence of 5 x lo-’ M DHT alone or in the presence
assayed their ability to activate the MMTV promoter in of 1 O-’ M progesterone, ZKll2993, or ZK98299 as indicated.
The data are presented as percent activation, where 100%
the presence of PR-A. This analysis was performed in
represents the maximal induction achieved by hAR at 5 x 1 O-’
the presence of either progesterone, ZKI 12993, or
M DHT. The experimental data represent mean values + the
ZK98299. As shown in Fig. 9A, the activity of AR average deviation from the mean of quadruplicate estimations.
induced by dihydrotestosterone (DHT) was influenced
directly by addition of progesterone. However, when
AR activity was assayed in the presence of PR-A, the A similar series of experiments was performed to
sensitivity of progesterone-mediated inhibition was sub- examine the effect of PR-A on MR regulation of MMTV-
stantially increased (Fig. 9B). Additional experiments LUC transcription and showed that PR-A efficiently
were performed in the presence of antiprogestins. Sur- inhibits MR transcriptional activity in a hormone-de-
prisingly, ZK112993 was an efficient inhibitor of DHT- pendent manner (data not shown). Cumulatively, these
activated AR. The efficacy of this antihormone (and data suggested that PR-A is capable of inhibiting the
potency; data not shown) was increased by the intro- transcriptional activity of all members of the GR subfam-
duction of PR-A into the cells. Similarly, the antiproges- ily of receptors.
tin ZK98299 had minimal direct effects on AR transcrip-
tional activity, but the addition of PR-A to the cells
increased AR’s sensitivity to this compound. This result DISCUSSION
reveals a mechanism whereby ZK98299, although hav-
ing a low affinity for AR, could in this cell and promoter This study describes a functional role for the A form of
context function as an efficient antiandrogen (Fig. 9B). the hPR (PR-A) that is distinct from that of PR-B. Using
MOL ENDO. 1993
1252
cell cotransfection assays we observed that in promoter
and cell contexts, where PR-A was inactive as a tran-
scriptional activator, it functioned as an extremely po-
tent, ligand-dependent repressor of PR-B transcrip-
tional activity. In most cases, the PR-A effect was
dominant, since repression was detected even at sub-
stoichiometric concentrations of PR-A. Interestingly,
PR-A functions also as a transcriptional activator in
certain cellular contexts. Most notable was its activity
in HeLa cells, where PR-A effectively stimulated the
tyrosine amino transferase promoter in a hormone-
dependent manner. Our results clearly demonstrate the
context dependence of PR function and suggest that
the interaction of the receptor with additional cell-spe-
cific proteins or processes are the key determinants of
the nature of receptor activity. The fact that PR-A
functions also as a repressor of MR, AR, and GR
function suggests that all members of this subgroup of
proteins may require similar transcriptional cofactors for
activity depending on the specific promoter and cellular
context. Therefore, the PR gene contains the potential
to encode both activators (PR-B or PR-A) or a repressor
(PR-A) of transcription. Differential biosynthesis of these
two receptor isoforms could represent a mechanism
employed to create functionally diverse responses to
progesterone in defined cells. In addition, fluctuations
in the PR-A-to-PR-B ratio in target cells may provide
developmental states whereby the genetic response to
hormones could be drastically altered.
Several examples of single genes encoding
activators and repressors have been defined (25). In
fact, the paradigm established for the transcriptional
activator LAP resembles closely that observed for PR.
In the case of LAP, alternate initiation of translation
results in the production of full-length LAP and a trun-
cated version, LIP, which functions as a repressor of
LAP function (26). Similarly, it has been shown that the
A and B isoforms of PR are produced either by tran-
scription from two distinct promoters within the same
gene (8) or as a result of alternate initiation of translation
from a single mRNA (7). An additional example of this
motif is provided by the mouse transcription factor
mTFE3, where a physiological role for two opposing
transcription factors produced from the same gene has
been postulated (27). The transcription factor TFE3 by
alternate splicing gives rise to two functionally opposite
proteins, TFES-L and TFES-S, the ratios of which vary
from tissue to tissue. In vitro, it has been shown that
subtle alterations in the level of TFE3-S profoundly
affect TFES-L function, suggesting a similar conse-
quence in the intact animal.
In the case of PR-A, however, its repressive capacity
may be much more widespread in that it appears to
have the ability to down-regulate GR, MR, and AR
responses also, at least under selected conditions. The
absolute and relative levels of PR-A and B expressed
in most normal human tissues are unknown, making it
difficult as yet to define a clear physiological role for
PR-A as an activator or a repressor (19). Studies in the
chick have identified variations in the levels of PR-A
Vol7No.10
and PR-B as a consequence of endocrine manipulations
(28, 29). In the mammary gland of the mouse the levels
of mPR-A and mPR-B vary depending on the develop-
mental state of the tissue (30). Additionally, seasonal
variations in the expression of the PR isoforms in
chicken reproductive tissues also have been identified
(31). Although the functional consequences of these
variations have not been described yet, taken together
with our results, we suggest that minor alterations in
the relative concentrations of the two PR isoforms could
Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021
have a dramatic overall effect on MR-, GR-, AR-, and
PR-B-mediated transcription of certain promoters. It is
quite possible that during development, or in a specific
cell type, an altered ratio of these transcription factor
isoforms may result in the alteration of the cellular
phenotype.
The PR-A receptor exhibits its dominant repressive
effect on all members of the GR subfamily of nuclear
receptors in a promoter- and cell-specific manner. We
were unable to observe any effects on a VDR-mediated
regulation of a VDRE-tk promoter. The transcriptional
activity of the SV-40 or Rous sarcoma virus promoters
were likewise unaffected by PR-A expression. The re-
stricted nature of this regulatory process defines differ-
ences in the mechanism of action of the individual
members of the nuclear receptor superfamily and sug-
gests the requirement of specific cofactors for this
subgroup of steroid receptors.
Two lines of investigation demonstrated that repres-
both sion by PR-A occurred independently of DNA binding.
First a mutant of PR-A (PR-A587); bearing a point
mutation in the DNA binding domain, functioned as an
efficient repressor of PR-B-mediated frans-activation.
Second, the antiprogestin ZK98299, which binds to PR
and retards DNA binding, promotes PR-A repressor
function.
Promoter and cell specificity coupled with the ob-
served stoichiometry indicate that the mechanism for
PR-A-mediated repression differs from the more gen-
eral squelching or transcriptional interference event that
occurs as a result of overexpression of some transcrip-
tion factors. Ligand-dependent, DNA binding-independ-
ent, cross-interference between the estrogen receptor
and PR or GR has been reported previously (32). In
addition, estrogen receptor overexpression effectively
down-regulates PRL gene expression, a process oc-
curring also independently of DNA binding (33). How-
ever, in contrast with the mechanism encountered for
PR-A repression, antagonists of the interfering receptor
do not promote squelching. Significantly, the trans-
dominant effects of PR-A are induced by all agonists
and antagonists of the PR.
It is difficult to define precisely the mechanism of
action of PR-A. However, we propose that on pro-
moters where PR-B (or a member of this sub-family) is
transcriptionally active and PR-A is not, PR-A binds
with greater affinity but in a nonproductive manner to a
limiting cofactor required for PR-B function. The require-
ment for this interaction ultimately would be determined
by cell context, such that the same promoter could be
hPR-A Effects on hPR-B Function
differentially activated, depending on the cellular com-
plement of transcription factors. On other promoters
where both A and B are active, there may be no
requirement for such additional cofactors, or a different
class of transcriptional factors used by both isoforms
are recruited, or the quantities of common cofactors
are in vast excess.
Several types of negative regulation cf transcription
have been shown to involve members of the nuclear
receptor superfamily. In addition to transcriptional inter-
ference or squelching described above, additional
mechanisms of repression also have been observed
(11, 32, 33, 34). The autologous regulation of thyroid
hormone receptor (TR) function by splicing variants of
TRcul and TRn2 (35) most closely resembles that
observed with PR-A and PR-B. However, the regulatory
activity of TRcv2 on TRal function probably represents
a dominant negative effect resulting from the formation
of nonfunctional heterodimers or competition for similar
DNA-regulatory elements. Transcriptional interference
among homologous receptors, resulting from the for-
mation of nonproductive heterodimers or competition
at the DNA level, has been shown to occur between
TR, the retinoid acid receptor, VDR, v-e&a, and chicken
ovalbumin upstream promoter transcription factor (36-
39). Additionally, the demonstration of a functional in-
teraction of GR and APl revealed a mechanism by
which nuclear receptors can regulate transcription
41). The stoichiometry of the observed inhibition by PR-
A (>90% inhibition at equimolar concentrations of PR-
A and PR-B) and the suggestion by other groups that
antiprogestin-activated PR-A is unable to heterodimer-
ize with agonist-activated hPR-B would seem to rule
out heterodimerization as the sole process involved in
repression (11). Clearly, the mechanism of action of
PR-A is distinct from those mechanisms defined previ-
ously and demonstrates additional functional flexibility
among members of the receptor superfamily.
In the experiments described in this study PR-A, in
the presence of progesterone or antiprogesterone,
terferes dominantly with all other members of the GR
subfamily of receptors. Especially interesting is the
observation that ZK98299, which does not directly
effect AR-dependent gene transcription, acts as a po-
tent AR antagonist when PR-A is present. Much effort
has been expended of late to develop synthetic antipro-
gestins that bind specifically to PR. However, this work
suggests that direct binding of the compound to PR-A
may result in an indirect interference of the transcrip-
tional activity of other nuclear receptors and promote
repression of function of many genes driven by other
steroid hormones, i.e. glucocorticoids, mineralocorti-
coids, and androgens. Therefore, it may not be possible
using conventional strategies to develop a clean anti-
progestin, since it is likely that antagonist activated PR-
A could suppress glucocorticoid or androgenic
tions in certain cell types.
Our results substantiate the conclusion that nuclear
receptors are a class of transcription factors that can
act as both activators and repressors of transcription.
1253
In the case of the hormone progesterone, a series of
cellular processes converge on the isoforms of its cog-
nate receptor and the resultant combination of their
effects determine its transcriptional activity. Therefore,
PR function can be regulated at any of the following
levels: 1) progesterone concentration, as hormone pro-
duction is tightly regulated by endocrine signals and
pathways; 2) PR gene transcription, which generates
both activators and repressors of transcription; 3) re-
ceptor activation, for example, phosphorylation, which
Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021
could differentially effect PR-A and PR-B; 4) receptor
binding to DNA; 5) receptor interactions with transcrip-
tional cofactors, which could be regulated in a cell-
specific manner; and 6) cross-interference between ste-
roid receptors of the same class. It is clear that no one
mechanism can accommodate the phenotypic response
of the cell to steroid hormones, but rather the cell has
evolved multiple regulatory processes that act alone or
in concert to modulate cellular functions.
MATERIALS AND METHODS
Chemicals
Restriction and modification enzymes were obtained from
Promega Biotec (Madison, WI), Boehringer Mannheim (Indi-
anapolis, IN), or New England Biolabs (Beverly, MA). PCR
(40, reagents were obtained from Perkin Elmer Cetus (Norwalk,
CT). [l ,2-3H]Progesterone (47.3 Ci/mmol) was purchased from
Amersham (Arlington Heights, IL). Chemicals were purchased
from Sigma (St. Louis, MO). Secondary and alkaline phospha-
tase-conjugated antibodies for Western analysis were ob-
tained from BioRad (Richmond, CA); Immobilon-P (polyvinyld-
ifluoridene) transfer membranes were purchased from Millipore
(Bedford, MA).
Construction of the Mammalian Expression Vectors
Encoding the A Subtype of the hPR (phPR-A) and the
DNA Binding Mutant hPR-A5B7
The plasmid phPR-B, containing the cDNA for the B subtype
in- of the hPR under the control of the SV40 enhancer/human
metallothionein-II promoter (42) was digested with BarnHI. Of
the three resulting DNA fragments [5.1,2.7, and 0.24 kilobases
(kb)] only two (5.1 and 2.7 kb) were isolated. The ligation of
the correctly oriented fragments resulted in a plasmid, called
phPR-A, which was identical to phPR-B except it lacked the
245 bases that contain the ATG sequence for the B isoform.
For the construction of the DNA binding mutant of hPR-A, we
followed the procedure described by others (24); cystine 587
was substituted with an alanine by PCR using primers carrying
a double-point mutation (5’-3’-CCTGTGGGAGCGCTAAGGT-
CTTC and counterpart) synthesized by National Biosciences
Inc. (Plymouth, MN). The presence of the double-point muta-
tion in the receptor molecule was verified by DNA sequencing.
The construction of the plasmids pRShGR, pRShMR, and
pRShVDR have been described previously (43, 44, 45). These
expression vectors use the Rous sarcoma virus promoter and
SV40 polyadenylation signal.
func- Construction of the Repotter Plasmids PRE,tk-LUC and
TATA2950-LUC
For the construction of PRE&LUC, the MMTV long terminal
repeat promoter sequence of MMTV-LUC reporter plasmid
(22) was substituted with the PRE,tk sequence from PRE,tk-
MOL ENDO. 1993 Vol7No.10
1254

chloramphenicol acetyltransferase (CAT) [PRE2tk-CAT con- excitement predicted for the future. Mol Endocrinol
tains two copies of the PRE of the tyrosine aminotransferase 41363-369
TAT promoter linked to the herpes simplex virus tk promoter 3. Beato M 1989 Gene regulation by steroid hormones. Cell
(42)]. Both reporter plasmids were digested with Xhol and 561335-344
HindIll; a 6.7-kb fragment from MMTV-LUC and a 0.12-kb 4. Evans RM 1988 The steroid and thyroid receptor super-
fragment from PRE2tk-CAT were isolated and ligated, resulting family. Science 240:889-895
in a plasmid called PRE2tk-LUC. 5. Schrader WT, O’Malley BW 1972 Progesterone-binding
For the construction of TATA2950-LUC, the pTATCAT, a components of chick oviduct: characterization of purified
plasmid containing the 2950-base fragment of the TAT gene subunits. J Biol Chem 247:51-59
promoter sequence (a gift from Dr. Gunther Schutz, German 6. Horwitz KB, Alexander PS 1983 In situ photolinked nu-
Cancer Research Center, Heidelberg, Germany), and the clear progesterone receptors of human breast cancer
MMTV-LUC plasmids were digested with Sac1 and Xhol, re- cells: subunit molecular weights after transformation and

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

spectively, blunt-ended with T4 DNA polymerase and then translocation. Endocrinology 113:2195-2201
digested with HindIll. The 2.9-kb fragment resulting from the 7. Conneely OM, Maxwell BL, Toft DO, Schrader WT, O’Mal-
digestion of the pTAT-CAT plasmid, and the 6.7-kb fragment, ley BW 1987 The A and B forms of the chicken proges-
containing the LUC gene and the backbone resulting from terone receptor arise by alternate initiation of translation
MMTV-LUC digestion products, were ligated to create the of a unique mRNA. Biochem Biophys Res Commun
plasmid TATA2950-LUC. 149:493-501
8. Kastner P, Krust A, Turcotte B, Stropp U, Tora L, Gro-
Cell Culture nemeyer H, Chambon P 1990 Two distinct estrogen-
regulated promoters generate transcripts encoding the
two functionally different human progesterone receptor
Monkey kidney CV-1 fibroblasts and human endometrial HeLa forms A and B. EMBO J 9:1603-1614
cells were routinely maintained in Dulbecco’s modified Eagle’s
9. Lessey BA, Alexander PS, Horwitz KB 1983 The subunit
medium (Biowhittaker, Walkersville, MD) supplemented with
structure of human breast cancer progesterone receptors:
10% fetal bovine serum (Hyclone Laboratories, Logan, UT).
characterization by chromatography and photoaffinity la-
Human hepatoma HepG2 cells were maintained in Eagle’s
beling. Endocrinology 112:1267-l 274
minimal essential medium containing 10% fetal bovine serum.
10. Christensen K. Estes PA. Onate SA. Beck CA. DeMarzo
A, Altmann Ml Lieberman BA, St John J, Nordeen SK,
Transient Transfection Assays Edwards DP 1991 Characterization and functional prop-
erties of the A and B forms of human progesterone
Cells were seeded in 12-well, 96-well, or 1 O-cm tissue culture receptors synthesized in a baculovirus system. Mol En-
plates. DNA was introduced into cells using calcium phosphate docrinol 5:1755-l 770
coprecipitation (22). Twenty micrograms of DNA per ml trans- 11. Meyer ME, Pornon A, Ji J, Bocquel MT, Chambon P,
fection buffer were used in each transfection reaction. In this Gronemeyer H 1990 Agonist and antagonist activities of
mix, the concentration of the LUC plasmids and that of the RU486 on the functions of the human progesterone re-
internal control plasmid (pCH110, which contains the gene for ceptor. EMBO J 9:3923-3932
the P-galactosidase enzyme) remained constant (5 pg each 12. Tora L, Gronemeyer H, Turcotte B, Gaub MP, Chambon
plasmid DNA), while the receptor plasmid concentration varied P 1988 The N-terminal region of the chicken progesterone
as indicated for each experiment. Different amounts of recep- receptor specifies target gene activation. Nature
tor parental plasmid pSV2-neo were included to keep constant 333:185-l 88
the total amount of the SV40 enhancer vectors. pGEM4 plas- 13. Bocquel MT, Kumar V, Stricker C, Chambon P, Grone-
mid DNA was added to balance the total DNA concentration meyer H 1989 The contribution of the N- and C-terminal
to 20 pglreaction. For the 96-well plate experiments, transfec- reqions of steroid receptors to activation of transcription
tions were performed on a Biomek 1000 automated laboratory is -both receptor and cell-specific. Nucleic Acids Res
work station (Beckman, Fullerton, CA), and cells were incu- 17:2581-2595
bated with the precipitate for 6 h. Cells were washed with 14. Meyer ME, Qurin-Stricker C, Lerouge T, Bocquel MT,
PBS and incubated for 24 h with or without hormones as Gronemeyer H 1992 A limiting factor mediates the differ-
indicated in the text. Cell extracts were prepared as previously ential activation of promoters by the human progesterone
described (22) and assayed for LUC and fl-galactosidase receptor isoforms. J Biol Chem 267:10882-l 0887
activities, 15. Klein-Hitpass L, Tsai SY, Weigel NL, Allan GA, Riley D,
Rodriauez R. Schrader WT. Tsai MJ. O’Mallev BW 1990
The progesterone receptor stimulates cell-free transcrip-
Acknowledgments tion by enhancing the formation of a stable pre-initiation
complex. Cell 60:247-257
The authors would like to thank Abby Esty (Ligand Pharma- 16. Shemshedini L, Ji J, Brou C, Chambon P, Gronemeyer H
ceuticals) for her help in the design of the experiments and 1992 In vitro activity of the transcription activation func-
Dale Mais for performing hormone binding analysis. tions of the proqesterone receptor. J Biol Chem
267:1834-l 839’ -
17. McDonnell DP, Vegeto E, O’Malley BW 1992 Identification
Received May 4, 1993. Revision received June 29, 1993. of a negative regulatory function for steroid receptors.
Accepted July 7, 1993. Proc Natl Acad Sci USA 89:10563-l 0567
Address requests for reprints to: Dr. Donald P. McDonnell, 18. Yoshinaga SK, Peterson CL, Herskowitz I, Yamamoto KR
Ligand Pharmaceuticals Inc., 9393 Towne Centre Drive, San 1992 Roles of SWll, SWl2, and SW13 proteins for tran-
Diego, California 92121. scriptional enhancement by steroid receptors. Science
258:1598-l 604
19. Horwitz KB 1992 The molecular biology of RU486: is
there a role for antiprogestins in the treatment of breast
REFERENCES cancer. Endocr Rev 13: 146-l 63
20. Cato ACB, Skroch P, Weinmann J, Butkeraitis P, Ponta
H 1988 DNA sequences outside the receptor-binding sites
1. Clark CL, Sutherland RL 1990 Progestin regulation of differentially modulate the responsiveness of the mouse
cellular oroliferation. Endocr Rev 11:266-301 mammary tumor virus promoter to various steroid hor-
2. O’Malley BW 1990 The steroid receptor superfamily: more mones. EMBO J 7:1403-1410
hPR-A Effects on hPR-B Function 1255

21. Strahle U, Klock G, Schutz G 1987 A DNA sequence of 34. Adler AJ, Danielsen M, Robins DM 1992 Androgen-spe-
15 base pairs is sufficient to mediate both glucocorticoid cific gene activation via a consensus glucocorticoid re-
and progesterone induction of gene expression. Proc Nat1 sponse element is determined by interaction with nonre-
Acad Sci USA 84:7871-7875 ceptor factors. Proc Natl Acad Sci USA 89:11660-l 1663
22. Berger TS, Parandoosh Z, Perry BW, Stein RB 1992 35. Koenig RJ, Lazar MA, Hodin RA, Brent GA, Larsen R,
Interaction of glucocorticoid analogues with the human Chin W, Moore DD 1989 Inhibition of thyroid hormone
glucocorticoid receptor. J Steroid Biochem Mol Biol action by a non-hormone binding c-erbA protein generated
411733-738 by alternate mRNA splicing. Nature 337:659-661
23. Klein-Hitpass L, Cato ACB, Henderson D, Ryffel GU 1991 36. Umesono K, Giguere V, Glass CK, Rosenfeld MG, Evans
Two types of antiprogestins identified by their differential RM 1988 Retinoic acid and thyroid hormone induce gene
action in transcriptionally active extracts from T47D cells. expression through a common responsive element. Na-
Nucleic Acids Res 19:1227-l 234 ture 336:262-265
24. Takimoto GS, Tasset DM, Eppert CA, Horwitz KB 1992 37. Glass CK, Holloway JM, Devary OV, Rosenfeld MG 1988

Downloaded from https://academic.oup.com/mend/article/7/10/1244/2714673 by guest on 05 December 2021

Hormone-induced progesterone receptor phosphorylation The thyroid hormone receptor binds with opposite tran-
consists of sequential DNA-independent and DNA-de- scriptional effects to a common sequence motif in thyroid
pendent stages: analysis with zinc finger mutants and the hormone and estrogen response elements. Cell 54:313-
progesterone antagonist ZK98299. Proc Natl Acad Sci 323
USA 89:3050-3054 38. Sande S, Sharif M, Chen H, Privalsky M 1993 v-erbA acts
25. Foulkes NS, Sassone-Corsi P 1992 More is better: acti- on retinoic acid receptors in immature avian erythroid
vators and repressors from the same gene. Cell 68:411- cells. J Virol 67:1067-l 074
414 39. Cooney AJ, Tsai SY, O’Malley BW, Tsai MJ 1992 Chicken
26. Desombres P, Schibler U 1991 A liver-enriched transcrip- ovalbumin upstream promoter transcription factor COUP-
tional activator protein, LAP, and a transcriptional inhibi- TF dimers bind to different GGTCA response elements,
allowinq COUP-TF to repress hormonal induction of the
tory protein, LIP are translated from the same mRNA. Cell
67:569-579 vitamin-D3, thyroid hormone and retinoic acid receptors.
Mol Cell Biol 12:4153-4163
27. Roman C, Cohn L, Calame K 1991 A dominant negative
40. Schule R, Rangarajan P, Kliewer S, Ransone L, Bolado J,
form of transcription activator mTFE3 created by differ-
Yang N, Verma IM, Evans RM 1990 Functional antago-
ential splicing. Science 254:94-97
nism between oncoprotein c-Jun and the glucocorticoid
28. Boyd-Leinen P, Fournier D, Spelsberg TC 1982 Nonfunc-
receptor. Cell 62:1217-l 226
tioning progesterone receptors in the developed oviducts 41. Yang-Yen HF, Chambard JC, Sun YL, Smeal T, Schmidt
from estrogen-withdrawn immature chicks and in aged
TJ, Drouin J, Karin M 1990 Transcriptional interference
nonlaying hens. Endocrinology 111:30-36 between c-Jun and the glucocorticoid receptor: mutual
29. Boyd-Leinen P, Gosse B, Rasmussen K, Martin-Dani G, inhibition of DNA binding due to direct protein-protein
Spelsberg TC 1984 Regulation of nuclear binding of the interaction. Cell 62:1205-l 215
avian oviduct progesterone receptor. J Biol Chem 42. Vegeto E, Allan GF, Schrader WT, Tsai MJ, McDonnell
259:241 l-2421 DP, O’Malley BW 1992 The mechanism of RU 486 antag-
30. Shyamala G, Schneider W, Schott D 1990 Developmental onism is dependent on the conformation of the carboxy-
regulation of murine mammary progesterone receptor terminal tail of the human progesterone receptor. Cell
gene expression. Endocrinology 126:2882-2889 69:703-713
31. Boyd PA, Spelsberg TC 1979 Seasonal changes in the 43. Arriza JL, Weinberger C, Cerelli G, Glaser TM, Handelin
molecular species and nuclear binding of the chicken BL, Housman DE, Evans RM, 1987 Cloning of human
oviduct progesterone receptor. Biochemistry 18:3685- mineralocorticoid receptor complementary DNA: struc-
3690 tural and functional kinship with glucocorticoid receptor.
32. Meyer ME, Gronemeyer H, Turcotte B, Bocquel MT, Science 237:268-275
Tasset D, Chambon P 1989 Steroid hormone receptors 44. Giguere V, Hollenberg SM, Rosenfeld MG, Evans RM
compete for factors that mediate their enhancer function. 1986 Functional domains of the human glucocorticoid
Cell 571433-442 receptor. Cell 46:645-652
33. Adler S, Waterman ML, He X, Rosenfeld MG 1988 Steroid 45. Umesono K, Murakami KK, Thompson CC, Evans RM
receptor-mediated inhibition of rat prolactin gene expres- 1991 Direct repeats as selective response elements for
sion does not require the receptor DNA-binding domain. the thyroid hormone, retinoic acid, and vitamin D3 recep-
Cell 52:685-695 tors. Cell 65:1255-l 266